# CiteAgentic - Full Corpus > CiteAgentic is an AI visibility platform for marketing teams and SEO agencies. It monitors brand mentions and citations, identifies competitor gaps, and turns findings into prioritized content and website improvements. Source: https://www.citeagentic.com --- # AI visibility platform for marketing teams and SEO agencies Track brand mentions, AI citations and competitor visibility. Turn AI search gaps into prioritised fixes and content with CiteAgentic’s AI visibility platform. ## Find the gap See which buyer questions name or cite your brand, which competitors appear, and which sources AI answers use. An overall visibility score can hide individual questions where buyers miss you. ## Choose the next move Connect visibility findings with content and website recommendations. Review the evidence, strengthen the relevant page, and use briefs and drafts to help your team do the work. ## Measure what changes Monitor visibility over time and revisit the questions you worked on. Keep brand mentions, recommendations and website citations distinct. A change in visibility does not by itself prove that a particular fix caused it. ## Product examples Homepage demos illustrate product workflows with fictional example data. They are not customer results or measured claims about the competitors shown. ## Coverage and plans CiteAgentic supports AI visibility monitoring, AEO audits and SEO monitoring. Supported engines include ChatGPT, Perplexity and Gemini. Google AI Overviews is a separate search signal. Engine access, frequency, site allowances and usage limits depend on the plan. Compare current [plans and pricing](https://www.citeagentic.com/pricing), or [book a demo](https://www.citeagentic.com/contact). ## Explore - [AI citation tracking](https://www.citeagentic.com/features/ai-citation-tracking) - [AEO tracker](https://www.citeagentic.com/features/aeo-tracker) - [AEO audits](https://www.citeagentic.com/features/aeo-scan) - [AI SEO and SEO services](https://www.citeagentic.com/services/digital-marketing/aiseo) --- # CiteAgentic documentation > CiteAgentic documentation — what each feature does, how to connect your CMS, analytics and search data, how to use the REST API and MCP server, and how to manage your account and sites. Source: https://www.citeagentic.com/docs _Documentation_ ## CiteAgentic documentation CiteAgentic finds the buyer questions you lose in AI answers, tells you why you lose them, and drafts the content that wins them back. These guides cover how to run it, connect it, and build on it. **TL;DR:** **TL;DR:** CiteAgentic's documentation covers four areas: (1) [Features](/docs/features) — what each screen does, (2) [Integrations](/docs/integrations) — nine CMS publishing channels plus Search Console, Analytics and scanner access, (3) [API and MCP](/docs/api) — a five-endpoint REST API and a 27-tool MCP server, (4) [Account and settings](/docs/account) — plan, team, credits and per-site configuration. Every page is available as Markdown by appending `?md`. ## What does CiteAgentic's documentation cover? There are four areas: features, integrations, API and MCP, and account settings. - **[Features](/docs/features)** — what each screen is for, and the difference between AI visibility and your audit score. - **[Integrations](/docs/integrations)** — connect your CMS so approved drafts publish themselves, connect Search Console and Analytics, and let our scanner reach your site. - **[API and MCP](/docs/api)** — programmatic access. A REST API for reporting, and an MCP server so Claude, Cursor and other AI clients can work with your data directly. - **[Account and settings](/docs/account)** — plan and billing, team roles, notifications, credits, and the five per-site settings tabs. ## Where should I start after signing up? Do three things, in this order: confirm the scan reached your site, connect where you publish, then connect Search Console and Analytics. 1. **Check your scan reached the site.** If pages came back unfetchable, a firewall is blocking us — [allowlist the scanner](/docs/integrations/scanner). Everything downstream depends on this. 2. **Connect where you publish.** A recommendation you cannot act on is a to-do list. [Pick your channel](/docs/integrations) and approved drafts go straight into your CMS. 3. **Connect Search Console and Analytics.** They are what turn "you are not cited" into "here is the traffic it costs you." ## How do I read these docs as Markdown? Every page here is available as clean Markdown for AI tools: append `?md` to any URL, or send `Accept: text/markdown`. There is a machine-readable index of the whole site at [llms.txt](/llms.txt), of the documentation at [/docs/llms.txt](/docs/llms.txt), and of each area — for example [/docs/integrations/llms.txt](/docs/integrations/llms.txt). For the REST API there is also an [OpenAPI 3.1 spec at /openapi.json](/openapi.json), readable without a key. We do this for the same reason we recommend it to customers: content an AI assistant can read cleanly is content it can cite. --- # About CiteAgentic > CiteAgentic is an AI search visibility platform for AEO and GEO, co-founded by Kapil Arora and Bec Wade. We track, fix, and prove how ChatGPT, Perplexity, Gemini, Google AI Overviews and Claude cite your brand - and use AI agents to help your team execute. Source: https://www.citeagentic.com/about _About_ ## We're building the visibility layer for the AI search era. Buyers ask AI before they Google. CiteAgentic tracks how AI engines answer about your brand, helps your team prepare reviewed fixes, and compares later observations. ## What is CiteAgentic? CiteAgentic is an answer engine optimization (AEO) platform that helps teams measure and improve a brand's visibility across AI search engines: ChatGPT, Perplexity, Gemini, Google AI Overviews, Claude and Grok. It audits why competitors get cited instead of you, writes paste-ready AI fix prompts, drafts the missing content, and compares visibility on subsequent scans. Why we exist SEO was about being ranked. AI search is about being cited. 01The shift Ten blue links became one generated answer. Two or three brands get named; everyone else is invisible. Your buyers already ask ChatGPT, Perplexity and Google AI before they ever see a ranking. 02Our answer CiteAgentic connects monitoring with action: inspect competitor answers, prioritise supported findings, and use AI agents to draft work for your team to review and approve. 03How we work Measure the engines in your plan, inspect the evidence, review a useful fix, then re-scan the same questions. Compare observed changes without assuming your edit caused them. Our story We built a marketplace in 2020. Search is what beat it. Nextask was a marketplace for local services. We launched it in 2020, straight into the lockdowns, and for a long time we told ourselves the lockdowns were the problem. They weren't. What actually beat us was search. Airtasker and hipages already sat on top of every query a customer typed, and they had the budget to stay there: the expensive tools, the engineers, the fast infrastructure, the content team publishing every week. We were a small business chasing the same customers with none of that. Ranking wasn't a contest of who did the better work for people. It was a contest of who could spend more, and the small operators quietly disappeared. SEO was the big end of town's favourite sport, because they were the only ones who could afford to play it. "We lost a business to a search engine. This one exists so the next small team doesn't have to." Kapil Arora, Co-Founder AI search is a real reset. When someone asks an assistant which builder to use, or what a job should cost, the engine isn't counting backlinks or rewarding whoever shipped the fastest page. It's looking for the clearest, best evidenced answer, and it will cite a ten person business over a household name if that business is the one that actually explained the thing. That's the opening we built CiteAgentic for. Smaller businesses usually have the better raw material: the real project numbers, the customers who will vouch for them, the research nobody else has bothered to do, the community that already trusts them. What they don't have is a team to turn any of it into pages an engine can read, or the hours to do it again next week. So we didn't stop at tracking. Tracking tells you that you're losing, which we already knew in 2020. CiteAgentic finds the questions your buyers actually ask, works out why a competitor gets named instead of you, drafts the page in your own voice from facts you supply, and publishes it straight to your site. You approve the work. The platform does the lifting. We're not trying to help anyone outspend anyone. We're trying to make sure the size of a marketing budget stops being the reason a good answer never gets heard. Team Built by people who saw the shift early. Founder-led by design: customers talk directly to the people building the product, and every reply lands within a day. Kapil Arora Co-Founder 21+ years across digital transformation, content management and AI automation. The last several years: answer engine optimisation, entity authority and citation strategy, the discipline CiteAgentic productises. Connect on LinkedIn Bec Wade Co-Founder Co-founder focused on customers and delivery: onboarding, service quality and making sure every recommendation turns into shipped work for real businesses. Has delivered for Values What we believe Honest numbers only AI answers are probabilistic. We never promise "the next scan will show you cited". We show real counts, measure before and after, and report exactly what changed. Execution over advisory A dashboard that stops at "here's what's broken" leaves the expensive work with you. Every diagnosis here arrives with the draft, the prompt or the outreach ready to review. Move at AI speed Engines refresh weekly and competitors publish daily. We ship fast, re-test constantly, and adapt the playbook as the engines change, because last quarter's tactics already expired. Trusted by brands and agencies ### Frequently Asked Questions **Q: What is answer engine optimization (AEO)?** Answer engine optimization (AEO) is the practice of structuring your content, entity signals, and off-site authority so AI answer engines like ChatGPT, Perplexity, and Google AI Overviews can identify your brand as a dependable source and cite it in their answers. It is search-first in its foundations and answer-first in its formatting. **Q: What is generative engine optimization (GEO)?** Generative engine optimization (GEO) is closely related to AEO. It focuses on earning mentions and citations inside the responses generative AI models produce. CiteAgentic treats AEO and GEO as one loop: measure where AI names you, fix the gaps, build the off-site proof, and re-scan to confirm the change. **Q: Who is CiteAgentic for?** CiteAgentic is for founders, marketers, and agencies who need buyers to find them when the buyer asks an AI assistant instead of Googling. From solo builders tracking one brand to agencies managing many, anyone who wants to know and improve how AI engines answer about their brand can use it. **Q: Why was CiteAgentic started?** The founders previously ran Nextask, a local services marketplace launched in 2020. It was beaten less by the COVID lockdowns than by search: Airtasker and hipages had the budget for the tools, engineers and weekly content that kept them at the top of every query, and smaller operators could not compete on spend. AI search changes that, because answer engines reward the clearest, best evidenced answer rather than the biggest budget. CiteAgentic exists so a business of any size can turn its own numbers, customers and research into answers AI engines cite, without needing a content team to do it. **Q: Who is behind CiteAgentic?** CiteAgentic was co-founded by Kapil Arora and Bec Wade. Kapil brings 21+ years of digital transformation and AI automation experience with organisations including Breville, Qantas, Woolworths, Telstra, Vodafone and JB Hi-Fi; Bec leads customers and delivery. The company is founder-led by design: customers talk directly to the people building the product. Be the answer, not the alternative Start your free trial and see how the engines your buyers already use answer today. Results in minutes. Book a demo Start free trial --- # Best AI SEO Tools | Compare Workflows & AI Visibility > Compare AI SEO tools for content, technical SEO and AI visibility. A dated shortlist with use cases, plan checks and a practical trial method. Source: https://www.citeagentic.com/compare # Choose the best AI SEO tools for the work your team needs to do. Need better content, stronger Google search coverage, or evidence of where AI recommends competitors? Start with that job, then compare the tools, plan limits and work required to act. Published capabilities reviewed 28 September 2026. CiteAgentic’s editorial shortlist, not an independent ranking or hands-on benchmark. ## Five tools to shortlist by business need | Tool and source | Consider for | Published focus | Check before choosing | | --- | --- | --- | --- | | [CiteAgentic](/pricing) | AI visibility with website fixes | Mentions, citations, SEO/AEO findings and fix prompts | Engine access, quotas and implementation effort | | [Peec AI](https://peec.ai/pricing) | Team AI search analysis | Brand/source analysis and suggested actions | Models, projects and action availability | | [Otterly](https://otterly.ai/pricing) | AI monitoring and GEO audits | Prompt research, citations and daily tracking | Required engine add-ons | | [Semrush](https://www.semrush.com/) | Broad SEO and AI visibility research | Keywords, backlinks, site audits and AI visibility | Toolkit coverage and limits | | [Surfer](https://surferseo.com/) | Content optimisation | Content creation and optimisation for search and AI visibility | Editing work on a real article | ## AI-assisted SEO and AI visibility solve different jobs Improving an article does not tell you whether ChatGPT mentions your business. Tracking mentions does not replace keyword research, technical fixes or editorial judgement. Evaluate each workflow you need. ## Use the same test before you pay 1. Choose real buyer questions and one page needing improvement. 2. Match language, market, engines and cadence. Inspect answers and cited URLs. 3. Complete one recommended change and record effort. 4. Compare total prompt, model, project, seat and export costs on the intended billing term. 5. Review subsequent observations and qualified enquiries. Score changes do not prove causation. Use current vendor pricing linked above. No unverified cheapest-tool ranking. ## Detailed comparisons and next steps - [Peec AI alternative](/compare/citeagentic-vs-peec) - [Otterly alternatives](/compare/citeagentic-vs-otterly) - [CiteAgentic vs AIclicks](/compare/citeagentic-vs-aiclicks): separate July review date; not reverified in this batch. - [Free SEO and AEO tools](/tools) - [Managed implementation services](/services) - [CiteAgentic pricing](/pricing) --- # Contact CiteAgentic > Get in touch about audits, partnerships, custom tiers, agency services, or anything else. We respond within 24 hours. Source: https://www.citeagentic.com/contact _Contact_ ## Get in touch Real replies within 24 hours, often within the hour during weekday US and Sydney time. No ticket queue, no autoresponder. 1 We read it, personally. Your message lands with the founder, not a triage queue. 2 A real reply within 24 hours with concrete next steps, often within the hour on weekdays. 3 If a call helps, we book one, a 30-minute working session, not a sales script. Agencies & enterprise Managing multiple brands, or need SSO, custom regions or white-label reports? Mention it in your message and we'll route you straight to a working session. See plans & Enterprise Prefer email? karora@citeagentic.com, we read every one. --- # AI Visibility Pricing and Plan Limits > Compare current AI visibility plan prices, engines, sites, prompts and scan allowances. Review monthly or annual billing and paid-trial terms. Source: https://www.citeagentic.com/pricing _Pricing_ ## Choose the coverage your business needs. See the limits before you start. Choose the engine coverage, sites and usage that fit your team. The plan cards show your included engines and monitoring frequency. Paid trials require a card; you can also start with a limited free preview. Compare brands, markets, prompts, included engines, scan frequency and credits. Expand each card for its full feature list. Software access and managed implementation services are separate. Need implementation support? Explore [managed services](/services) or [contact us](/contact) with your website and priorities. ### Pricing questions, answered direct --- # CiteAgentic API and MCP — programmatic access > Pull CiteAgentic visibility data, recommendations and scan history into your own tools — over the REST API, or over MCP from Claude, Cursor and other AI clients. One key, one permission model. Source: https://www.citeagentic.com/docs/api _Documentation · API_ ## Programmatic access to your visibility data Two interfaces over one credential: a small REST API for reporting, and an MCP server that lets an AI client read your data and run scans directly. **TL;DR:** **TL;DR:** CiteAgentic offers two programmatic interfaces over one API key: a read-only [REST API](/docs/api/rest) with five v1 endpoints for reporting, and an [MCP server](/docs/api/mcp) with 27 tools that can also start scans and generate content. Use REST for dashboards and warehouses; use MCP for AI clients like Claude and Cursor. Both need a paid plan and are governed by the same 13 [scopes](/docs/api/authentication). ## Should I use the REST API or MCP? Use the REST API for reporting into your own dashboards, and MCP for anything an AI client does or anything that starts work. The REST API is read-only; MCP is not. | You want to | Use | | --- | --- | | Pull scores and recommendations into a dashboard, warehouse or report | [REST API](/docs/api/rest) | | Let Claude, Cursor or another AI client work with your data | [MCP server](/docs/api/mcp) | | Start scans, generate prompts, or draft content programmatically | [MCP server](/docs/api/mcp) — the REST API is read-only | Both use the same key and the same scopes, so choosing one does not lock you out of the other. ## What is on each API page? Three pages: authentication, the REST API, and the MCP server. - [Authentication](/docs/api/authentication) — getting a key, the scope vocabulary, plan requirements, and what each rejection means. - [REST API](/docs/api/rest) — the five v1 endpoints, with examples and an honest list of what v1 does not do yet. - [MCP server](/docs/api/mcp) — endpoint, the 27 tools, cooldowns and idempotency. ## Is there a machine-readable spec? Yes, for the REST API: [OpenAPI 3.1 at /openapi.json](/openapi.json). The MCP server describes itself — call `tools/list` and it returns every tool with its JSON Schema, so no separate spec is needed. ## What rules apply to both interfaces? **A key sees exactly what its owner sees.** Scope is the sites that user owns or collaborates on, enforced by row-level security in the database rather than only in the route handler. **Quotas are the plan's quotas.** Anything created through the API counts against the same limits as the UI. This is another door into the same account, not a way around the plan. **Running work spends credits.** Read scopes and run scopes are separate for exactly this reason. Grant `run:` only to a key that is meant to start scans. --- # AEO Tools for Monitoring, Audits and Action > Explore CiteAgentic AEO tools for AI mentions, citations, page audits and buyer questions. Find the right free diagnostic or account workflow. Source: https://www.citeagentic.com/features _Platform capabilities_ ## Choose a tool by the problem you need to solve | Your decision | CiteAgentic capability | Useful starting point | | --- | --- | --- | | Are buyers seeing our brand in AI answers? | Sampled brand mentions and competitor responses | [AI visibility checker](/features/aeo-scan) | | Is our visibility changing? | Repeat observations for selected buyer questions | [AEO tracker](/features/aeo-tracker) | | Which websites are being cited? | Returned citation URLs and source comparisons | [AI citation tracking](/features/ai-citation-tracking) | | What should we fix on a page? | Page-level content and readiness findings | [Content audit tool](/features/aeo-page-audit) | | Are technical problems holding the site back? | SEO findings, affected pages and fix workflow | [SEO audit software](/features/seo-scan) | | Which questions should we investigate? | Generated ideas grouped by intent | [Keyword question tool](/tools/what-customers-ask-chatgpt) | ## AEO and GEO tools in one working process Answer engine optimisation and generative engine optimisation overlap: both concern how a business and its information appear in AI-generated answers. CiteAgentic supports this work through monitoring, diagnostics and evidence-based recommendations. It is one product with connected capabilities, not an independent ranking of competing tools. Explore AI citation tracking: inspect cited pages and turn source gaps into next steps. **Discover the gap** Start with a business, audience, market and buyer question. Review completed answers and citations to see whether the gap is a missing mention, a competing source or limited coverage. **Decide what to change** Use audit evidence and business context to prioritise relevant fixes. A copied fix prompt can brief your existing AI or development workflow; review the proposed changes before publishing. **Measure the next result** Recheck the finding after implementation and compare later answers using the same questions. Track enquiries and sign-ups separately from AI presence to assess commercial value. ## Try a focused check before choosing a plan Use the [free tools hub](/tools) for URL readiness, robots policy, brand recognition and generated question ideas. Free tools have a limited scope and may require an email to run or unlock details. Their output does not replace ongoing monitoring or a full technical investigation. **Illustrative workflow:** a service business sees a competitor recommended for a high-value question. It inspects the answer, reviews the relevant service page, corrects a missing explanation and compares later runs. The process helps test an improvement; it does not prove why an engine made its original choice. ## Check coverage before you commit Engine access, sites, prompt allowances, scan cadence, credits and history depend on the selected plan. Returned answers vary and some engines do not supply citation links. AI Overviews is a separate search signal, not an additional conversational engine. Review [pricing](/pricing) for current allowances. If your team needs implementation support, explore [managed AEO and GEO services](/digital-marketing/ai-visibility-aeo). If you are choosing a workflow for several clients or locations, start with [use cases](/use-cases). --- # SEO & AEO Services | Choose Your Implementation Support > Choose regional SEO, managed AEO and GEO services, or self-serve AI visibility software. Compare the scope and find the right next step for your team. Source: https://www.citeagentic.com/services _Choose who does the work_ ## You know visibility needs work. Get the right team behind it. Bring us in to improve your search pages and AI visibility, or use CiteAgentic to guide your own team. Choose the route that fits your problem, the work you need done and who will implement it. ### Three routes, each with a clear next step **We need better local search enquiries** For Brisbane and Gold Coast businesses that need technical SEO, stronger service pages and local search work alongside AI search optimisation. Review the scope and package options in our SEO Brisbane and Gold Coast services. **AI names competitors instead of us** For teams that need someone to investigate mentions and cited sources, prioritise content and evidence gaps, and implement agreed changes. Explore our answer engine optimization services and managed GEO work. **Our team can make the changes** Use the AI visibility tool to investigate buyer questions and competitor mentions. Follow repeated observations with the AEO tracker and compare software plans for the access you need. ### Agree the work before committing to an engagement The first decision is what needs doing and who will own it. We scope help around your existing team and providers. **Advisory and prioritisation** Review available scan evidence, buyer questions and priority pages together. Leave with the actions to investigate, the evidence still needed and implementation responsibilities your team can understand. **Technical and page improvements** Agree the fixes we will implement, from accessibility and internal links to service-page clarity and supported structured data. Work that needs your developer or business approval is identified in the scope. **Content based on actual buyer questions** Develop service content, FAQs, comparisons and evidence briefs using your business knowledge and observed gaps. Use customer evidence and service expertise to answer the questions that matter before an enquiry. **Progress you can review** See what shipped, what remains blocked and how comparable observations changed. Track qualified enquiries, search performance and AI citations separately, so your team can see which outcomes changed and what to prioritise next. Need campaign support beyond organic search? Explore the digital marketing service menu. Want an initial website check before discussing managed work? Try the free SEO audit or AEO readiness audit. Readiness findings do not measure whether AI currently recommends your brand. ### Services questions, answered ### Frequently Asked Questions **Q: Should I choose software or managed services?** Choose software when your team can investigate findings and implement changes. Discuss managed services when you need help deciding what to do or someone to deliver agreed work. The service pages explain the scope; the pricing page covers software access. **Q: Can you work alongside our existing SEO agency?** Yes. We can scope specialist AI visibility work or review a defined set of gaps. Agree responsibilities and access first so the engagement complements existing work instead of duplicating it. **Q: How is managed service pricing decided?** We discuss your goals, priority pages, services, markets and implementation needs, then agree the scope and fees. Confirm any software access requirements as part of that scope; software pricing is not a managed-service quote. **Q: What should I bring to the first conversation?** Your website, priority services, target locations, main competitors and the result you want. Existing Search Console data, scan reports and genuine customer evidence are useful if available, but we can discuss where to start without them. **Get a scope that matches the work you actually need.** Tell us whether the priority is local search, AI recommendations or implementation capacity. We will help you choose a service path and define a useful starting point. --- # Resources > Long-form guides on AEO, AI search, fix prompts, and the citation graph. Where the AI-search era's playbook gets written down. Source: https://www.citeagentic.com/resources _Resources_ ## The AI-search playbook, written down as we learn it. No 17-step listicles. No recycled SEO advice. Long-form guides grounded in what we observe across thousands of brand scans: how AI engines decide whom to cite, and what actually changes the outcome. ### Start here What answer engine optimization (AEO) is, and how it differs from classic SEO. **What is AEO visibility?** The practical definition: the share of AI answers that mention you. Why that number beats the alternatives, and the four signals that move it. Read → **AEO vs SEO - the actual difference** Why AEO isn't 'SEO with extra steps'. The signals are different, the playbook is different, the metrics are different. Read → **Will AEO replace SEO?** The opinion piece. AEO won't replace SEO, but it reshapes which parts of SEO matter, on a faster timescale than most teams are ready for. Read → **5 things every founder should know about AEO** Five claims, each with a paragraph of expansion, a practical example, and an action you can take this week. Read → ### From unknown to cited The complete path: crawler access, source selection, citable content, entity clarity, independent proof, and verifying the fix worked. **How to get cited by AI search engines** The complete guide: crawler access, source selection, extractable content, entity clarity, independent proof, and measuring whether a fix changed the answer. Read → **How to get mentioned by AI** The prioritised implementation checklist, in pipeline order: access, content, entity, independent proof, measurement. Each step with a way to verify it. Read → **Which AI crawlers should you allow?** OAI-SearchBot vs GPTBot, Googlebot vs Google-Extended, PerplexityBot: what each crawler controls, safe robots.txt configs, and how to verify access end to end. Read → **How do AI engines choose sources?** What public documentation supports about source selection: query fan-out, relevance to sub-questions, and why competitors get cited instead of you. Read → **How to structure a citable page** The page formula: direct answer, evidence, example, qualification, next step. Plus honest structured data and how to verify the page after publishing. Read → **How do AI engines identify your business?** Consistent naming, About pages, Organization schema, sameAs, Business Profile and directories: making every record tell the same story. Read → **Why reviews and communities decide recommendations** Engines trust sources you do not control. Which ones matter by business type, how to find the sources supporting competitors, and how to participate legitimately. Read → **How to measure whether a fix worked** Baselines, mentioned vs recommended vs cited, prompt-level tracking, and a before-and-after test design that survives AI answer variance. Read → **Traffic but no AI citations?** Solid Google traffic, invisible in AI answers. The six verifiable failure classes behind the gap, and the order to test them in. Read → **Citation graphs vs backlink graphs** Backlinks and AI citations answer different questions. What each graph measures, when you need each, and what only the citation graph can tell you. Read → **Surfer SEO alternatives for AI search** Tools built for Google rankings score the wrong target for AI answers. The honest tool landscape, split by what you're actually optimising for. Read → ### How we build Why our fixes come as copy-paste prompts, and how the scores are computed. **How scoring works** How the visibility, sentiment, share-of-voice and citation scores are computed, and the guardrails that keep small samples from lying. Nothing hidden. Read → ### Head to head Honest side-by-sides against the other AI-visibility tools. **CiteAgentic vs Otterly** Both track your visibility in AI answers. Only one writes the fixes and proves they worked. Read → **CiteAgentic vs AIclicks** They cover the most AI platforms. We tie every fix to a question you are losing. Read → **CiteAgentic vs Peec AI** Their strength is analytics. Ours is doing the fixes and proving they worked. Read → --- # AI Visibility Use Cases for Businesses and Agencies > Find an AI visibility workflow for founders, local businesses, marketing teams and agencies. Connect the right questions, evidence and next actions. Source: https://www.citeagentic.com/use-cases _Use cases_ ## Founders: find out whether you make the shortlist Your buyer is comparing solutions, not searching for a new brand by name. Start with a small set of category, use-case and alternatives questions that match your product today. Inspect which companies appear and what the responses say they are good at. **Example workflow:** a new project-management product tests questions about agency approvals. If a competitor appears, review the answer and source before deciding whether the product page needs a clearer workflow explanation or genuine customer proof. **Measure:** valid answers naming your brand, relevant citations and qualified demo enquiries. Start with the [AI visibility checker](/features/aeo-scan), then use the [tracker](/features/aeo-tracker) for repeat observations. ## Local businesses: connect the service and the place Use questions that include the service, customer need and real service area. Keep Brisbane and Gold Coast observations separate when location is part of the question. Do not imply an office or service capability the business does not have. **Example workflow:** an accounting firm reviews recommendations for a specific advisory service in its market. It checks its service page for clear scope, credentials and accurate location information, then rechecks the question set after publishing improvements. **Measure:** mentions by market, useful referral pages and qualified calls or enquiries. Start with a [free AEO audit](/tools/free-aeo-audit) or [regional SEO support](/services/digital-marketing/aiseo). ## Marketing teams: turn a visibility gap into an owned task When someone asks why a competitor appears, begin with the actual answer. Distinguish a missing mention from a missing citation, then connect the observation to a relevant page and accountable owner. Do not rewrite an entire site because one response changed. **Example workflow:** a B2B team notices repeated comparison questions naming a competitor. It checks factual accuracy, reviews the cited sources and decides whether a comparison page or implementation guide would help buyers evaluate the product. **Measure:** repeated results on the same prompts, issue resolution and conversions from the affected content. Use [citation tracking](/features/ai-citation-tracking) and the [content audit tool](/features/aeo-page-audit). ## Agencies: keep each client's evidence and decisions separate Choose questions and markets for each client rather than recycling one generic prompt list. Use site-scoped results to explain what changed, why it matters and what work should happen next. Multi-brand access and reporting allowances depend on the plan. **Example workflow:** an agency reviews two clients with different services. Each gets its own question set, audit findings and next action. Reporting distinguishes completed fixes, monitoring observations and work still waiting for approval. **Measure:** findings resolved, client-specific visibility trends and qualified outcomes. Compare [plans](/pricing) and explore the [AEO tools](/features) before agreeing client deliverables. ## Choose self-service or implementation support Use the product when your team can review and apply the recommendations. Choose [managed services](/services) when you need help researching, prioritising and implementing the work. A strategy enquiry should include your website, target buyers, service areas and the decision you need to make. **Start with one question your next buyer would ask.** Review the generated ideas against real customer language, select the relevant ones and set up monitoring coverage that fits your plan. --- # Terms of Service > The agreement governing your use of CiteAgentic, under the laws of Australia. Source: https://www.citeagentic.com/legal/terms _Legal_ ## Terms of Service Last updated: 26 May 2026 These Terms of Service ("Terms") are an agreement between you (and the organisation you represent, "you") and **[CiteAgentic Pty Ltd, ABN/ACN 54238243639]** ("CiteAgentic", "we", "us"). By signing up for or using CiteAgentic, you agree to these Terms. If you are agreeing on behalf of an organisation, you confirm you have authority to bind it. ## 1. The service CiteAgentic is an AEO and audit platform that queries AI engines on your behalf, audits your domain across SEO, AEO, performance, security and conversion, and produces scores, recommendations and AI fix prompts. We may update, improve or change features from time to time. ## 2. Eligibility and accounts You must be at least 16 and able to form a binding contract. You are responsible for all activity under your account, for keeping your credentials secure, and for the content of the brands, prompts and competitors you configure. ## 3. Acceptable use You agree not to: - Track or audit a brand or domain you have no legitimate business interest in, or without the brand owner's authorisation (your own brand, your employer's, or a client's with written authorisation) - Scrape or access CiteAgentic other than through the documented REST API - Reverse-engineer, resell or redistribute our recommendations or fix prompts as a competing product - Use the service unlawfully, to infringe others' rights, or to compromise its security or integrity We may suspend or terminate access for breach of this section. ## 4. Your data and content As between you and us, you own the data and content you submit (brand information, prompts, competitors, connected-account data). You grant us a licence to host, process and use it to provide and improve the service, and as described in our [Privacy Policy](/legal/privacy). ## 5. Our intellectual property We own all rights in the CiteAgentic platform. The outputs we generate for you (scores, recommendations, AI fix prompts) are licensed to you for unlimited internal business use under your subscription. Aggregated, de-identified benchmark data is owned by us. ## 6. Industry benchmarks We compute aggregated cohort benchmarks (averages and percentiles per category) across customers. These are de-identified and minimum-N gated so they cannot be attributed to an individual brand. ## 7. Third-party services and AI engines The service relies on third-party providers and AI engines (see our Privacy Policy). Their availability and outputs are outside our control, and your use of any connected service (such as Google Search Console) is also subject to that provider's terms. We may engage third‑party service providers (including hosting, analytics and AI providers) as sub‑processors to deliver the Service. Some may be located outside Australia. We take reasonable steps to ensure they protect Your Data in line with our Privacy Policy. When you connect a third‑party account (for example, Google Search Console), you authorise us to access and process data from that account on your behalf. You are responsible for maintaining your rights to use and share that data. ## 8. Fees, billing and taxes Paid plans are billed in advance through our payment provider (Stripe). Fees are stated on our [Pricing page](/pricing) and are exclusive of GST and other applicable taxes unless stated otherwise. You can cancel any time from Settings; cancellation takes effect at the end of the current billing period. Except where required by law (including the Australian Consumer Law), fees already paid are non-refundable, though we will act reasonably where a fault is ours. Subscriptions automatically renew at the end of each billing period unless you cancel through the Service. Some plans include a bundle of usage credits (for example, AI tokens or analysis runs) per billing period. The type and quantity of included credits, and how they are consumed. Unused credits expire at the end of the applicable billing period and do not roll over, unless we state otherwise at the time of purchase. If you use all included credits, you may continue using the Service by purchasing additional credits or a higher plan at the then‑current prices shown in the Service. Additional credits are billed in advance and are non‑refundable, subject to your non‑excludable rights under the Australian Consumer Law. We may suspend or limit credit‑based features if you have no remaining credits or if payment for credits fails. ## 9. Free tier and trials We may offer a free tier or trial. We can change or withdraw free offerings at any time, and usage limits described on our Pricing page apply. ## 10. Warranties and the Australian Consumer Law We provide the service using reasonable care and skill. **Nothing in these Terms excludes, restricts or modifies any consumer guarantee, right or remedy you have under the Australian Consumer Law (ACL) that cannot lawfully be excluded.** Subject to that, and to the maximum extent permitted by law, the service is provided "as is" and we do not warrant that it will be uninterrupted, error-free, or that it will produce any particular business outcome. ## 11. Limitation of liability To the maximum extent permitted by law, and subject to your non-excludable rights under the ACL: - Neither party is liable for indirect, incidental or consequential loss, or loss of profits, revenue, data or goodwill. - Our total aggregate liability arising out of or in connection with the service is limited to the amount you paid us for the service in the 12 months before the event giving rise to the liability. - Where our liability for a breach of a non-excludable ACL guarantee cannot be excluded but can be limited, our liability is limited (at our option) to re-supplying the service or paying the cost of having it re-supplied. ## 12. Indemnity You indemnify us against claims, losses and costs arising from your data or content, your use of the service in breach of these Terms, or your infringement of any third-party or brand-owner rights. ## 13. Suspension and termination You may stop using the service and close your account at any time. We may suspend or terminate access if you materially breach these Terms, fail to pay, or use the service in a way that risks harm to us or others. On termination, your right to use the service ends and we will handle your data as described in the Privacy Policy. Before or within a reasonable time after termination, you may export available reports and outputs from the Service, except where we have suspended your access due to serious breach or legal obligation. ## 14. Changes to these Terms We may update these Terms. For material changes we will give you at least 14 days' notice by email or in-product. Your continued use after the changes take effect constitutes acceptance. ## 15. Governing law These Terms are governed by the laws of ** Queensland, Australia**, and you and we submit to the non-exclusive jurisdiction of the courts of that place. ## 16. General These Terms are the entire agreement between you and us about the service. If any provision is unenforceable, the rest continues to apply. Our failure to enforce a right is not a waiver. You may not assign these Terms without our consent; we may assign them to a successor in connection with a sale or reorganisation of our business. ## 17. Contact Questions about these Terms: **[karora@citeagentic.com]** - **[CiteAgentic Pty Ltd]**, [Registered address, State, Australia]. --- # Case Studies > See how brands are using CiteAgentic to win in AI search. Real results from SaaS, agencies, and e-commerce customers. Source: https://www.citeagentic.com/case-studies _Case Studies_ ## How brands are winning in AI search Real results from customers shipping AEO fixes. See how SaaS, agencies, and e-commerce are adapting to the era when AI engines decide whom to cite. ## Featured stories Fintech / SaaS 2.8× increase in AI search visibility How a fintech platform recovered lost AI citations by fixing schema and topical gaps in 45 days. Read story → Agency Shipping AEO fixes for 12+ clients An SEO agency added AEO monitoring and paste-ready fix prompts to their client delivery stack. Read story → E-commerce Citation rate climbed 67% in 60 days An e-commerce brand went from invisible in AI answers to showing up in 4 of 5 competitor product queries. Read story → Real Estate / Local 4× organic traffic growth in 90 days A Sydney property agency captured AI-driven buyer intent by fixing AEO readiness gaps, achieving 325% revenue growth. Read story → ## Why brands switch to CiteAgentic Real audits, not estimates We run actual scans of your pages through every major AI engine. No simulations. No guesswork. Paste-ready fix prompts Don't just get told what's broken. Get the exact prompt you hand to Claude, ChatGPT, or Cursor to fix it. Monitor continuously Schedule recurring scans. Watch citation trends week over week. Catch regressions before they tank visibility. Free tier = real value Get industry benchmarks, page audits, and fix prompts on the free plan. Pro tier unlocks monitoring and recommendations. ## Getting started Run your first AEO audit today and see how your site stacks up: Get started free → --- # CiteAgentic account and site settings > Manage your CiteAgentic account — profile, plan and billing, team members and roles, notification preferences, sites, beta features and account deletion — plus the five per-site settings tabs. Source: https://www.citeagentic.com/docs/account _Documentation · Account_ ## Account, team and site settings Two levels of configuration: your account, which covers billing, team and preferences, and each site, which covers what we scan and where we publish. **TL;DR:** **TL;DR:** Account settings live at `/account` in seven sections (profile, plan and billing, team, notifications, sites, beta features, danger zone); site settings live at `/sites/:id/settings` in five tabs (business profile, regions, scanning, integrations, advanced). Team roles are owner, admin, analyst and viewer. Account deletion is a 30-day soft delete you can cancel at any point in that window. ## What can I change in account settings? At **/account**, in seven sections. | Section | Covers | | --- | --- | | **Profile** | Name, email, and password. Minimum 12 characters; if you signed up with Google you can set a password here without knowing an old one. | | **Plan and billing** | Your plan, credit balance, top-ups and invoices. Billing is self-serve through Stripe's customer portal. | | **Team** | Invite people and set a role. | | **Notifications** | Three toggles: weekly digest, alert emails, product emails. | | **Sites** | Every site on the account, and where to add another. | | **Beta features** | Opt into features still being finished. Today that is Reporting, which reveals Report Studio in the sidebar. | | **Danger zone** | Delete the account. | ### What can each team role do? There are four roles. Only the owner can add or remove people; admin, analyst and viewer differ in how much they can change. | Role | Can | | --- | --- | | **Owner** | Everything, including adding and removing people. Only the owner can change the team. | | **Admin** | Manage sites, scans and content. | | **Analyst** | Run scans and work with the output. | | **Viewer** | Read only. | ### How do I delete my account? Deletion is a 30-day soft delete, and it is cancellable for the whole of that window. Nothing is destroyed on the day you click it. You can also export everything first — **Account → Profile → Export** returns your data as JSON. ## What can I change per site? At **/sites/:id/settings**, in five tabs. These are per site, so a change to one site never affects another. | Tab | Covers | | --- | --- | | **Business profile** | What the business is, what it sells, and its Google Business Profile URL. This is what grounds prompt generation and content drafting, so it is worth getting right. | | **Regions** | Which markets to scan. The first region is the primary one; free scans cover the primary region only. | | **Scanning** | Scan frequency and scope. | | **Integrations** | [Search and analytics connections](/docs/integrations), [publishing channels](/docs/integrations), and an outbound CMS webhook. | | **Advanced** | Danger zone for the site. | ## What spends credits? AI work spends credits: scans across engines, content drafts, agent runs. Your plan includes an allowance and you can top up. Reads never spend credits — only work that calls a model does. This is also why API keys separate `read:` scopes from `run:` scopes, so an automation cannot spend credits it was never meant to touch. --- # CiteAgentic MCP server — connect Claude, Cursor and other AI clients > Connect an AI client to CiteAgentic over the Model Context Protocol. Endpoint, authentication, the 46 tools (38 named plus an 8-tool gateway that reaches every other app function) it exposes, and the scopes that govern them. Source: https://www.citeagentic.com/docs/api/mcp _Documentation · MCP_ ## Give your AI client your visibility data CiteAgentic runs a Model Context Protocol server, so Claude, Cursor and any other MCP client can read your scans, prompts, competitors and recommendations — and run new scans — without you copying anything across. **TL;DR:** **TL;DR:** CiteAgentic runs an MCP server at `POST https://citeagentic.com/mcp`, authenticated with `Authorization: Bearer ` and speaking JSON-RPC over Streamable HTTP (protocol version 2025-06-18). It exposes 46 tools (38 named plus an 8-tool gateway that reaches every other app function) covering sites, scans, prompts, competitors, recommendations and content, governed by 25 scopes. Visibility scans respect a one-hour per-site cooldown, so a retry does not bill you twice. ## What is the CiteAgentic MCP endpoint? It is `POST https://citeagentic.com/mcp`. The server speaks JSON-RPC over Streamable HTTP at that single endpoint: ``` POST https://citeagentic.com/mcp Authorization: Bearer ``` `GET /mcp` answers `405` on purpose — this is a POST-only transport. One key works for both the MCP server and the [REST API](/docs/api/rest) — there is one credential, not two. See [Authentication](/docs/api/authentication) for how to get one and how scopes work. ## What tools does the MCP server expose? Twenty-seven tools, grouped by what they touch. | Area | Tools | | --- | --- | | Sites | `cite_list_sites`, `cite_get_portfolio_summary`, `cite_get_site_brief` | | Scans | `cite_get_latest_scan`, `cite_get_scan_status`, `cite_run_visibility_scan`, `cite_run_audit_scan` | | Audit | `cite_get_audit_scores`, `cite_get_audit_findings` | | Visibility | `cite_get_visibility_scores` | | Prompts | `cite_list_prompts`, `cite_get_prompt_detail`, `cite_generate_prompts`, `cite_refine_prompt`, `cite_get_prompt_competitor_results` | | Competitors | `cite_list_competitors`, `cite_add_competitor`, `cite_get_competitor_gaps`, `cite_run_competitor_comparison` | | Recommendations | `cite_get_recommendations` | | Content | `cite_get_content_opportunities`, `cite_generate_content_draft`, `cite_get_voice_profile` | | Buyer evidence | `cite_get_review_themes`, `cite_get_buyer_criteria`, `cite_get_positioning_claims` | | Agent | `cite_ask` | ## What scopes does each tool need? Each tool declares the scope it needs, and a key only carries the scopes you grant it. Read and write are separated, and so is *running* work that costs credits. | Scope | Grants | | --- | --- | | `read:sites` | Site list, portfolio summary, site brief | | `read:visibility` | Visibility scores and history | | `read:audits` | Audit scores and findings | | `read:prompts` | Prompt list and per-prompt detail | | `read:competitors` | Competitor list and gap analysis | | `read:recommendations` | The recommendation queue, including fix prompts | | `read:content` | Content opportunities and drafts | | `read:voice` | Voice profile, review themes, buyer criteria | | `write:prompts` | Generate and refine prompts | | `write:competitors` | Add a competitor | | `write:content` | Generate a content draft | | `run:visibility` | Start a visibility scan | | `run:audits` | Start an audit scan | Grant the narrowest set that does the job. A key that only needs to report on visibility does not need `run:` scopes, and a scan started by mistake spends credits. ## What should I know before wiring up MCP? **Scans are rate-limited per site.** A visibility scan started through MCP respects a one-hour cooldown, the same as one started in the app. A second request inside that window returns the existing run rather than starting a duplicate and billing you twice. **Write tools accept an idempotency key.** Pass `idempotencyKey` on a tool that creates something and a retry returns the original result instead of creating a second copy. Worth using: agent loops retry more than people do. **Every call is recorded** against the key that made it, so you can see which client did what. **Quotas are the plan's quotas.** Prompts, competitors and drafts count against the same limits as the UI. MCP is another way into the same account, not a way around it. ### Frequently Asked Questions **Q: Which AI clients work with this?** Any MCP client that supports Streamable HTTP with a bearer token — Claude Code, the Claude desktop app, Cursor, and others. The server advertises protocol version 2025-06-18. **Q: Is the MCP key the same as the API key?** Yes. One key in Account → API keys works for both POST /mcp and the REST API. Scopes govern what it can reach. **Q: What happens if my plan does not include API access?** Every call returns tier_disallows_api. Nothing partially works. --- # CiteAgentic REST API v1 — endpoints and responses > The CiteAgentic REST API v1 — five read endpoints covering sites, visibility, recommendations and scan runs, authenticated with the same bearer key as the MCP server. Source: https://www.citeagentic.com/docs/api/rest _Documentation · API_ ## REST API v1 A small read-only surface for pulling visibility data and recommendations into your own dashboards, warehouses and reports. **TL;DR:** **TL;DR:** CiteAgentic REST API v1 has five endpoints, all `GET` and all read-only: `/api/v1/brands`, `/api/v1/brands/:id`, and that brand's `/visibility`, `/recommendations` and `/runs`. Authenticate with `Authorization: Bearer `. There are no write endpoints, no pagination and no webhooks in v1 — to start scans or generate content, use the [MCP server](/docs/api/mcp). ## What can the REST API do? Five endpoints, all `GET`, all read-only. If you need to *start* work — run a scan, generate prompts, draft content — that lives on the [MCP server](/docs/api/mcp), not here. v1 is deliberately a reporting surface. All requests need `Authorization: Bearer `. See [Authentication](/docs/api/authentication). ## Is there an OpenAPI spec? Yes. The full OpenAPI 3.1 description is at [/openapi.json](/openapi.json) (also served at `/api/v1/openapi.json`). It declares every endpoint, query parameter, response shape and error, so an AI agent or a client generator can work from a contract rather than from this prose. It needs no key to read. A test pins the spec to the routes that actually exist in both directions, so an endpoint cannot be added, removed or renamed without the spec failing the build. ## What are the REST API endpoints? | Method | Path | Returns | | --- | --- | --- | | `GET` | `/api/v1/brands` | Every brand the key's owner can see | | `GET` | `/api/v1/brands/:id` | One brand's detail | | `GET` | `/api/v1/brands/:id/visibility` | Visibility scores over time | | `GET` | `/api/v1/brands/:id/recommendations` | The recommendation queue, each with its fix prompt | | `GET` | `/api/v1/brands/:id/runs` | Scan run history | ### Example ```bash curl https://citeagentic.com/api/v1/brands \ -H "Authorization: Bearer $CITEAGENTIC_KEY" ``` Then pull the recommendations for one of them: ```bash curl https://citeagentic.com/api/v1/brands//recommendations \ -H "Authorization: Bearer $CITEAGENTIC_KEY" ``` Each recommendation carries `payload.fixPrompt` — the paste-ready prompt we generate for it. That is included on purpose: if you are building your own workflow on top of CiteAgentic, the fix prompt is the part worth having. ## What do the error responses mean? | Response | Meaning | | --- | --- | | `401 missing_bearer_token` | No `Authorization: Bearer` header | | `401 invalid_api_key` | Unknown or revoked key | | `403 tier_disallows_api` | Valid key, but the plan does not include API access | | `404` | The brand does not exist, or the key's owner cannot see it | A brand belonging to another account returns `404`, not `403` — an API should not confirm that something exists to someone who cannot read it. ## What are the rate limits? Requests are rate-limited per IP, shared with the rest of the application. There is no separate per-key quota today, so a busy integration and a busy browser session draw on the same allowance. ## What is missing from v1? Four things: write endpoints, pagination, webhooks, and per-key rate limits. Being straight about the edges, because guessing wastes your time: - **No write endpoints.** Creating prompts, competitors or drafts is MCP-only for now. - **No pagination or cursors.** Collections return in full. - **No webhooks or event subscriptions.** Poll, or use MCP. - **No per-key rate limits or usage endpoint.** Calls are logged against the key, but there is no endpoint to read that log. If one of these is blocking something you are building, tell us — the order we build them in follows what customers actually ask for. --- # CiteAgentic features — what each screen does > A map of the CiteAgentic product: Overview, Discover, Improve, Authority, Content Studio, Plan and SEO — what each surface answers and where its data comes from. Source: https://www.citeagentic.com/docs/features _Documentation · Features_ ## What each screen is for CiteAgentic is organised around one loop: find the buyer questions you lose in AI answers, understand why you lose them, fix it, and prove the fix moved something. Each surface owns one stage of that loop. **TL;DR:** **TL;DR:** CiteAgentic is organised as one loop across four areas — Overview (where you stand), Discover (which buyer questions matter), Improve (what to change), Authority (which independent sources AI trusts) — plus Content Studio, Plan and SEO. AI visibility and audit score are deliberately two different numbers: the audit measures whether your page *can* be cited, visibility measures whether it *is*. ## What are the four areas of CiteAgentic? The four areas are Overview, Discover, Improve and Authority. They are stages of one job, not separate products, and the left rail follows that order deliberately. | Area | Answers | Where it lives | | --- | --- | --- | | **Overview** | Where do I stand, and what is the single most valuable thing to do next? | Dashboard | | **Discover** | Which buyer questions matter, who wins them today, and what do buyers actually care about? | Prompts, Topics, Buyers, Competitors | | **Improve** | What do I change, in what order, and did it work? | Recommendations, Site Audit, Pages | | **Authority** | Which independent sources do AI engines trust, and am I present in them? | Sources, Authority Map | ## What else is in the product? Four surfaces sit outside the visibility loop: Content Studio, Plan, SEO and Reporting. - **Content Studio** — turns a visibility gap into a draft. Ideas, a content audit, and a calendar that publishes on a schedule through a [connected channel](/docs/integrations). - **Plan** — the work queue. Tasks grouped by how often they recur, not by an invented phase. - **SEO** — classic search performance: dashboard, keywords, local (when the site is local), competitor compare, and reports. Separate from AI visibility on purpose, because the two are measured differently. - **Reporting** — Report Studio, currently behind a beta flag you can enable in **Account → Beta features**. ## What is the difference between AI visibility and audit score? AI visibility measures whether AI answers actually mention you; audit score measures whether your pages are *able* to be cited. One is an outcome, the other is readiness, and a perfect audit score does not guarantee a citation. **AI visibility** is the share of AI answers that mention, recommend or cite you. It is measured by asking real engines real buyer questions and reading what comes back. **Audit score** is how ready your pages are to be cited — structure, evidence, accessibility to AI crawlers. It is measured by reading your site. A perfect audit score does not guarantee a citation. On-page work improves your *eligibility*; independent third-party corroboration improves the *likelihood*. The product keeps them apart because collapsing them into one number would let a page look finished when nothing about the outcome had changed. ## Where are the detailed guides for each screen? Detailed guides for individual surfaces are being written. In the meantime, the [resource library](/resources) covers the underlying concepts — how AI engines choose sources, what makes a page citable, and how scoring works. --- # Cookie Policy > CiteAgentic cookie policy. Source: https://www.citeagentic.com/legal/cookies _Legal_ ## Cookie Policy Last updated: 2026 This policy explains what cookies and similar technologies CiteAgentic uses, why we use them, and how you can control them. It should be read alongside our [Privacy Policy](/legal/privacy), which covers how we handle personal information more broadly. ## What cookies are A cookie is a small text file that a website stores on your browser when you visit it. Cookies let a site remember things between page loads and visits - that you're signed in, what theme you prefer, or how many people land on a given page. Some cookies are set by the site you're visiting ("first-party"); others are set by services that the site embeds, such as analytics providers ("third-party"). We also use `localStorage`, a similar browser-storage mechanism, for a couple of small client-side preferences - we treat it the same way as a cookie for the purposes of this policy. ## Cookies we set | Name | Purpose | Type | Duration | |---|---|---|---| | `trace_session` | Authenticated session (HttpOnly - not readable by page scripts) | Essential | 30 days | | `trace_csrf` | CSRF token, JS-readable, echoed back as a header on state-changing requests | Essential | 30 days | | `trace_theme` | Remembers your light/dark theme choice | Preference | Persistent | | Google Analytics (`_ga`, `_ga_*`) | Aggregated website usage analytics (production only) | Analytics | Up to 2 years | | `ca:analytics:consent:v1` (local storage) | Remembers your analytics choice | Preference | 180 days | ### Essential cookies `trace_session` and `trace_csrf` are what make the product work. The session cookie is how the server recognises you as signed in across page loads; it's marked `HttpOnly`, so it can't be read or tampered with by page scripts, which protects it from cross-site-scripting attacks. The CSRF cookie is a security pairing token: when you take an action that changes data (saving a brand, running a scan, updating billing), the app reads this cookie and echoes its value back to the server as a header, proving the request came from our own page rather than a malicious third-party site. Neither cookie is used for advertising, profiling, or tracking you across other websites - they exist purely to keep your account secure and your session alive. ### Preference cookies `trace_theme` simply remembers whether you last viewed the app in light or dark mode, so you don't have to reselect it on every visit. It carries no personal information beyond that single preference value. ### Analytics cookies Google Analytics starts automatically on our production website and app, unless you previously turned analytics off or your browser sends Global Privacy Control. We do not display an automatic consent popup. It measures visits and outcomes such as account creation, trial activation, completed demo bookings and scans. Our custom conversion events do not include your email, password or entered website details. Advertising consent remains disabled. PostHog collection is disabled. Previously saved analytics opt-outs remain respected for 180 days. You can also use Global Privacy Control or the Google Analytics opt-out browser add-on linked below. ## Cookies we don't set - No advertising or retargeting cookies - No cross-site fingerprinting or device-fingerprinting scripts - No social-media tracking pixels embedded in our pages - No sale or sharing of cookie-derived data with data brokers ## How to control or opt out of cookies Most browsers let you view, block, or delete cookies through their settings - look for "Privacy", "Cookies", or "Site data" in your browser's preferences. Because every browser organises these settings a little differently, the exact path varies: in Chrome and Edge it's under Settings → Privacy and security → Cookies and other site data; in Firefox it's Settings → Privacy & Security → Cookies and Site Data; in Safari it's Settings → Privacy → Manage Website Data. You can typically choose to block third-party cookies only, clear cookies on exit, or block all cookies from a specific site. Keep in mind that blocking the essential `trace_session` and `trace_csrf` cookies will prevent the signed-in dashboard from working - you simply won't be able to stay logged in. Everything outside the authenticated app (the marketing site, free tools, blog, and docs) will continue to work normally without them. If you'd like to opt out of Google Analytics specifically, without affecting any other cookies, Google provides a dedicated [opt-out browser add-on](https://tools.google.com/dlpage/gaoptout) that prevents its JavaScript from sharing visit information. ## Do Not Track Some browsers offer a "Do Not Track" signal. There isn't yet a common industry standard for how sites should respond to it, so we don't currently change our behaviour based on that signal - but the controls above let you achieve the same outcome directly. ## Changes to this policy We may update this Cookie Policy as our product, providers, or applicable law change. If we make a material change, we'll update the "Last updated" date above; for significant changes we'll aim to give existing customers reasonable notice (for example, by email or an in-app notice). ## Contact Questions about this policy or how we use cookies? Reach us at **[karora@citeagentic.com]**. --- # Privacy Policy > How CiteAgentic collects, uses, discloses and protects your personal information under the Australian Privacy Act 1988 (Cth). Source: https://www.citeagentic.com/legal/privacy _Legal_ ## Privacy Policy Last updated: [26 May 2026] This Privacy Policy explains how **[CiteAgentic Pty Ltd, ABN 54238243639]** ("CiteAgentic", "we", "us", "our") handles personal information. We are an Australian business and we comply with the *Privacy Act 1988* (Cth) and the Australian Privacy Principles ("APPs"). Where we collect personal information about individuals in the EU/UK, we also apply equivalent protections. By using CiteAgentic, you agree to the handling of your information as described here. ## What this policy covers This policy applies to personal information we collect through: - The CiteAgentic website and marketing pages - Sign-up, onboarding and your authenticated use of the product - Communications you have with us (email, support, scheduling) - Payments you make as a paying customer It does **not** cover third-party websites we link to, or the practices of the AI engines and other providers described below, which have their own privacy policies. ## Information we collect | Category | Examples | |---|---| | **Account data** | Name, email, password hash, and OAuth tokens (encrypted) for services you connect, such as Google Search Console | | **Brand & input data** | Brand names, domains, prompts, competitors, ICPs and products you submit for tracking and auditing | | **Generated data** | Scan results, scores, recommendations and citation data produced when we query AI engines on your behalf | | **Usage data** | Pages and dashboards you visit, features you use, and diagnostic/error logs | | **Payment data** | Billing details processed by our payment provider (we do not store full card numbers) | | **Support data** | The content of messages you send us | We do **not** collect: your own AI-engine API keys (we run engines on our infrastructure with our keys), or your source code. ## How we collect it We collect information directly from you (when you sign up, configure tracking, pay or contact us), automatically (through cookies, analytics and server logs), and from third parties you authorise (for example, Google when you connect Search Console). ## Why we use it We use personal information to: - Provide, operate and improve the audits, monitoring and recommendations you signed up for - Authenticate you and secure your account - Process payments and manage your subscription - Send transactional messages (digests you opted into, password resets, service notices) - Compute aggregated, de-identified cohort benchmarks (minimum-N gated so they cannot be re-identified to an individual brand) - Provide support and respond to your enquiries - Meet our legal and record-keeping obligations ## Disclosure to third parties (subprocessors) We disclose personal information to the service providers below so we can run CiteAgentic. We do not sell or rent personal information. *Please confirm/adjust this list to match your live stack:* | Provider | Purpose | Location | |---|---|---| | Stripe | Payments and billing | USA / Australia | | Resend | Transactional email | USA | | Render | Application hosting | USA | | Supabase (Postgres) | Database hosting | [REGION] | | Google Analytics | Website analytics | USA | | Google (Search Console, Maps, OAuth) | Connected-account data you authorise | USA | | DataForSEO | Search-results and AI-overview data | USA / EU | | AI engines - OpenAI, Anthropic, Google, Perplexity, Brave, xAI, Microsoft | We send your prompts to these engines and receive their responses; we do not send them your account details | USA / global | ## Overseas disclosure (APP 8) Several of our providers are located outside Australia, primarily in the United States and the European Union. Before disclosing personal information overseas we take reasonable steps to ensure the recipient handles it consistently with the APPs, including through contractual data-protection terms. By using CiteAgentic you acknowledge these overseas disclosures. ## A note on AI engines To monitor your AI search visibility, the prompts you configure are sent to third-party AI engines. Do not include personal or confidential information in prompts that you do not want transmitted to those engines. ## Cookies and analytics We use a session cookie for authentication, a CSRF cookie for security, and Google Analytics cookies, with automatic collection and opt-out controls described in our Cookie Policy, to understand visits and product outcomes. See our [Cookie Policy](/legal/cookies) for the breakdown. ## Data security We protect personal information with measures including encryption of OAuth tokens, encrypted transport (HTTPS), access controls and reputable hosting providers. No system is perfectly secure, but we take reasonable steps to protect your information and to respond to any data breach in line with the Notifiable Data Breaches scheme. ## Retention We keep account data for as long as your account is active. When you delete your account there is a 30-day grace period, after which your data is erased from our databases and backups, except where we are required to retain certain records (for example, tax and transaction records) under Australian law. ## Accessing and correcting your information (APP 12 & 13) You can access and correct most of your information directly in Settings. You may also: - **Export** your data - our REST API exposes it, and we will email a JSON copy on request - **Correct** inaccurate information - **Delete** your account and data - **Object** to or opt out of direct marketing at any time To make a request, email **[karora@citeagentic.com]**. We will respond within a reasonable period (usually within 30 days) and may need to verify your identity first. ## Direct marketing (APP 7) We only send marketing communications you have opted into, and every such message includes an unsubscribe option. ## Children CiteAgentic is a business tool not directed to anyone under 16, and we do not knowingly collect their personal information. ## Complaints If you have a privacy concern, contact us first at **[karora@citeagentic.com]** and we will investigate. If you are not satisfied with our response, you may complain to the Office of the Australian Information Commissioner (OAIC) at [oaic.gov.au](https://www.oaic.gov.au). ## Changes to this policy We may update this policy from time to time. We will post the updated version here and, for material changes, notify you by email. ## Contact **[CiteAgentic Ltd]** [Gold Coast, QLD, Australia] Email: getlaunchy@gmail.com --- # Digital Marketing Agency for Brisbane & Gold Coast > Digital marketing agency support for Brisbane and Gold Coast businesses. Connect SEO, AEO, Google Ads and social content to qualified enquiries. Source: https://www.citeagentic.com/digital-marketing _Brisbane and Gold Coast_ ## Choose the channel for the job More channels do not automatically mean more customers. We review your offer, existing performance, service area and capacity before recommending where to spend time and budget. Work is scoped around priorities and deliverables, with implementation responsibilities agreed before it starts. | If you need to… | Consider… | What the work focuses on | | --- | --- | --- | | Improve relevant organic discovery | [SEO services](/services/digital-marketing/aiseo) | Technical priorities, service-page intent, local relevance and measurement | | Understand AI recommendations | [Managed AEO and GEO](/digital-marketing/ai-visibility-aeo) | Buyer questions, mentions, citations and evidence gaps | | Capture existing search demand | [Google Ads management](/digital-marketing/google-ads) | Campaign structure, conversion tracking, budget and landing pages | | Publish consistent, useful content | [Social media management](/digital-marketing/social-media-management) | Platform choice, content production, approvals and reporting | | Evaluate paid AI discovery | [ChatGPT advertising](/digital-marketing/chatgpt-ads) | Eligibility, offer readiness and a measured test plan | ## What a practical engagement includes **An agreed starting point** Review the website, available analytics and current campaigns. Identify the commercially important service, audience and market, plus the evidence missing from the decision. **Work with a named purpose** Define pages, campaigns and content deliverables, approval steps and responsibility for access and implementation. Separate media spend and third-party costs from management work. **A review that informs the next action** Compare completed work and results against the agreed objective. Explain what to continue, change or stop, using lead quality and conversion evidence alongside channel metrics. ## Serving Brisbane and the Gold Coast We can plan campaigns and content for businesses serving Brisbane, the Gold Coast or both. Targeting follows your actual service area and customer demand. Location-specific pages should earn their place through useful local information, not duplicated copy with a suburb name swapped in. These are service areas, not claims of separate offices. ## What we measure Agree conversion events before launch, such as a qualified enquiry, booked consultation or completed purchase. Search Console impressions and clicks help assess organic discovery; advertising reports help assess spend and conversions; AI scans report sampled mentions and citations. These are separate measures and are not automatically attributable to the same sale. ## Before we recommend a budget Bring your website, priority service, target market and any existing account access or reports. We clarify scope, management fees, expected client input and reporting cadence in the proposal. We do not promise a ranking, a lead volume or an AI recommendation. **Get a channel plan built around your next business priority.** Tell us what you sell, where you serve customers and what a qualified enquiry looks like. We will scope the next useful work from there. --- # CiteAgentic integrations — connect your CMS, analytics and search data > Connect the tools CiteAgentic reads from and publishes to — WordPress, GitHub, Wix, Webflow, Contentful, Storyblok, Contentstack, Sanity, Drupal — plus how to let our scanner reach your site. Source: https://www.citeagentic.com/docs/integrations _Documentation · Integrations_ ## Connect the tools you already publish with CiteAgentic finds the buyer questions you lose in AI answers and drafts the content that answers them. Integrations cover the two ends of that loop: letting us read your site, and getting approved content back out into your CMS. **TL;DR:** **TL;DR:** CiteAgentic publishes approved drafts into nine CMS platforms — GitHub, WordPress, Wix, Webflow, Contentful, Storyblok, Contentstack, Sanity and Drupal — configured per site under Site Settings → Integrations. Every credential is encrypted with AES-256-GCM and verified with a live **Test & connect** call before it is saved. Separately, [scanner access](/docs/integrations/scanner) covers allowlisting our crawler so audits see your whole site. ## Which CMS platforms can CiteAgentic publish to? There are nine: GitHub, WordPress, Wix, Webflow, Contentful, Storyblok, Contentstack, Sanity and Drupal. Approved drafts publish straight into your CMS. Every channel is configured the same way — **Site Settings → Integrations → Publishing channels** — and stores its credentials encrypted. Pick the one that matches where your site actually lives. - [GitHub](/docs/integrations/github) — for sites built from a repo: Astro, Hugo, Next.js, Jekyll, Gatsby, Eleventy, Docusaurus. Commits Markdown; your existing deploy pipeline takes it live. - [WordPress](/docs/integrations/wordpress) — via an Application Password. Publishes as a page or a post, draft or live. - [Wix](/docs/integrations/wix) — adds items to a Wix CMS collection using a Wix API key. - [Webflow](/docs/integrations/webflow) — creates CMS items via the Data API v2, staged or live. - [Contentful](/docs/integrations/contentful) — creates entries in your content model with a Content Management token. - [Storyblok](/docs/integrations/storyblok) — creates stories in the content type you choose. - [Contentstack](/docs/integrations/contentstack) — creates entries and optionally publishes them to an environment. - [Sanity](/docs/integrations/sanity) — creates documents with a real Portable Text body, not a string dump. - [Drupal](/docs/integrations/drupal) — creates nodes through core JSON:API. No contributed modules. You can connect more than one. Each channel belongs to a site, and a draft is published to the channel you pick at publish time. ## How do I let the scanner reach my site? Add an allowlist rule for our User-Agent in your firewall or CDN. It is one line, and it is the fix whenever an audit reports pages it could not fetch. - [Scanner access](/docs/integrations/scanner) — how our scanner identifies itself, and how to allowlist it so audits collect complete data. We never spoof a browser or a bot to get past your security tools. ## Which publishing channel should I use? Match the channel to how your site is built, not to what you would prefer to use. | If your site is | Use | | --- | --- | | Generated from a Git repository | [GitHub](/docs/integrations/github) | | Running on WordPress | [WordPress](/docs/integrations/wordpress) | | Built in Wix or Webflow | [Wix](/docs/integrations/wix) or [Webflow](/docs/integrations/webflow) | | Backed by a headless CMS | [Contentful](/docs/integrations/contentful), [Storyblok](/docs/integrations/storyblok), [Contentstack](/docs/integrations/contentstack) or [Sanity](/docs/integrations/sanity) | | Running on Drupal | [Drupal](/docs/integrations/drupal) | ## What happens to my credentials? Every secret you paste is encrypted with AES-256-GCM before it is stored, and it is never shown again — not in the UI, not in an export, not to support. Connections are per site, so a credential added to one site is not readable from another. Each channel also has a **Test & connect** step that does a real read against your account before saving, so a wrong token, a missing permission or a required field we cannot fill is reported at setup rather than on your first scheduled publish. --- # AI Visibility Checker for Brand Mentions > Check whether sampled AI answers name your brand, recommend competitors or cite your website. Review answer evidence and choose the next action. Source: https://www.citeagentic.com/features/aeo-scan _Brand mentions and competitor answers_ ## Check the questions that could lead to a sale A useful AI visibility check starts with a specific buyer, need and market. “Payroll compliance support for an Australian enterprise” is a more actionable test than asking an engine whether it knows a brand name. Keep branded questions separate from unbranded discovery questions when judging performance. **Read the answer behind the metric** See whether the response names your business, recommends it or names a competitor. Inspect the answer before deciding that a mention represents a meaningful buying opportunity. **Inspect the available citations** Review the URLs returned by citation-capable engines. A mention and a link to your website are different signals; an answer can contain either, both or neither. **Choose a grounded next step** Use competitor wins, missing topics and audit findings to decide what to investigate. A competing brand appearing in an answer is evidence of a gap, not proof of the exact cause. ## Run, interpret and check again 1. Add your website and business context, then choose buyer questions and supported engines. 2. Run the check and review which responses completed. Failed responses are not evidence of brand absence. 3. Inspect missed questions alongside page readiness and competitor citations. 4. Improve a relevant page or evidence source, then compare future responses using the same question set. **Illustrative result:** your brand appears in three of ten valid answers and a competitor appears in six. Read those answers and their sources before deciding whether to improve a comparison page, a service explanation or supporting proof. These numbers demonstrate interpretation, not a customer result. ## A checker is a snapshot; a tracker builds the history Answers can change across engines, dates, locations and wording. This checker samples the selected prompts and engines. It does not measure every conversation in consumer ChatGPT or guarantee the same answer for every person. Use the [AEO tracker](/features/aeo-tracker) for ongoing comparisons. Use the [AEO checker](/tools/will-ai-recommend-this-page) when you want to inspect page readiness without asking AI engines. Learn how to interpret an [AI visibility score](/features/ai-visibility/score). ## Coverage and limits Available engines, prompt allowances, scan frequency and credits are shown in your current plan. Some responses have no source links. Engine errors, cached daily results and cost or usage limits can reduce completed coverage; review scan status before comparing totals. **Find the buyer questions where your brand is missing.** Start with a limited free preview or choose a plan for broader coverage. Paid trials require a card; review the terms at checkout. --- # SEO Audit Software for Finding and Fixing Issues > SEO audit software that turns technical and on-page findings into a prioritised fix queue. Review evidence, act on issues and check the next scan. Source: https://www.citeagentic.com/features/seo-scan _Technical and on-page audits_ ## From audit report to a shorter fix queue **Find the access and indexing problems** Check response codes, redirects, robots directives, canonical tags and sitemap signals. A blocked or misdirected page needs attention before copy improvements can help it. **Review the page evidence** Inspect titles, descriptions, heading structure, structured data, links and image attributes. Findings identify the observed issue and page scope so you can assess the right fix. **Prioritise the work** Group related issues and compare severity, impact and effort. Where your plan includes fix prompts, use the finding context to brief your developer or AI coding tool. ## Choose the scan depth your site needs A single-page check is a useful starting point. Deeper scans can assess multiple pages, internal linking, near-duplicate content and thin content within the configured crawl scope. Availability, page limits and AI-assisted checks depend on your plan and scan settings. A crawl is not a promise that every URL on a large site was inspected. Connect Google Search Console to add available search-performance, sitemap and URL-inspection evidence. These checks use the permissions and data Google makes available. Performance testing is a separate scan; a lab result and real-user field data should be interpreted separately. ## A practical fix-and-check workflow 1. Run a baseline and inspect the affected URLs, rather than reacting only to a score. 2. Choose a high-impact issue you can verify, such as an unintended noindex directive on a service page. 3. Apply the change in your website, review it and run a follow-up scan. 4. Compare the finding before and after. Measure search traffic and enquiries over time; a passed check is not a ranking guarantee. **Illustrative example:** a service page returns a redirect to an unrelated location. Correct the destination, then check the response and canonical again. This is an example workflow, not a customer result. ## Free check or ongoing audit software? Use the [free SEO audit](/tools/free-seo-audit) for an initial URL diagnostic. Choose the account workflow when you need site context, scan history and a managed queue of issues. For answer structure and evidence gaps, pair it with the [page content audit](/features/aeo-page-audit). For observed AI mentions, use the [AI visibility checker](/features/aeo-scan). **Start with the page that matters most to revenue.** Check a key service or product page, review the findings and choose whether ongoing scans fit your workload. Compare current allowances on the pricing page. --- # Performance - dual-profile Core Web Vitals > Lighthouse lab + real-world 2026 throttle. Tells you which gate is lying - your CI or your users. CrUX field data, opportunity list, paste-ready fix prompts. Source: https://www.citeagentic.com/features/perf-scan _Performance_ ## Dual-profile Core Web Vitals. Tells you which gate is lying - your CI, or your users. Most perf audits run Lighthouse and call it done. We run Lighthouse and a separate real-world 2026 throttle profile, then surface the delta. When your lab score is 95 and your CrUX field data says 'poor', we tell you exactly what's hiding in the gap. ### Both profiles, every metric **Lab profile** Lighthouse on the same throttle CI uses. LCP, INP, CLS, FCP, TTFB, total page weight, render-blocking resources, image-format opportunities. **Real-world profile** 2026 mobile throttle (slower CPU, slower network than Lighthouse defaults). Plus CrUX field data when available - your actual users' p75 LCP/INP/CLS from Chrome. **Delta analysis** Per-metric `delta_*` finding when lab and real-world diverge by 20%+. "Lighthouse is masking a real-world LCP regression - your CI gate is lying to you" - with a fix prompt for the specific bottleneck. **Lighthouse mismatch ruleId** A category-level finding when the gap is large enough to suggest your CI gate doesn't reflect user experience at all. The most consequential perf finding most teams have never seen. **Run it on any URL. Free.** Pro tier adds weekly cadence + regression alerts. The audit itself never costs anything. --- # AEO Tracker — Monitor Your Brand in AI Search > Use CiteAgentic’s AEO tracker to monitor brand mentions, AI citations and competitors over time. Compare repeated answers and prioritise your next improvement. Source: https://www.citeagentic.com/features/aeo-tracker See if your business makes the answer{' '}and what’s missing with an AEO tracker.} sub="Track mentions of your business, see where competitors appear instead, and identify content gaps to prioritise." primary={{ label: 'Track my business', href: '/?signup=1' }} secondary={{ label: 'See plans & pricing', href: '/pricing' }} visual={ Three questions. Clear next steps. AM I MISSING BUYER OPPORTUNITIES?Find out where AI leaves you out.See which relevant buying questions bring up your brand—and which don't. WHO MAKES THE SHORTLIST?See who buyers hear about instead.Compare competitors in the same answers and inspect the sources AI uses. WHERE SHOULD I SPEND MY TIME?Choose a focused improvement.Use the findings to prioritize missing content, clearer product information or website fixes. } /> What you're buyingAEO tracking.Without checkingevery answer yourself. An AEO tracker monitors how your business appears in AI answers to questions your customers might ask. AEO stands for answer engine optimization.CiteAgentic brings brand mentions, source links and competitor comparisons into one place. Scheduled monitoring helps you follow changes over time, instead of repeating searches and maintaining a spreadsheet.You choose the questions and market that matter to your business. Available engines, schedules and limits depend on your plan. Less guesswork.More focused marketing.You don't need another number to watch. You need to know which questions, pages and gaps deserve attention. Find gaps before you write more content.See where your business is absent from relevant answers. Use those questions to guide what you explain on your product, service and comparison pages. Understand what competitors are known for.Read how AI describes competing businesses and which sources it cites. Identify useful explanations or proof your own site may be missing. Check whether your visibility is changing.After you improve a page, follow the same questions over time. See whether your brand appears more often and inspect the answers behind the change. How it worksFrom your businessto your next useful action. 01 / SET UPTell us what you sell.Add your website, customers, market and competitors. Choose the buying questions you want to monitor. 02 / MONITORSee where you appear.Review mentions, citations and competitors across the AI engines included in your plan. Follow changes as new runs complete. 03 / IMPROVEWork on the right gap.Use recommendations to prioritize an improvement. Update your content or fix a page issue, then review later results. Compare like with like. Keep your buyer questions and engine selection consistent when comparing runs. A new question or engine changes the sample; it is not automatically an improvement or decline. Market context depends on the engine’s supported controls and does not reproduce every buyer’s personalised experience. One completed run establishes a baseline. A trend needs later comparable observations. Review failures separately from successful answers that do not mention you, and inspect cited URLs through AI citation tracking before choosing a content change. For a single visibility sample, explore the AI visibility scan. For the denominator and limits behind scores, read how scoring works. ### Frequently Asked Questions **Q: Can I track my business in ChatGPT answers?** Yes. CiteAgentic tracks brand mentions and available source links in answers to your selected questions on supported engines, including ChatGPT. You can inspect each answer to understand whether your brand was recommended, cited or simply mentioned. Coverage depends on your plan. **Q: How often does CiteAgentic check my visibility?** Monitoring follows your plan and configuration. Supported schedules include daily, weekly and monthly runs; some access is manual. Check the pricing page for your plan’s current frequency, engines and limits. **Q: How is an AEO tracker different from an SEO rank tracker?** An SEO rank tracker follows website positions in search results. An AEO tracker follows how AI answers mention your business and competitors, and which sources they cite. Use it alongside search analytics to understand these different ways customers discover you. **Q: Is this useful for a US SaaS business?** Yes. Choose questions about the software you sell, the customers you serve and the US market. Compare the same questions and engines over time. Results represent those monitored answers, not every user’s personalized experience. **Q: Is AEO tracking the same as GEO tracking?** The terms overlap: AEO means answer engine optimization, while GEO means generative engine optimization. Both involve understanding how your business appears in AI-generated answers. CiteAgentic tracks mentions, citations and competitors to help guide improvements. **Q: Will tracking guarantee more leads or sales?** No. Tracking shows where your business appears and helps you choose what to improve. AI answers can change, and some return no source links. Measure website visits, leads and sales separately; an increase in mentions alone does not prove an increase in revenue. Start with the questions that matter Find out if AI puts yourbusiness on the shortlist. See where you appear, who appears instead and what to improve. Track my business → THE PLATFORMAI visibility monitoring → THE SOURCESAI citation tracking → THE NEXT FIXAEO page audits → --- # AEO vs SEO - the actual difference > Why AEO isn't 'SEO with extra steps'. Different signals, different playbook, different metrics, different era. Source: https://www.citeagentic.com/resources/aeo-vs-seo _Resources · AEO vs SEO_ ## AEO is not 'SEO with extra steps'. The signals are different, the playbook is different, the metrics are different. Every team trying to do AEO with their SEO toolset is fighting the last war. Here's the actual difference, not the marketing one. **TL;DR:** SEO ranks a page for a keyword through the link graph. AEO gets a passage cited for a prompt through the citation graph. Cross-encoder rerankers truncate at ~512 tokens, so your lede decides whether you survive retrieval. A Princeton-led GEO study found that adding statistics, quotations, and sourced claims lifted brand visibility in AI answers by up to 40%. ## What changed between the PageRank era and the AI-search era? **Google PageRank era (2000–2024):** the link graph was the authority signal. You ranked because authoritative pages linked to you. Tools optimised for it - keyword density, on-page SEO, backlink profile, anchor distribution, click-through rate. [Google's own documentation on how Search works](https://developers.google.com/search/docs/fundamentals/how-search-works) describes this as a crawl → index → rank pipeline anchored in PageRank. **AI-engine era (2024–):** the citation graph is the authority signal. You're cited because authoritative *sources* (which an LLM-grade reader can extract from) talk about you in language that overlaps with how buyers ask questions. Different game. A [2023 study on Generative Engine Optimization](https://arxiv.org/abs/2311.09735) by researchers at Princeton, UMass Amherst, and Georgia Tech found that adding statistics, quotations, and source citations to content increased brand visibility in AI-generated responses by up to 40%. ## What are the six key differences between AEO and SEO? The six differences are: (1) the unit of optimisation, (2) the signal of authority, (3) the query distribution, (4) the freshness expectation, (5) the metric, and (6) the fix. Each one is a different playbook, detailed below. ### 1. The unit of optimisation SEO optimises a **page** for a **keyword**. AEO optimises a **passage** for a **prompt**. Cross-encoder rerankers truncate at ~512 tokens; your lede decides whether you survive the rerank stage. The page that ranks #1 in Google might be invisible in ChatGPT because the answer is buried at section 4. ### 2. The signal of authority SEO authority = backlinks from authoritative domains. AEO authority = citations *to* the brand from authoritative sources, with consistent entity coherence (sameAs links, named-entity density, Hearst-pattern co-occurrence). A Wikipedia link is worth more than 100 blog backlinks because Wikipedia is what AI engines extract from at training and inference time. ### 3. The query distribution SEO queries are **keywords** ("best CRM", "dentist near me"). AEO queries are **questions** ("what is the best CRM for a 5-person remote sales team", "is X-tool worth it for a startup"). The long tail moved into prompt-shape; AI engines collapse the head terms. ### 4. The freshness expectation SEO indexing is days-to-weeks. AEO retrieval is **live web search at inference time** (ChatGPT search, Perplexity, Gemini grounding). A two-day-old blog post can outrank a five-year SEO winner if the AI engine's reranker thinks it's more relevant. ### 5. The metric SEO metric: ranking position in SERPs. AEO metric: **brand visibility percentage** = (# prompts where AI engines mentioned your brand) / (# prompts run). Position no longer applies - there's no SERP, there's an answer. ### 6. The fix SEO fixes are technical (page speed, meta tags, schema, link building) plus content. AEO fixes are mostly content + structural (lead-answer placement, Hearst patterns, entity coherence, sourced claims). The sourced-claim ratio matters more than the link graph. ## Which parts of traditional SEO still matter? Most of the technical surface is still required: HTTPS, mobile-friendly pages, schema, fast LCP, and a robots.txt that doesn't block by mistake. If Google can't crawl you, AI engines often can't either (they share the GPTBot/CCBot distinction but they all use traditional crawlers as a fallback). What changes is the **where** of the optimisation work - toward content structure, citation earning, and entity coherence, away from link building and keyword density. ## What CiteAgentic audits CiteAgentic's scan consists of both audits. The SEO audit catches everything SEOptimer catches plus some they miss (DNS pack, tech-stack detection, dual-profile CWV). The AEO audit catches what no SEO tool measures (chunkability, lead-answer placement, retrieval signals, sourced-claim ratio). And the AEO **monitoring** product tracks brand visibility in actual AI engines, weekly. Tools for the era you're optimising for, not the one you optimised for last decade. **See where you sit on the AEO scoreboard.** 90 seconds, no credit card. The first scan is real, not a teaser. --- # How to connect Wix to CiteAgentic and publish content automatically > Connect a Wix CMS collection to CiteAgentic using a Wix API key, so approved drafts are added to your site as collection items. Step-by-step with screenshots, field mappings, and the exact errors you may hit. Source: https://www.citeagentic.com/docs/integrations/wix _Publishing channels · Wix_ ## Publish to your Wix CMS Connect a Wix CMS (Wix Data) collection and CiteAgentic adds each approved piece as a new item, with the title and body mapped to your own collection fields. No OAuth app, no Velo code, no plugin. **TL;DR:** **TL;DR:** Connect a Wix CMS (Wix Data) collection with a Wix API key and CiteAgentic adds each approved draft as a new collection item, mapping title and body to your own fields. You need three values: your Wix site ID, the collection ID, and an API key with Wix Data permissions. It writes to a collection rather than Wix Blog, because collections accept plain HTML while the Blog API requires Wix's Ricos JSON. Connecting Wix to CiteAgentic takes about five minutes and needs three values: your Wix site ID, the CMS collection ID that will receive the content, and a Wix API key with Wix Data permissions. Once connected, every approved draft can be pushed to Wix on demand or on a schedule. ## What is the Wix integration? The Wix integration writes each approved CiteAgentic draft into a Wix CMS (Content Manager) collection as a new item. It uses the Wix Data REST API with an API key you generate yourself, so there is no app to install and nothing to approve in an OAuth screen. It writes to a **collection**, not to Wix Blog. That is deliberate: collections accept plain HTML in a text field, while the Blog API requires Wix's Ricos rich content JSON. Writing to a collection means your draft arrives intact, and you control exactly how it renders by binding a page or repeater to that collection in the Wix editor. ## What do you need before you start? You need three things, all from your own Wix account: - **A Wix site with a CMS (Content Manager) collection** to receive the items. - **Your Wix site ID**, also called the `metaSiteId`. - **A Wix API key** with Wix Data (Content Manager) read and write permissions. If your site has never used the CMS, add a collection first. The connection test fails with a clear message when Wix Data is not enabled on the site. ## How do you create the CMS collection? Create the collection in your Wix dashboard before connecting, so CiteAgentic has somewhere to write. 1. In your Wix dashboard, open **CMS** (also labelled Content Manager). 2. Create a collection for this content, or choose an existing one. `BlogPosts` is a common choice. 3. Note the collection's **ID**, which is its system name rather than its display name. Open the collection's settings to see it. 4. Note the **field keys** that hold the title and the body. The defaults CiteAgentic assumes are `title` and `content`. Your collection needs a text field for the title and a rich-text or long-text field for the body. A slug field is optional. ## Where do you find your Wix site ID and API key? Both live in your Wix account settings, in two different places. **For the API key:** 1. Go to **Settings → API Keys** in your Wix account dashboard. 2. Click **Generate API Key** and give it **Wix Data** permissions, read and write. 3. Copy the key. It starts with `IST.` and Wix shows it exactly once, so store it somewhere safe before closing the dialog. **For the site ID:** Find it under **Settings → Headless / API**, or read the `metaSiteId` out of your Wix dashboard URL. It is a UUID, formatted like `a1b2c3d4-0000-0000-0000-eeeeffff1111`. ## How do you connect Wix in CiteAgentic? Open **Site Settings → Integrations → Publishing channels**, then click **+ Wix** and fill in the form. The fields map as follows: | Field | Required | What to enter | | --- | --- | --- | | Label | No | A name you will recognise later, e.g. "My Wix site" | | Wix site ID | Yes | The `metaSiteId` UUID | | CMS collection ID | Yes | The collection's system ID, e.g. `BlogPosts` | | Title field key | No | Defaults to `title` | | Content field key | No | Defaults to `content` | | Slug field key | No | Only if your collection has a slug field | | Wix API key | Yes | The `IST.…` key, stored encrypted and never shown again | Click **Test & connect**. CiteAgentic fetches the target collection, which validates the API key, the site ID and the collection's existence in a single call. On success it saves the channel and shows the collection's display name back to you. ## How do you publish a draft to Wix? There are two ways, and both use the same connection. **On demand:** open any draft in Content Studio and choose **Share ▾ → Publish to your Wix site**. The item is created immediately and the draft is marked published. **On a schedule:** link the Wix channel to a day in the content calendar. Once you approve that piece, it publishes to Wix on its scheduled date without further action. Each piece is inserted as a collection item, with your title in the title field and the rendered HTML in the content field. How it then appears on your live site depends on the page or repeater you have bound to that collection in the Wix editor. CiteAgentic writes the data; Wix controls the presentation. ## What errors might you see, and what do they mean? The connection test surfaces Wix's own error codes translated into plain language. These are the three you are most likely to hit: - **"Wix Data (CMS) is not enabled on this site."** Wix returned `WDE0110`. The CMS app is not installed on the site. Add a CMS collection in your Wix editor, then retry. - **"Collection … not found for this site."** The collection ID is wrong, or it belongs to a different site. Check you used the system ID rather than the display name. - **"Authentication failed."** The API key is wrong, expired, or lacks Wix Data permissions. Generate a new key with read and write on Wix Data. If a scheduled publish fails, the channel is flagged with the error in Site Settings and the failure is recorded against that draft, so nothing fails silently. ### Frequently Asked Questions **Q: Does this publish to Wix Blog or a CMS collection?** It writes to a Wix Data (CMS) collection. Collections accept plain HTML in a field and need no OAuth app, while the Blog API requires Wix\ **Q: Where do I find the collection ID?** In Content Manager, open your collection\ **Q: What permissions does the Wix API key need?** Wix Data (Content Manager) read and write. The connection test fails with a specific message if the key lacks access or if Wix Data is not enabled on the site. **Q: Which collection fields get filled?** The title field and content field you specify, defaulting to title and content, plus the slug field if you provide one. Those fields must already exist in your collection or the write fails. **Q: Is my Wix API key stored securely?** Yes. The key is encrypted at rest with AES-256-GCM and is never shown again after you save it. Only non-secret config such as the site ID and collection ID is readable back. **Q: Can I connect more than one Wix site?** Yes. Publishing channels are per site in CiteAgentic, so each of your sites can have its own Wix connection pointing at its own collection. **Ready to connect?** Site Settings → Integrations → Publishing channels → + Wix. Add your site ID, collection ID, and a Wix Data API key. --- # AI Visibility Tool — Mentions, Citations & Competitors > Evaluate CiteAgentic’s AI visibility tool. Inspect brand mentions, cited URLs and competitors in AI answers, then choose what to improve and track over time. Source: https://www.citeagentic.com/features/ai-visibility _AI Search Visibility Monitoring_ ## An AI visibility tool that shows which brands make ChatGPT’s shortlist. Track how your brand appears across ChatGPT, Google AI Overviews, Perplexity and other leading AI engines. Separate mentions, recommendations and citations - and see how each changes over time. ## What is AI search visibility monitoring? AI search visibility monitoring is the practice of continuously measuring where, and how, AI answer engines represent your brand when buyers ask questions in your category. CiteAgentic tracks it across ChatGPT, Perplexity, Gemini, Google AI Overviews, Claude, Grok, and separates three outcomes that are easy to confuse: a brand mention (AI names you), a recommendation (AI puts you on the shortlist), and a website citation (AI links your page as a source). ### A mention, a recommendation, and a citation are not the same **Mention** The AI names your brand in its answer. You are in the conversation, but not necessarily endorsed. **Recommendation** The AI puts you on the shortlist it hands the buyer - named as one of the options worth considering. **Citation** The AI links your website as a source it drew from. A separate outcome that can require different content and authority improvements. ### How AI visibility monitoring works 1. **Choose buyer questions** - select the category, comparison, and problem-based questions your buyers actually ask AI. 2. **Run them across AI engines** - CiteAgentic records the answers under consistent engine, market, and prompt conditions. 3. **Separate the outcomes** - detect mentions, recommendations, cited domains, cited URLs, position, and sentiment. 4. **Compare competitors and sources** - see which brands win the answer and which third-party sources are repeatedly cited alongside winning answers. 5. **Monitor changes over time** - track whether changes to your content, authority, and campaigns are followed by improved visibility. 6. **Retest and verify** - confirm whether your brand is now mentioned, recommended, or cited more consistently. ### Turn monitoring into action CiteAgentic connects every visibility gap to the prompt, competitor, cited source, and page involved - so you can decide whether to improve an existing page, create new content, or build independent authority elsewhere. | Signal found | Decision it enables | |---|---| | Competitors are recommended, but you are absent | Investigate their positioning and the evidence AI is rewarding | | Your brand is mentioned but never cited | Improve the relevant pages and their citation readiness | | Independent sources cite competitors instead of you | Prioritise outreach, reviews, PR, and inclusion in those sources | | Sentiment or company facts are wrong | Correct your first-party information and trusted profiles | ### See the evidence behind every score Open any tracked prompt to inspect the original AI answer, the brands recommended, the cited URLs, sentiment, and how the result changed over time. Every number on your dashboard traces back to a real answer you can read. ### Four ways monitoring reads your AI visibility **Track whether you appear** Visibility score, prompt tracking, multi-engine coverage, and geographic tracking - how often each engine names you, across the questions and markets that matter. See the AI Visibility Score → **Understand how you appear** Brand mentions, recommendations, sentiment by engine, and factual accuracy - not just whether you show up, but how AI describes you. **See what AI cites** Domain and page-level citations, Google AI Overviews, and citation trends - the exact sources and pages AI draws from for each answer. **Measure against the market** Share of voice, competitor comparison, and winning versus losing prompts, plus a weekly email digest so changes reach your inbox. ### AI answers change. Your measurement process shouldn't. Serious buyers want to know whether one AI answer reflects what their customers really see. Here is how CiteAgentic keeps monitoring trustworthy: - **Real engine answers, consistent conditions** - prompts run under controlled engine, market, and language settings, using vendor-native geo targeting where each engine supports it. - **Trend over snapshot** - AI answers vary between runs, so monitoring follows the trend across scheduled runs rather than over-reading a single response. - **Outcomes defined, not blurred** - a mention, a recommendation, and a citation are detected and reported separately. - **Sentiment and citations checked each run** - sentiment is classified per mention, and cited domains and URLs are parsed from the answer. - **Inspect the original answer** - the raw answer behind every result is stored, so you can read exactly what the engine said. Read the full [scoring methodology](/resources/how-scoring-works) for how the numbers are built. ### Monitoring powers the entire AI visibility loop Discover what buyers ask, measure where you appear, diagnose why competitors win, improve your content and authority, then retest to verify the result. Monitoring is the hub the whole loop reports back to. [Discover buyer prompts](/features/ai-prompt-research) → **Monitor visibility** → [Diagnose competitors](/features/ai-competitor-intelligence) → [Optimize content](/features/aeo-content-optimization) → [Build authority](/features/ai-recommendations-authority) → Verify results ### Frequently Asked Questions **Q: What does AI search visibility monitoring measure?** It measures where and how AI answer engines represent your brand when buyers ask questions in your category - whether you are mentioned, recommended, or cited as a source, plus sentiment and the prompts you win and lose. CiteAgentic tracks the trend over time rather than reading a single answer. **Q: What is the difference between a brand mention and a citation?** A brand mention is when an AI answer names your brand. A website citation is when the AI links your page as a source it drew from. They are different outcomes with different fixes, so CiteAgentic tracks them separately rather than collapsing them into one number. **Q: How often is my AI visibility updated?** CiteAgentic rechecks your tracked prompts on a scheduled cadence, so you follow a trend over time rather than reading a single answer. Cadence and coverage vary by plan. **Q: Can I monitor competitors as well as my own brand?** Yes. CiteAgentic tracks which competitors appear for the same buyer questions, their share of voice and position, and the sources AI cites when recommending them. Competitor limits vary by plan. **Q: Can I monitor visibility by country or market?** Yes. You can track prompts under supported geographic and language conditions to see how recommendations differ between markets. Market limits vary by plan. **Q: Which AI engines are monitored?** ChatGPT, Perplexity, Gemini, Google AI Overviews, Google AI Mode, Claude, Grok. Google AI Overviews is tracked separately because it surfaces and cites sources differently from the chat engines. Engine coverage varies by plan. **Q: How is this different from an SEO rank tracker?** A rank tracker tells you your position in a list of blue links. AI visibility monitoring reads the actual AI answer - whether you are named, described, and cited - because in AI search the engine names two or three options, and buyers may consider those options without reviewing a traditional results page. **See where AI answers name your brand.** Monitor mentions, recommendations and citations across every major AI engine, and turn each gap into a decision about content, pages, or authority. --- **Explore:** [AEO tracker](/features/aeo-tracker) · [AI citation tracking](/features/ai-citation-tracking) · [Compare plans](/pricing) · [AI Visibility Score](/features/ai-visibility/score) · [What is AEO visibility?](/resources/what-is-aeo-visibility) · [How scoring works](/resources/how-scoring-works) --- # Security Scan - headers, secrets, HTTPS, CORS > Ten-minute audit catches the four classes of cheap-to-fix security exposure. Header suite, secret scanning, transport, CORS misconfig. Source: https://www.citeagentic.com/features/security-scan _Security Scan_ ## Ten minutes. Four classes of cheap-to-fix exposure. Real risk, real fix. Most security audits are PDFs full of theory. This one runs in ten minutes, finds the four exposure classes that account for ~80% of real-world breaches at SaaS scale, and ships a paste-ready fix prompt for each. ### Four exposure classes, not 47-page checklists **Secret scanning** Stripe live/test keys, AWS access + secret keys, GCP keys, GitHub PAT/OAuth, Slack tokens, OpenAI/Anthropic keys, Twilio SID, SendGrid, Supabase anon, private keys, JWT in HTML/JS bundles. Public-side leaks only - what an attacker scrapes from your site. **Security headers** HSTS, CSP, X-Frame-Options, X-Content-Type-Options, Referrer-Policy. Five headers that close five real attack classes - and the five whose absence we see most. **Transport** HTTPS coverage, mixed content (passive + active), HSTS preload, certificate sanity. The "are users protected on the wire" set. **CORS** Wildcard origin with credentials, missing preflight handling, exposed authenticated APIs. The misconfig that lets a malicious site act on your authenticated APIs from a victim's browser. **Free, every tier.** Security audits should never be paywalled. The Pro lever is monitoring (re-run weekly, alert on regressions, link findings to your AEO-tracked prompts) - the audit itself is always free. --- # CiteAgentic API authentication — keys, scopes and plan requirements > How CiteAgentic API keys work: bearer authentication shared by the REST API and the MCP server, the scope vocabulary, plan requirements, and what happens when a key is rejected. Source: https://www.citeagentic.com/docs/api/authentication _Documentation · API_ ## One key, two interfaces The REST API and the MCP server share a single credential and a single permission model. Get a key once and both work. **TL;DR:** **TL;DR:** One API key authenticates both the REST API and the MCP server, sent as `Authorization: Bearer `. Keys carry a subset of 13 scopes that separate read, write and run — `run:` scopes spend credits, the others do not. Keys are stored as a SHA-256 hash and cannot be recovered, only replaced. API access requires a paid plan; without it every call returns `tier_disallows_api`. ## How do I get a CiteAgentic API key? **API keys are currently issued by our team rather than self-serve.** Email **support@citeagentic.com** from the address on your account and tell us which scopes you need. Self-serve key management is being built; this page will change when it ships. API and MCP access is a paid-plan capability. If your plan does not include it, every call returns `tier_disallows_api` — the API does not partially work on a plan without it. ## How do I authenticate a request? Send it as a bearer token: ``` Authorization: Bearer ``` That header is the same for `POST /mcp` and for every `GET /api/v1/*` endpoint. Keys are stored as a SHA-256 hash, never in plain text. That means we cannot show you a key again after it is issued, and we cannot recover a lost one — a lost key is revoked and replaced. ## What scopes are available? A key carries a list of scopes, and each tool or endpoint declares the one it needs. Read, write and *run* are deliberately separate, because running work spends credits and the other two do not. | Scope | Grants | | --- | --- | | `read:sites` | Site list, portfolio summary, site brief | | `read:visibility` | Visibility scores and history | | `read:audits` | Audit scores and findings | | `read:prompts` | Prompt list and per-prompt detail | | `read:competitors` | Competitor list and gap analysis | | `read:recommendations` | The recommendation queue, including fix prompts | | `read:content` | Content opportunities and drafts | | `read:voice` | Voice profile, review themes, buyer criteria | | `write:prompts` | Generate and refine prompts | | `write:competitors` | Add a competitor | | `write:content` | Generate a content draft | | `run:visibility` | Start a visibility scan | | `run:audits` | Start an audit scan | Ask for the narrowest set that does the job. A key that only reports on visibility does not need `run:` scopes, and a scan started by mistake spends credits that a read-only key could never have spent. ## What data can an API key reach? A key resolves to the user who owns it, and sees exactly what that user sees: the sites they own, plus any they collaborate on. This is enforced by row-level security in the database, not only by the route handler, so a bug in a route cannot leak another tenant's data. Quotas are the plan's quotas. Prompts, competitors and drafts created through the API count against the same limits as the UI. The API is another way into the same account, not a way around it. ## Why was my API call rejected? There are four rejections, and they mean different things: a bad credential, a plan without API access, or a key missing the scope this call needs. | Response | Meaning | | --- | --- | | `401 missing_bearer_token` | No `Authorization: Bearer` header was sent | | `401 invalid_api_key` | The key does not match any active key — check for a typo, or it was revoked | | `403 tier_disallows_api` | The key is valid, but the account's plan does not include API access | | `403 insufficient_scope` | The key is valid and the plan allows API access, but this key lacks the scope this call needs | The distinction matters when you are debugging: `invalid_api_key` is a credential problem, `tier_disallows_api` is a billing problem, and `insufficient_scope` means ask us to reissue with a wider scope list. ## Can I see what a key has done? Every call is recorded against the key that made it, including the tool or endpoint name and whether it succeeded. If you hand a key to an automation, its activity is attributable to that key rather than blended into your own. --- # Content Audit Tool for AI Answer Readiness > A page-level content audit tool for answer clarity, structure, schema and evidence gaps. Review findings, improve the page and recheck readiness. Source: https://www.citeagentic.com/features/aeo-page-audit _Page-level answer readiness_ ## What this content audit covers This is a page-level AEO content audit. It helps you assess whether an important product, service or resource page explains its topic clearly and supplies usable evidence. It is not a complete content inventory, traffic analysis or automatic consolidation plan for your entire website. **Clear answers and structure** Review whether the page addresses the question directly and uses understandable headings. Expand missing explanations with information useful to the buyer, rather than adding words to chase a score. **Evidence and business context** Inspect factual detail, source links and brand context. Correct unsupported claims and add relevant, verifiable proof. A generated recommendation is a starting point for review, not new evidence. **Technical readiness** Check the structured data and access signals available in the audit. These help explain how the page is exposed to crawlers, but passing a check cannot guarantee that an engine will cite it. ## Diagnose, rewrite, recheck 1. Choose a commercially important page and the buyer question it should answer. 2. Review the findings and confirm that each applies to the visible page. 3. Use an available fix prompt or brief to make a specific change. Keep brand claims and customer evidence accurate. 4. Run the audit again to check the issue. Then use visibility monitoring to assess mentions and citations separately. **Illustrative example:** a payroll remediation page lists services but never explains what an employer receives. A useful edit could describe the assessment, evidence required, agreed deliverables and handover. It should not invent a turnaround time, a regulatory assurance or customer outcomes. ## What to use alongside it | Your question | Next step | | --- | --- | | Can crawlers access the page? | [Robots TXT checker](/tools/will-chatgpt-find-my-website) | | What is the quick readiness baseline? | [Free AEO checker](/tools/will-ai-recommend-this-page) | | Are technical issues affecting multiple pages? | [SEO audit software](/features/seo-scan) | | Does AI actually mention or cite us? | [AI visibility checker](/features/aeo-scan) | | Is the page producing enquiries? | Review your analytics and Search Console evidence alongside the audit | ## Scope before score URL checks depend on the content the scanner can fetch. Login walls, blocked requests and client-rendered content can limit the evidence. The free pasted-text checker cannot inspect HTML-only signals such as structured data. Treat unavailable checks as unavailable, rather than a clean bill of health. Account audit depth and fix-prompt access depend on your plan. Start with a free page check, then choose the broader workflow if you need it. **Audit the answer your next customer needs.** Review a URL or draft, identify the weakest supported signals and make one verifiable improvement before checking again. --- # How to connect Drupal to CiteAgentic and publish content automatically > Connect a Drupal site to CiteAgentic using the core JSON:API module and basic auth, so approved drafts are created as nodes. Covers the two setup steps most sites miss. Source: https://www.citeagentic.com/docs/integrations/drupal _Publishing channels · Drupal_ ## Publish to your Drupal site Connect a Drupal site and CiteAgentic creates each approved piece as a node in the content type you choose, with the body written as HTML in your chosen text format. Core modules only, no contributed code. **TL;DR:** **TL;DR:** Connect a Drupal 9+ site with the core **JSON:API** and **HTTP Basic Authentication** modules enabled, plus a user account with permission to create content. CiteAgentic creates each approved draft as a node in your chosen content type, with the body written as HTML in your chosen text format. Core modules only — no contributed code. JSON:API defaults to read-only, so the two setup steps most sites miss are enabling basic auth and switching JSON:API to read-write. Connecting Drupal takes about ten minutes, and most of that is two configuration steps on the Drupal side that are easy to miss. Get those right and the connection itself needs four values: your site URL, the content type machine name, and a username and password. ## What is the Drupal integration? The Drupal integration creates each approved CiteAgentic draft as a node on your site using the core **JSON:API** module, authenticating over HTTP basic auth. Nothing contributed is required: both modules ship with Drupal core. Because Drupal's body field accepts HTML plus a text format, your article is written as real HTML rather than as markdown. Headings, lists, links and emphasis all arrive intact. By default nodes are created **unpublished**, so an editor reviews them in Drupal before anything appears on the site. ## What are the two setup steps most sites miss? These are the two reasons a Drupal connection fails, and both are on the Drupal side rather than in CiteAgentic. **1. JSON:API ships read-only.** Out of the box, JSON:API answers `GET` requests perfectly and refuses every `POST`. Go to **Configuration → Web services → JSON:API** and select **"Accept all JSON:API create, read, update, and delete operations"**. This is worth understanding because a site in read-only mode looks completely healthy from the outside. CiteAgentic probes for write mode during the connection test rather than letting you discover it on your first publish. **2. HTTP Basic Authentication is not enabled by default.** The `basic_auth` module is part of core but is off in a standard install. Without it, Drupal *ignores* the Authorization header entirely and treats the request as anonymous, which surfaces as a permissions error that looks like a wrong password. Enable it at **Extend → HTTP Basic Authentication**. ## What do you need before you start? - **Drupal 9 or later** with the **JSON:API** and **HTTP Basic Authentication** modules enabled. - **JSON:API set to accept write operations**, as above. - **A content type machine name**, e.g. `article` or `page`. - **A Drupal user** with permission to create content of that type. Create a dedicated API user with a role scoped to content creation rather than reusing an administrator account. The password is stored encrypted, but a narrower account limits the blast radius either way. ## Where do you find the content type machine name? Go to **Structure → Content types**. The machine name is shown beneath each type's label, and is lowercase with underscores, e.g. `basic_page` rather than "Basic page". Entering the display label is a common mistake, so CiteAgentic rejects anything that is not a machine name before it even calls your site. ## How do you connect Drupal in CiteAgentic? Open **Site Settings → Integrations → Publishing channels**, click **+ Drupal**, and fill in the form. The fields map as follows: | Field | Required | What to enter | | --- | --- | --- | | Drupal site URL | Yes | Base URL where `/jsonapi` is reachable | | Content type | Yes | Machine name, e.g. `article` | | Body field | No | Defaults to `body` | | Text format | No | Basic HTML (default), Full HTML, or Restricted HTML | | URL alias | No | Let Drupal decide (default), or use the CiteAgentic slug | | On publish | No | Create unpublished (default) or published | | Drupal username | Yes | A user who can create this content type | | Password | Yes | Stored encrypted and never shown again | Click **Test & connect**. CiteAgentic checks that JSON:API is present, that your credentials and content type work, and that the site accepts writes, naming whichever of those three is wrong. ## Which text format should you choose? **Basic HTML** suits most sites and is the default. It permits headings, lists, links, and emphasis while stripping anything unusual. **Full HTML** preserves every tag, which matters if your articles include embeds or custom markup. Only use it if the publishing user is trusted, since Full HTML performs no filtering. **Restricted HTML** is intended for anonymous comments and strips most structure. It is included for completeness but rarely the right choice for articles. ## Should CiteAgentic set the URL alias? Only if your site does not already generate one. If you use **Pathauto**, leave **URL alias** on "Let Drupal decide". Pathauto creates an alias from its own pattern, and writing a second alias leaves two paths pointing at the same node. If you do not use Pathauto and want readable URLs, choose "Use the CiteAgentic slug" and the node is created at `/your-article-slug`. ## How do you publish a draft to Drupal? **On demand:** open any draft in Content Studio and choose **Share ▾ → Publish to Drupal**. The node is created immediately. **On a schedule:** link the Drupal channel to a day in the content calendar. Once you approve that piece, it is created on its scheduled date. CiteAgentic links back to the node's alias if one exists, and to `/node/ID` otherwise. ## What errors might you see, and what do they mean? - **"No JSON:API found at /jsonapi."** The JSON:API module is not enabled, or your site serves it from a custom path. - **"JSON:API is in read-only mode."** Write operations are disabled. Configuration → Web services → JSON:API. - **"Authentication failed."** Wrong username or password, **or** the `basic_auth` module is not enabled. Check the module first: without it Drupal never reads the Authorization header. - **"Content type … not found."** The machine name is wrong. - **"Content type must be a machine name."** You entered a display label such as "Basic Page" instead of `basic_page`. - **"The user cannot create … content."** Credentials are valid but the role lacks the create permission for that content type. If a scheduled publish fails, the channel is flagged with the error in Site Settings and the failure is recorded against that draft, so nothing fails silently. ### Frequently Asked Questions **Q: Do I need any contributed modules?** No. JSON:API and HTTP Basic Authentication both ship with Drupal core. You do need to enable HTTP Basic Authentication, and to switch JSON:API out of read-only mode. **Q: Why does my connection test say read-only when the site works fine?** JSON:API answers GET requests by default but refuses POST until an administrator enables write operations at Configuration → Web services → JSON:API. A read-only site looks perfectly healthy from the outside, which is why CiteAgentic probes for write mode during the connection test. **Q: Are nodes published live straight away?** Not by default. Nodes are created unpublished so an editor reviews them in Drupal first. Choose **Q: Does the body arrive as HTML or markdown?** HTML. Drupal\ **Q: Will this conflict with Pathauto?** Only if you tell it to. Leave URL alias on **Q: Can I use OAuth instead of basic auth?** Not currently. This integration uses HTTP basic auth against a Drupal user, which needs only core modules. Create a dedicated API user with a narrowly scoped role rather than reusing an administrator account. **Ready to connect?** Enable JSON:API write operations and HTTP Basic Authentication in Drupal, then Site Settings → Integrations → Publishing channels → + Drupal. --- # Connect GitHub — publishing setup > Connect a GitHub repo to CiteAgentic so approved content is committed as Markdown and your static site deploys it automatically. Step-by-step with a fine-grained access token. Source: https://www.citeagentic.com/docs/integrations/github _Publishing channels · GitHub_ ## Publish to your Git-backed site Best for sites built from a repo — Astro, Hugo, Next.js, Jekyll, Gatsby, Eleventy, Docusaurus. CiteAgentic commits each approved piece as a Markdown file; your existing deploy pipeline takes it live. **TL;DR:** **TL;DR:** Connect a GitHub repo with a fine-grained personal access token carrying **Contents: read & write**, and CiteAgentic commits each approved draft as `/.md` with title and date frontmatter. Your existing deploy pipeline (Vercel, Netlify, Cloudflare Pages, GitHub Pages) takes it live. Set the folder path to whatever directory your site already reads Markdown from — `content/posts/` for Hugo, `src/content/blog/` for Astro, `_posts/` for Jekyll. ## How it works CiteAgentic commits your content as `content/.md` (the **slug is generated automatically from the title**) with title + date frontmatter. Because your site already builds from that repo, the commit triggers your host (Vercel, Netlify, Cloudflare Pages, GitHub Pages, …) to rebuild and deploy — so the post goes live with no extra step. CiteAgentic doesn't host your site; it pushes to the repo your site already deploys from. ## What you'll need - A **GitHub repository** that your website builds from. - A **personal access token** with **Contents: read & write** on that repo. ## Step 1 — Create a personal access token We recommend a **fine-grained** token, scoped to just the one repo: 1. In GitHub, click your avatar → **Settings → Developer settings → Personal access tokens → Fine-grained tokens**. (Direct link: `https://github.com/settings/personal-access-tokens/new`.) 2. Give it a name and expiration, and set **Resource owner** to the account/org that owns the repo. 3. Under **Repository access**, choose **Only select repositories** and pick your content repo. 4. Under **Permissions → Repository permissions**, set **Contents → Read and write** (Metadata: Read is added automatically). 5. Click **Generate token** and copy it — it starts with `github_pat_…` and is shown once. A classic token works too (`https://github.com/settings/tokens/new` → check the `repo` scope), but it grants access to all your repos, so the fine-grained token above is safer. ## Step 2 — Connect it in CiteAgentic 1. Open **Site Settings → Publishing channels**. 2. Click **+ GitHub** and fill in: - **Repository** — `owner/repo` - **Branch** — optional; defaults to the repo's default branch - **Folder path** — where Markdown files are committed. Defaults to `content/`. Point it at your generator's content directory (examples below). - **Personal access token** — the one you just created 3. Click **Test & connect**. ### Where do I find the folder path? It isn't a value GitHub gives you — it's a choice about your own repo, and it has one correct answer: the folder your site already reads Markdown from. Open the repo on github.com and find where your existing posts live. Whatever folder they sit in is what goes in this field. If the site has no posts yet, use the table below for your generator. **Test & connect** checks the folder for you. It reports either `Publishing into content/ (2 Markdown files already there)` — which confirms you picked the right one — or that the folder doesn't exist yet, along with the content folders it did find in the repo. A folder that doesn't exist is not an error; it's created on the first publish. But if the test names folders you recognise and yours isn't one of them, that's the signal to change it before publishing rather than after. ### Folder path by generator | Generator | Typical folder path | | --- | --- | | Hugo | `content/posts/` | | Astro (content collections) | `src/content/blog/` | | Jekyll | `_posts/` | | Next.js / Eleventy | `content/` or `posts/` | | Docusaurus | `blog/` or `docs/` (often under `website/`) | ## Publishing - On demand: **Share ▾ → Publish to your repo** on any approved draft. - Automatic: link the GitHub channel to a day in the **content calendar** and it commits on that date — your host deploys it the rest of the way. **Ready to connect?** Site Settings → Publishing channels → + GitHub. Add your repo, content folder, and a Contents:read-write token. --- # How to connect Sanity to CiteAgentic and publish articles as Portable Text > Connect a Sanity project to CiteAgentic with an API token, so approved drafts are created as documents with a proper Portable Text body. Field mappings, drafts, and error decoding. Source: https://www.citeagentic.com/docs/integrations/sanity _Publishing channels · Sanity_ ## Publish to your Sanity dataset Connect a Sanity project and CiteAgentic creates each approved article as a document in your schema type, with the body converted to real Portable Text rather than dumped in as a string. **TL;DR:** **TL;DR:** Connect a Sanity project with an API token holding Editor permissions and CiteAgentic creates each approved draft as a document in your schema type, converting the body to real **Portable Text** rather than storing it as a string. You need the project ID, the dataset name, the document type, and the token. Documents can be created as drafts for review. Connecting Sanity takes about five minutes and needs four values: your project ID, the dataset, the schema type for new documents, and an API token with write access. ## What is the Sanity integration? The Sanity integration creates each approved CiteAgentic draft as a document in your dataset, using the Content Lake mutate endpoint. It authenticates with an API token you generate yourself, so there is no app to install. By default documents are created as **Studio drafts**, so nothing appears on your site until you open the document in Sanity and publish it. ## Does it write Portable Text or a string? Portable Text, by default, and this is the main difference from our other CMS integrations. Sanity's standard rich-text field stores **Portable Text**: a flat array of typed blocks, each with a style, inline spans, and annotations. Writing markdown into that field as a plain string would leave you with raw `##` and `**` characters rendered on the page. CiteAgentic converts its markdown into Portable Text blocks before writing, so your article arrives structured. The conversion covers: - **Headings** `#` through `######` map to styles `h1`–`h6` - **Paragraphs** map to `normal` blocks - **Bulleted and numbered lists** map to blocks with `listItem: "bullet"` or `"number"` - **Blockquotes** map to the `blockquote` style - **Bold, italic and inline code** map to the `strong`, `em` and `code` marks - **Links** become `markDefs` entries of `_type: "link"` referenced by key from the span, which is how Portable Text models annotations Images and horizontal rules are dropped rather than emitted as literal markdown, because there is no plain-block equivalent and leaking syntax into prose is worse than omitting it. If your body field genuinely is a plain string, switch **Body format** to **Markdown string** and the raw markdown is written instead. ## What do you need before you start? You need four things from your Sanity account: - **A project**, and its **project ID**. This is a short alphanumeric ID such as `abc12xyz`, not the project name. - **A dataset**, usually `production`. - **A schema type** for the new documents, e.g. `post` or `article`. - **An API token** with **Editor** permissions on that dataset. ## Where do you find your project ID, dataset and token? All three live at [sanity.io/manage](https://www.sanity.io/manage). **Project ID:** open the project. The ID is shown on the project page and in the URL. It is short and alphanumeric. **Dataset:** open **Datasets**. Most projects have `production`, and some also have `development`. **API token:** open **API → Tokens** and add a token with **Editor** permissions. Sanity shows it once, so copy it before closing the dialog. ## How do you connect Sanity in CiteAgentic? Open **Site Settings → Integrations → Publishing channels**, click **+ Sanity**, and fill in the form. The fields map as follows: | Field | Required | What to enter | | --- | --- | --- | | Project ID | Yes | The short alphanumeric ID, e.g. `abc12xyz` | | Dataset | Yes | Usually `production` | | Document type | Yes | Your schema type, e.g. `post` | | Title field | No | Defaults to `title` | | Body field | No | Defaults to `body` | | Body format | No | Portable Text (default) or Markdown string | | Slug field | No | Written as `{ _type: "slug", current: "…" }` | | On publish | No | Studio draft (default) or published document | | API token | Yes | Stored encrypted and never shown again | Click **Test & connect**. CiteAgentic runs a small GROQ read against your dataset, which validates the project ID, the dataset and the token in one call. One honest limitation: a read cannot prove **write** access. If your token is read-only the connection test passes and the first publish fails. Give the token Editor permissions when you create it. ## How do drafts work? Sanity has a native draft mechanism, and the integration uses it rather than inventing one. A document whose `_id` begins with `drafts.` is unpublished and appears in the Studio awaiting review. That is what **Create as a Studio draft** does. Choosing **Create as a published document** omits the prefix, so the document is live in the dataset immediately. ## How do you publish a draft to Sanity? **On demand:** open any draft in Content Studio and choose **Share ▾ → Publish to Sanity**. The document is created immediately. **On a schedule:** link the Sanity channel to a day in the content calendar. Once you approve that piece, it is created on its scheduled date. Each piece becomes a document with your title field, the Portable Text body, and the slug object if you nominated a slug field. CiteAgentic links back to the document in your Studio, because your front end owns the public URL. ## What errors might you see, and what do they mean? - **"Authentication failed."** The token is wrong, expired, or has no access to this dataset. - **"Project … or dataset … not found."** Check the project ID is the short alphanumeric ID rather than the project name, and that the dataset exists. - **"Project ID looks wrong."** You entered a project name. Use the ID from sanity.io/manage. - **"Session not found."** Sanity's response to a malformed or revoked token. Generate a new one. - **"Sanity rate-limited the request."** Wait a moment and retry. If a scheduled publish fails, the channel is flagged with the error in Site Settings and the failure is recorded against that draft, so nothing fails silently. ### Frequently Asked Questions **Q: Will my article arrive as structured rich text?** Yes. CiteAgentic converts its markdown into Portable Text blocks, so headings, lists, quotes, bold, italic, inline code and links all arrive as real Portable Text rather than as literal markdown characters. **Q: What if my body field is a plain string?** Switch Body format to **Q: Are documents published live straight away?** Not by default. They are created with Sanity\ **Q: What permissions does the API token need?** Editor, so it can write to the dataset. The connection test only performs a read, so a read-only token passes the test but fails on first publish. **Q: What happens to images in the draft?** They are dropped from the Portable Text body. Uploading assets to Sanity requires a separate asset pipeline, so emitting markdown image syntax into structured prose would be worse than omitting it. **Q: Is my API token stored securely?** Yes. It is encrypted at rest with AES-256-GCM and never shown again after you save it. Only non-secret config such as the project ID and dataset is readable back. **Ready to connect?** Site Settings → Integrations → Publishing channels → + Sanity. Add your project ID, dataset, document type, and an Editor API token. --- # Messaging Audit - conversion + clarity scorecard > Hero, CTAs, social proof, pricing graded against a thousand SaaS pages. Per-element grouping, instant fixes, first-impression analysis. Source: https://www.citeagentic.com/features/messaging-audit _Messaging Audit_ ## Confusion kills conversions. Score your page against a thousand SaaS landings. Hero, CTAs, social proof, pricing - every element graded individually. Findings group by page element so you see all the hero issues together, all the CTA issues together. Each finding ships with an instant fix and the AI prompt to ship it. ### Six elements every SaaS landing page lives or dies on **Hero** H1 specificity, sub-headline value-prop, primary CTA visibility, hero image relevance, "first impression" gut score from the LLM judge. **CTAs** Primary CTA position, copy strength, button vs link affordance, secondary CTA logic, conversion-path clarity. **Social proof** Logo strip presence + caliber, testimonials with attribution, case studies, integration count, founder/team visibility. **Pricing** Whether pricing is shown on the homepage at all, plan comparison clarity, feature labelling, FAQ depth, "what's included" specificity. **Trust + risk** Security signals, refund/cancel policy mention, customer-data handling, contact paths, founder/about page accessibility. **First impression** A 5-second LLM-judge gut read: "what does this page sell, who is it for, would I read more". The most honest score on the page. ### 4-tier severity + per-element grouping + instant fixes The audit was rebuilt this week to match the LaunchRecord-parity feedback - sharper severity scale, findings grouped by page element instead of one long flat list, an inline 'instantFix' you can ship in a minute, plus the standard paste-ready AI fix prompt for the bigger work. **Free for the first scan.** Pro tier links findings to your AEO-tracked prompts: "your hero scores 47/100 on clarity → that's why you're invisible on these 4 prompts about CRM pricing." --- # CiteAgentic scanner: identification and allowlisting > How the CiteAgentic scanner identifies itself, why a firewall or CDN can block scans, and how to allowlist it so audits collect complete data. We never bypass or spoof our way past security tools. Source: https://www.citeagentic.com/docs/integrations/scanner _Docs · Scanner access_ ## Our scanner announces itself. Here's how to let it in. When CiteAgentic audits a page, the request identifies itself honestly in the User-Agent header. Some firewalls and CDNs block automated requests by default, which shows up in your audit as pages we couldn't fetch. The fix is a one-line allowlist rule, not a workaround. **TL;DR:** **TL;DR:** The CiteAgentic scanner identifies itself honestly in its User-Agent header and never spoofs a browser or another company's bot to get past a security tool. If your audit reports pages we could not fetch, a firewall or CDN is blocking us, and the fix is a one-line allowlist rule. This page covers the exact rule for Cloudflare, AWS WAF and the common WordPress security plugins. ## How our scanner identifies itself Every request from the CiteAgentic scanner carries this User-Agent: ``` Mozilla/5.0 (compatible; CiteAgenticBot/1.0; +https://citeagentic.com/docs/scanner) ``` Earlier versions identified as `Launchy/1.0`. If you allowlisted that token before, keep the rule or update it; matching `CiteAgenticBot` covers all current requests. The `+https://citeagentic.com` link is there so any server administrator who sees the request can find out who we are. We fetch only public pages of sites our customers have asked us to audit, we respect the page's robots directives for our purposes, and we cap how much we fetch per page and per scan. ## What we will never do Some tools try to slip past security controls by pretending to be a normal browser or, worse, by impersonating Google's or OpenAI's crawlers. We don't, as a hard policy: - **No spoofed User-Agents.** We never present ourselves as Googlebot, GPTBot, a human browser session, or anything we aren't. - **No WAF evasion.** If your firewall challenges or blocks our request, we record "couldn't fetch" honestly and tell you. We don't rotate IPs or solve challenges to get around it. - **No fetching behind logins.** Public pages only. This matters beyond principle: impersonating verified crawlers gets IPs blacklisted and produces audit data your real visitors and real AI crawlers would never see. An audit is only useful if it sees what the world sees. ## Why a scan says "couldn't fetch" If your audit shows pages we couldn't collect, one of these is almost always the cause: - **CDN or WAF bot rules**: Cloudflare, Akamai, AWS WAF and similar block or challenge unrecognised automated agents by default. - **JavaScript challenges**: interstitial pages ("Checking your browser…") that require executing a challenge before serving content. - **IP-based rate limiting or blocking**: security rules that throttle or reject repeated automated requests. - **User-Agent filtering**: rules that reject any request whose agent string isn't a known browser. None of these mean your site is broken. They mean your security layer is doing its job against unknown bots, and our scanner is, deliberately, an announced bot. ## How to allowlist the scanner Add one allow rule in your CDN/WAF for requests whose User-Agent contains: ``` CiteAgenticBot ``` or, more strictly, the full token `CiteAgenticBot/1.0; +https://citeagentic.com/docs/scanner`. - **Cloudflare**: Security → WAF → Custom rules → create a rule with expression `(http.user_agent contains "CiteAgenticBot")` and action **Skip** (bot fight mode / managed challenge). - **Akamai**: add a match target on the `User-Agent` header containing `CiteAgenticBot` with a bypass action in your bot category. - **AWS WAF**: add a string-match rule on the `user-agent` header with action **Allow**, placed above your bot-control rule group. - **Other stacks**: any rule engine that can match a request header can allow the same token. Re-run your scan afterwards. Previously blocked pages should collect normally. If they still fail, the block may be IP-based; contact us and we'll share the details your administrator needs. **See exactly which pages were blocked** Every CiteAgentic audit lists pages it couldn't fetch and why, instead of silently skipping them. Run a scan and check the Technical Accessibility section. ### Frequently Asked Questions **Q: Does CiteAgentic bypass Cloudflare or other firewalls?** No. If a firewall blocks or challenges our scanner, the audit reports the page as not collectable and tells you why. The supported fix is allowlisting our announced User-Agent, never evasion, IP rotation, or challenge solving. **Q: What User-Agent does the CiteAgentic scanner use?** Mozilla/5.0 (compatible; CiteAgenticBot/1.0; +https://citeagentic.com/docs/scanner). Earlier versions identified as Launchy/1.0. The URL in the agent string identifies us to any server administrator reviewing logs. **Q: Why does my audit say some pages could not be fetched?** Your CDN or firewall is blocking unrecognised automated agents, commonly Cloudflare or Akamai bot rules, JavaScript challenges, IP rate limits, or User-Agent filtering. Allowlist the CiteAgenticBot agent token and re-run the scan. **Q: Is allowlisting the scanner a security risk?** The rule only admits requests carrying our agent string to your public pages, which are already publicly readable. It does not disable your WAF for browsers or other bots. Strict setups can scope the rule to the full token including our URL. **Q: Does the scanner affect how AI engines see my site?** No. ChatGPT, Perplexity and the others use their own crawlers (OAI-SearchBot, PerplexityBot and friends) with their own access rules. Our guide on AI crawler access covers allowing those separately. --- # How to connect Webflow to CiteAgentic and publish content automatically > Connect a Webflow CMS collection to CiteAgentic with a Site API token, so approved drafts are created as CMS items. Covers the collection ID, the Rich Text body field, and staged vs live publishing. Source: https://www.citeagentic.com/docs/integrations/webflow _Publishing channels · Webflow_ ## Publish to your Webflow CMS Connect a Webflow CMS collection and CiteAgentic creates each approved piece as an item in it, with the body written into your Rich Text field. No OAuth app, no custom code. **TL;DR:** **TL;DR:** Connect a Webflow CMS collection with a Site API token carrying CMS read and write scope, and CiteAgentic creates each approved draft as a collection item with the body written into a Rich Text field. You need the collection ID (the hex string in its URL) and the body field's slug, which defaults to `post-body`. Items can be created staged or live. ## How it works CiteAgentic calls the Webflow Data API v2 and creates one CMS item per approved draft. Webflow requires every item to carry a `name` and a `slug`; both come from the draft title. The body is written as HTML into a Rich Text field you nominate. New items can be created **staged** or **live**. Staged items exist in the CMS immediately but only appear on the site at the next site publish, which is the safer default when someone reviews content before it ships. ## What you'll need - A **Webflow site with the CMS enabled** and at least one collection to publish into. - A **Site API token** with CMS read and write scope. - The **collection ID** of the target collection. - The **field slug** of the Rich Text field that holds the body. ## Step 1 — Generate a Site API token 1. In Webflow, open **Site settings → Apps & integrations → API access**. 2. Click **Generate API token**. 3. Give it CMS **read and write** scope. 4. Copy the token. Webflow shows it once. ## Step 2 — Find the collection ID Open the collection in the Designer's CMS panel, or open its settings. The hex string in the collection URL is the ID, for example `580e63fc8c9a982ac9b8b745`. ## Step 3 — Find the body field slug A field's *slug* is the API name, which is not always the label you see in the Designer. A field labelled "Post Body" usually has the slug `post-body`. Open the collection's field settings to confirm it. CiteAgentic defaults to `post-body`. The body field must be a **Rich Text** field. A plain-text field will store the HTML as visible markup rather than rendering it. ## Step 4 — Connect it in CiteAgentic 1. Open **Site Settings → Publishing channels**. 2. Click **+ Webflow** and fill in: - **CMS collection ID** — from step 2 - **Body field slug** — from step 3, defaults to `post-body` - **Published site URL** — optional; lets CiteAgentic link to the item once it is live - **Publish as** — staged or live - **Site API token** — from step 1 3. Click **Test & connect**. The connection test reads the collection and warns you about any **required fields it will not be filling**. That warning matters: Webflow rejects an item that is missing a required field, so a collection with a required "Featured image" or "Author" reference needs a default set in Webflow before publishing will succeed. ## Publishing - On demand: **Share ▾ → Publish** on any approved draft. - Automatic: link the Webflow channel to a day in the **content calendar**. ### Frequently Asked Questions **Q: Why did publishing fail with a validation error?** Almost always a required field on the collection that CiteAgentic does not fill, or a slug that already exists. The error message names the specific field, because we read Webflow\ **Q: Staged or live?** Staged creates the item in the CMS without putting it on the site; it appears at the next site publish. Live publishes immediately. Choose staged if anyone reviews content before it ships. **Q: Can I publish into more than one collection?** Yes. Add one Webflow channel per collection; each carries its own collection ID and body field. **Q: Does the body arrive as Markdown or HTML?** HTML, written into the Rich Text field. Point the body field at a Rich Text field, not a plain-text one. **Ready to connect?** Site Settings → Publishing channels → + Webflow. Add your collection ID, body field slug, and a CMS read-write token. --- # Peec AI Alternative | Compare Website Fix Workflows > Compare a Peec AI alternative for AI visibility, SEO and AEO findings, fix prompts and follow-up scans. Dated sources and practical evaluation steps. Source: https://www.citeagentic.com/compare/citeagentic-vs-peec # Need a Peec AI alternative that connects visibility gaps to website fixes? Peec lists daily tracking, source analysis, prompt suggestions and recommended actions. Its pricing page also lists agent actions; availability depends on the selected plan. Compare those workflows with CiteAgentic’s audit findings, fix prompts and follow-up scans, using the same business questions. Vendor-authored comparison. Published capabilities reviewed 28 September 2026; not a hands-on benchmark. ## The short answer Consider CiteAgentic when you want AI visibility evidence alongside SEO and AEO findings you can act on. Keep Peec AI on your shortlist if its reporting and optimisation workflow already fits your team. This is a vendor-authored comparison of published capabilities, not an independent product benchmark. ## At a glance | Decision | CiteAgentic | Peec AI | | --- | --- | --- | | AI answers and citations | Allowed engines depend on plan | Check selected plan and models | | Suggested next actions | Grounded in scan findings | Vendor lists recommendations | | Website fix workflow | SEO/AEO audit findings and fix prompts | Evaluate depth using a real page in your trial | | After-change measurement | Follow-up scans compare observed visibility | Test the reporting workflow you need | | Price and limits | See current pricing; tier limits apply | Check current quote and model selection | ## What should you compare beyond a visibility score? Use the same questions, language, market and observation window. Check whether an answer mentions your brand, cites a URL, or recommends a competitor. These are different signals. A sampled answer is not a count of all customer searches. Compare the evidence behind the score. ## What work can your team actually complete? In CiteAgentic, review an audit finding, its affected page and the suggested fix prompt. Check the proposed change before applying it. Run that same exercise in the other tool and judge the specificity, effort and relevance of the guidance. Choose the workflow that helps your team ship a useful fix. ## Does a better score prove a fix worked? A follow-up scan can show an observed change. It cannot isolate your edit from model changes, changing sources or answer variation. Keep the prompt set stable and review several observations alongside qualified visits and enquiries. Treat before-and-after results as evidence, not causal proof. ## Pricing and limits Published pages checked 28 September 2026. Follow the pricing links for current terms; this is not a checkout quote. Check included scans, quotas and fix-prompt access before choosing a plan. Plan selection: See pricing: Confirm currency, models and projects. Check Peec’s current pricing for your currency, billing term and required models. Confirm the full quote before purchasing. ## Before switching ### How should I evaluate a Peec AI alternative? Define the questions, markets and engines you need first. Test equivalent coverage, review cited URLs and complete one website fix. Compare the time and cost of that workflow, not only the headline score. ### Will switching improve my AI visibility? Changing software alone does not improve visibility. Results depend on the work you implement and how AI systems retrieve and answer. No tool can guarantee mentions or citations. ### Can I move my historical reports? Keep your existing exports and baseline. Recreate a representative prompt set and run both tools alongside each other. Do not assume historical scores transfer or use the same methodology. ## Sources and next steps - [Peec AI capabilities](https://peec.ai/) - [Peec AI pricing and limits](https://peec.ai/pricing) - [CiteAgentic pricing](/pricing) - [Compare AI SEO tools](/compare) - [Review CiteAgentic audit features](/features/seo-scan) --- # How to connect Storyblok to CiteAgentic and publish content automatically > Connect a Storyblok space to CiteAgentic with a personal access token, so approved drafts are created as stories in your content type. Step-by-step with field mappings, regions, and the errors you may hit. Source: https://www.citeagentic.com/docs/integrations/storyblok _Publishing channels · Storyblok_ ## Publish to your Storyblok space Connect a Storyblok space and CiteAgentic creates each approved piece as a story in the content type you choose, with the title and markdown body written into your own fields. No OAuth app, no plugin. **TL;DR:** **TL;DR:** Connect a Storyblok space with a personal access token and CiteAgentic creates each approved draft as a story in the content type you choose, writing title and Markdown body into your own fields. You need the space ID, the content type name, and the field names for title and body. Pick the right region when your space is not in the EU, or requests reach the wrong API host. Connecting Storyblok takes about five minutes and needs three values: your numeric space ID, the content type new stories should use, and a personal access token. Once connected, every approved draft can be pushed to Storyblok on demand or on a schedule. ## What is the Storyblok integration? The Storyblok integration creates each approved CiteAgentic draft as a new story in your space using the Storyblok Management API. It authenticates with a personal access token you generate yourself, so there is no app to install and nothing to approve in an OAuth screen. By default stories are created **as drafts**, so nothing goes live on your site until you open it in Storyblok and publish. You can switch that to publish immediately if you prefer. ## Why does it write markdown instead of richtext? Storyblok's Richtext field stores content as structured JSON in a ProseMirror-style tree, not as text. Writing into it reliably would mean synthesising that tree and guessing at your component nesting. Storyblok's **Markdown** and **Textarea** field types store plain strings, and CiteAgentic already renders every draft to clean markdown. So the integration writes markdown into a field you nominate. Point it at a Markdown field and your content arrives fully formatted, with headings, lists and links intact. If your content type only has a Richtext body today, add a Markdown field to it and point the integration at that. ## What do you need before you start? You need three things, all from your own Storyblok account: - **A Storyblok space**, and its numeric **space ID**. - **A content type component** for the new stories, for example `page` or `article`. It must already exist in the space. - **A personal access token** with write access to that space. You also need to know which **region** your space is hosted in. A token issued against a US space will fail against the EU host, which is the single most common setup mistake. ## Where do you find your space ID, content type and token? All three live in your Storyblok account, in three different places. **Space ID:** open your space and go to **Settings → General**. It is a long number, e.g. `288868932106293`. **Content type:** open **Block Library** and find the content-type block your pages use. You need its technical name (lowercase, no spaces), not its display name. **Personal access token:** go to **My account → Account settings → Personal access tokens**, create a token, and copy it. Storyblok shows it once. ## How do you connect Storyblok in CiteAgentic? Open **Site Settings → Integrations → Publishing channels**, click **+ Storyblok**, and fill in the form. The fields map as follows: | Field | Required | What to enter | | --- | --- | --- | | Space ID | Yes | The numeric space ID from Settings → General | | Region | No | Where the space is hosted. Defaults to EU | | Content type | Yes | The component technical name, e.g. `page` | | Title field key | No | Defaults to `title` | | Body field key | No | A Markdown or Textarea field. Defaults to `content` | | Parent folder ID | No | Numeric folder ID. Blank creates at the space root | | On publish | No | Create as draft (default) or publish immediately | | Personal access token | Yes | Stored encrypted and never shown again | Click **Test & connect**. CiteAgentic fetches the space to validate the token, the region and the space ID, then lists the space's components to confirm your content type actually exists. If it does not, the error names the ones that do, so you can correct a typo without leaving the screen. ## Which regions are supported? Storyblok shards its Management API by data residency, and CiteAgentic supports all five hosts: | Region | Management API host | | --- | --- | | EU (default) | `mapi.storyblok.com` | | United States | `api-us.storyblok.com` | | Canada | `api-ca.storyblok.com` | | Australia | `api-ap.storyblok.com` | | China | `app.storyblokchina.cn` | Pick the one matching your space. If you are unsure, try EU first: a mismatch returns an authentication error rather than doing anything harmful. ## How do you publish a draft to Storyblok? There are two ways, and both use the same connection. **On demand:** open any draft in Content Studio and choose **Share ▾ → Publish to Storyblok**. The story is created immediately and the draft is marked published. **On a schedule:** link the Storyblok channel to a day in the content calendar. Once you approve that piece, it is created in Storyblok on its scheduled date. Each piece becomes a story whose `name` and slug come from the draft title, with your title field and markdown body field filled in. Storyblok stories have no inherent public URL because your front end owns routing, so CiteAgentic links back to the story in the Storyblok editor. ## What errors might you see, and what do they mean? The connection test translates Storyblok's responses into plain language. These are the ones you are most likely to hit: - **"Authentication failed."** The token is wrong or expired, **or the Region does not match where the space is hosted**. Check the region before regenerating the token. - **"Space … not found."** The space ID is wrong, or it belongs to an account the token cannot reach. - **"Space ID must be the numeric space ID."** You entered a space name or slug. Use the number from Settings → General. - **"Content type … not found."** The component name is wrong. The message lists the components that do exist in the space. - **"Storyblok rate-limited the request."** Storyblok allows 3 requests per second on Starter and 6 on Growth and above. Wait a moment and retry. If a scheduled publish fails, the channel is flagged with the error in Site Settings and the failure is recorded against that draft, so nothing fails silently. ### Frequently Asked Questions **Q: Does this publish to a Richtext field?** No. Richtext stores ProseMirror-style JSON rather than text, so the integration writes markdown into a Markdown or Textarea field you nominate. If your content type only has a Richtext body, add a Markdown field and point the integration at that. **Q: Are stories published live straight away?** Not by default. Stories are created as drafts so you can review them in Storyblok first. Switch the **Q: Where do I find my space ID?** Open the space and go to Settings → General. It is a long number such as 288868932106293, not the space name or slug. **Q: What access does the personal access token need?** Write access to the target space. A personal access token grants access to every space your account can reach, based on your permissions in each. **Q: Can I create stories inside a folder?** Yes. Put the folder\ **Q: Is my token stored securely?** Yes. It is encrypted at rest with AES-256-GCM and never shown again after you save it. Only non-secret config such as the space ID and content type is readable back. **Ready to connect?** Site Settings → Integrations → Publishing channels → + Storyblok. Add your space ID, content type, and a personal access token. --- # Connect WordPress — publishing setup > Connect your WordPress site to CiteAgentic so approved content publishes straight to your blog. Step-by-step setup using a WordPress Application Password. Source: https://www.citeagentic.com/docs/integrations/wordpress _Publishing channels · WordPress_ ## Publish to WordPress, hands-free Connect your WordPress site once and CiteAgentic can push approved, scheduled content straight into your blog — as a draft for review, or live. Setup takes about two minutes. **TL;DR:** **TL;DR:** Connect WordPress with your username and an **Application Password** (Users → Profile → Application Passwords, WordPress 5.6+). No plugin and no OAuth app are needed. CiteAgentic creates each approved draft through the REST API as either a page or a blog post, and as a draft or published immediately. Setup takes about two minutes. ## What you'll need - A self-hosted **WordPress 5.6 or newer** site with the REST API reachable at `/wp-json` (the default). - Your **WordPress username**. - A **WordPress Application Password** (not your normal login password — WordPress blocks login passwords on the API for security). ## Step 1 — Create an Application Password 1. Log in to your WordPress admin (`/wp-admin`). 2. Go to **Users → Profile**, then scroll to **Application Passwords**. (Direct link: `https://YOUR-SITE/wp-admin/authorize-application.php`.) 3. Enter a name like `CiteAgentic` and click **Add New Application Password**. 4. Copy the generated password — it looks like `abcd EFGH ijkl MNOP qrst UVWX`. The spaces are fine to keep. WordPress only shows it once. ## Step 2 — Connect it in CiteAgentic 1. In CiteAgentic, open **Site Settings → Publishing channels**. 2. Click **+ WordPress** and fill in: - **WordPress site URL** — e.g. `https://blog.acme.com` - **WordPress username** — your admin username - **Application password** — the one you just generated - **Publish as** — `Draft` (review in WordPress first) or `Published` (live immediately) 3. Click **Test & connect**. A green "Connected as …" confirms it. ## How publishing works - Publish any approved draft on demand from **Share ▾ → Publish to your site**. - Or link the WordPress channel to a day in the **content calendar** and it publishes automatically on that date — no further action. - Each piece is created as a WordPress post using your draft's title and rendered HTML body. New posts use your chosen **Publish as** status. ### Frequently Asked Questions **Q: Why not my normal WordPress password?** WordPress core rejects account passwords over the REST API. Application Passwords are the supported, revocable way to grant API access — you can delete one anytime without changing your login. **Q: Where do I find my site URL?** It is the base address of your site, e.g. https://blog.acme.com. CiteAgentic talks to it at /wp-json/wp/v2/posts. **Q: Will posts go live immediately?** Only if you set **Q: Is my password stored safely?** Yes — credentials are encrypted at rest and never shown again in the app. **Ready to connect?** Head to Site Settings → Publishing channels → + WordPress and paste your username + Application Password. --- # How scoring works > Methodology for brand visibility, sentiment, share-of-voice, citation diversity, and the category benchmark. Min-N gates, dedup keys, the complete formula set. Source: https://www.citeagentic.com/resources/how-scoring-works _Resources · methodology_ ## How every CiteAgentic score is computed. Open methodology - every formula, every gate, every threshold. Scoring is data; data should be inspectable. **TL;DR:** Every CiteAgentic score is an inspectable formula. Brand visibility is mentioned_calls divided by succeeded_calls, times 100. Citation diversity is Shannon entropy over channels, normalised to 0 to 100. The category benchmark only reports aggregates once a cohort has at least 5 brands. Full formulas, dedup keys, and gates below. ## Brand visibility For each scan run, brand visibility is `mentioned_calls / succeeded_calls × 100`, expressed as a percent. - `succeeded_calls` = engine calls that returned an answer (excludes failed engines) - `mentioned_calls` = subset where the deterministic mention detector found at least one of the brand's name variants in the answer Per-engine visibility is the same formula scoped to one engine. Overall is across all engines. **Why deterministic detection?** Whole-word case-insensitive match across every variant the LLM extracted at onboarding (name, lowercase, common misspellings, the bare domain). Whole-word boundaries dodge the "Apple" matching "snapple" trap. An LLM-as-judge per mention would balloon cost and add noise on cases that are unambiguous. ## Sentiment After every scan, every confirmed mention is classified once via a single batched LLM call. Output: `positive | neutral | negative`. Stored on `brand_mentions.sentiment` for use in the regression rec and per-engine sentiment tabs. **Why batched?** Per-mention LLM calls scale to N×M cost (mentions × engines). One batched call per scan, with up to 60 mentions per request, scales to one call total per run. ## Share of voice Per scan, share of voice is `entity_mentions / total_succeeded_calls × 100` per entity (the brand and each tracked competitor). Sums to ≤ 100% per scan; entities not mentioned are 0%. **Time series:** the dashboard plots SoV per day, per entity. Top 3 competitors by total volume in the window are surfaced as comparison lines. ## Citation diversity Shannon entropy of channel distribution, normalised to 0–100. ``` H = −Σ p_i × ln(p_i) for each channel i with p_i = count_i / total H_max = ln(N) where N = number of observed channels diversity = round(H / H_max × 100) ``` - All citations in one channel → 0 - Perfectly distributed across all channels → 100 Used as the headline "citation diversity" stat on the citations tab. ## Citation velocity Week-over-week delta: ``` thisWeek = sum(citation_count for buckets in [now-7d, now]) lastWeek = sum(citation_count for buckets in [now-14d, now-7d]) delta = thisWeek − lastWeek deltaPct = round(delta / lastWeek × 100) if lastWeek > 0 dir = up | down | flat ``` ## Category benchmark Each tracked brand contributes one data point: its most-recent finished run's brand-visibility percentage. ``` benchmark.category = brand.category (e.g. 'b2b-saas-marketing') benchmark.brandCount = number of active brands in the category with ≥1 finished run benchmark.avgVisibility = round(mean of all brands' visibility) benchmark.p50/p75/p90 = percentile of the same series ``` **Min-N gate:** `brandCount ≥ 5`. Below the threshold the cohort is "gated" - we return the count but no aggregates so individual brands aren't reverse-engineered from the average. ## Recommendations dedup Every recommendation has a `dedup_key` (kind + content-hash). On re-mining, `INSERT … ON CONFLICT(brand_id, dedup_key) DO NOTHING` preserves the prior status. So a `dismissed` rec stays dismissed even if the underlying condition still applies. Sentiment regression recs bucket the dedup key weekly so a recurring regression re-emits. ## Audit scoring Each platform's scoring lives in `server/scoring.js`. Common pattern: ``` score = 100 − (15 × critical_count + 5 × warning_count + 1 × info_count) × category_multiplier × compound_multiplier ``` Capped at 0. Category multipliers are tuned so foundational/gating signals (Bot Access, Server-render, Retrieval, Chunkability) outweigh ones that are over-measured industry-wide (e.g. another schema warning). Compound multipliers fire when paired failures co-occur: JS-only render + no JSON-LD = 1.6×, no FAQ schema + no question-form headings = 1.4×. **See the methodology applied to your own brand.** Free scan, no credit card. Walk through the actual numbers and how they were computed. --- # Google Ads Brisbane Management > Google Ads management for Brisbane and Gold Coast businesses. Plan campaigns, conversion tracking, landing pages and budgets around qualified leads. Source: https://www.citeagentic.com/digital-marketing/google-ads _Paid search management_ ## Management starts with the lead you want A form submission is not always a good lead. Before planning keywords or spend, agree which services, locations and customer types the campaign should attract. For a Brisbane business also serving the Gold Coast, the locations and budgets should reflect real capacity rather than an assumed statewide reach. **Campaign setup and search intent** Plan campaign structure, keyword themes, ads, negative keywords and location settings. Match the search need to the relevant service page and a clear next step. **Conversion measurement** Agree which enquiries, calls or sales should count. Review the tracking setup and test the conversion path before using reported conversions to guide budget decisions. **Ongoing optimisation** Review available search terms, wasted spend, conversion quality, ads and landing-page friction. Make changes against the agreed goal and explain the reason for each priority. ## Landing pages are part of the decision An ad can bring a visitor to the wrong page faster. We review whether the landing page explains the service, answers buying questions, provides genuine proof and makes the enquiry step easy. Copy and development work are scoped explicitly, rather than assumed to be included in every management engagement. ## What the reporting should tell you - How much was spent and which campaigns or searches drove the available conversions. - Whether those enquiries fit your service, market and customer criteria. - Where measurement is incomplete, including calls or offline sales not yet connected. - Which budget, targeting or landing-page change is recommended next and why. Where lead-quality feedback is available, use it alongside cost per conversion. Click-through rate and cheap clicks alone do not establish business value. ## Fees, ownership and scope Management fees, setup work, media budget and optional landing-page work are separated in the proposal. Your advertising budget is not an unlimited allowance bundled into a generic service price. Agree account access, approval responsibilities and reporting cadence before launch. Results depend on competition, offer, budget, website and the quality of measurement; there is no guaranteed cost per lead. ## When another channel may fit better If buyers need more explanation before they enquire, fix the service page first. If the goal is organic discovery, review [SEO services](/services/digital-marketing/aiseo). If you want a coordinated channel plan, start with our [digital marketing agency service](/digital-marketing). **Discuss Google Ads for your Brisbane or Gold Coast service.** Share your website, priority service, target area and current spend if you already advertise. We will review the scope before recommending a campaign. --- # Connect Contentful — publishing setup > Connect your Contentful space to CiteAgentic so approved content publishes straight into your content model. Step-by-step setup using a Content Management token. Source: https://www.citeagentic.com/docs/integrations/contentful _Publishing channels · Contentful_ ## Publish to Contentful, hands-free Connect your Contentful space once and CiteAgentic can create approved, scheduled content as entries in your content model — as a draft for review, or published live. Setup takes a couple of minutes. **TL;DR:** **TL;DR:** Connect a Contentful space with a Content Management token and CiteAgentic creates each approved draft as an entry in your content model, as a draft or published. You need the space ID, the content type's API identifier, and the field IDs for title and body. The body field must be **long text (Markdown)**, not Rich Text — Rich Text requires a document AST that a Markdown string cannot satisfy. ## What you'll need - A **Contentful space** with a content type (model) for your articles — e.g. `blogPost`. - The content type's **API identifier** and the **field IDs** for its title and body. The body should be a **Text (long text)** field — point CiteAgentic at it and we write Markdown into it. (Rich text fields aren't supported; they need a special document format.) - A **Content Management token** (CMA token). ## Step 1 — Create a Content Management token 1. In Contentful, go to **Settings → API keys**. 2. Open the **Content management tokens** tab and click **Generate personal token** (or create a Personal Access Token from your account settings). 3. Give it a name like `CiteAgentic` and copy the token — it looks like `CFPAT-…`. Contentful only shows it once. ## Step 2 — Find your IDs - **Space ID** — **Settings → General settings → Space ID**. - **Environment** — usually `master` (leave blank to default to it). - **Content type ID** — **Content model → your model → (⋯) → Edit → API identifier**. - **Title / body field IDs** — the API identifiers of the fields on that model (defaults: `title` and `body`). ## Step 3 — Connect it in CiteAgentic 1. Open **Site Settings → Publishing channels**. 2. Click **+ Contentful** and fill in: - **Space ID**, **Environment** (or leave `master`) - **Content type ID** — e.g. `blogPost` - **Title field ID** / **Body field ID** — default to `title` / `body` - **Locale** — defaults to `en-US` - **Publish as** — `Draft` (review in Contentful first) or `Published` (live immediately) - **Content Management token** — the `CFPAT-…` token from Step 1 3. Click **Test & connect**. A green "Connected to content type …" confirms it. ## How publishing works - Publish any approved draft on demand from **Share ▾ → Publish to your space**. - Or link the Contentful channel to a day in the **content calendar** and it publishes automatically on that date — no further action. - Each piece is created as a Contentful entry using your draft's title and Markdown body. New entries use your chosen **Publish as** status. ### Frequently Asked Questions **Q: Why a Content Management token and not a Delivery key?** Delivery/Preview keys are read-only. Creating and publishing entries requires the Content Management API, which uses a CMA (personal access) token. You can revoke it anytime. **Q: My body field is Rich text — will it work?** No. Rich text fields store a structured document AST that plain Markdown cannot satisfy. Use a Text (long text) field for the body, or add one to your model and point CiteAgentic at it. **Q: Will entries go live immediately?** Only if you set **Q: Is my token stored safely?** Yes — credentials are encrypted at rest and never shown again in the app. **Ready to connect?** Head to Site Settings → Publishing channels → + Contentful and paste your space ID + Content Management token. --- # AI Visibility Score Explained: Mentions and Citations > Understand your AI visibility score, mention rate, citation rate and sample coverage. Learn what changes mean and which action to take next. Source: https://www.citeagentic.com/features/ai-visibility/score _Understand your results_ ## Start with the metric label and its denominator AI visibility is a family of measures. Brand mentions, recommendations and website citations answer different questions and should not be added into one percentage. In CiteAgentic, inspect the metric label, reporting window, questions and completed engine responses before comparing results. | Measure | What it tells you | How to read it | | --- | --- | --- | | Named in AI answers | The business is explicitly named | Named responses divided by valid responses | | Recommended | The response recommends the business | Recommended responses divided by valid responses | | First choice | The business is classified as the top choice | Top-choice responses divided by valid responses, not simply the first name in a list | | Website used as a source | The answer cites your website | Owned citations divided by citation-eligible responses | | Any AI presence | Named or cited at least once | The union of mentions and owned citations, counted once per response | **Illustrative calculation:** a brand named in 8 of 20 valid answers has a 40% mention rate. If six of those eight also cite its website, they are still eight responses with any presence, not fourteen. Failed requests are excluded from valid-response rates. No valid responses means unavailable, not 0% visibility. ## How the windowed Presence Score differs Where the report displays the windowed Presence Score, it combines repeated observations for each question, engine and region. For each observed combination, the calculation is `(present responses + 0.5) / (valid responses + 1)`. It averages these values using the configured weights, defaults to equal weight per combination, multiplies by 100 and rounds. This smoothing makes limited history less extreme. A single successful observation does not make a combination permanently 100% visible. Unobserved combinations do not become zero scores; coverage and confidence must be read alongside the result. Snapshot rates use the most recent response per question and engine; the windowed score uses the history available in its reporting window. ## What a change does and does not prove Compare the same prompts, markets, engines and time windows. Adding harder questions or a new engine changes the sample and can change the score without your existing pages becoming worse. Confidence grows with coverage and repeated valid observations. A small movement in a small sample is a reason to inspect answers, not declare a win. A readiness score assesses a page against audit checks. A visibility score measures sampled answer presence. Neither is a Google ranking, a percentage of all AI users reached, or an estimate of sales. ## Turn the number into the next useful action **Low mention rate** Review commercially important questions where competitors appear. Check whether your page answers the need and whether the cited evidence suggests a content or authority gap. **Mentions without owned citations** Inspect the source URLs and which claims they support. Improve relevant pages and evidence where justified; do not assume every engine response can return citations. **A changing score** Check completed coverage and changes to the prompt set first. Then compare the actual responses before attributing the movement to a website edit. Use the [AI visibility checker](/features/aeo-scan) for a baseline, [AEO tracker](/features/aeo-tracker) for repeat observations and [citation tracking](/features/ai-citation-tracking) to inspect sources. Fix priorities should come from evidence and business value, not the score alone. --- # ChatGPT Advertising Readiness for Australian Businesses > Plan ChatGPT advertising for an Australian business. Review eligibility, offers, landing pages and measurement before agreeing an advertising test. Source: https://www.citeagentic.com/digital-marketing/chatgpt-ads _Paid AI discovery_ ## Can Australian businesses advertise in ChatGPT? **Availability checked 28 September 2026:** OpenAI lists Australia as available for self-service Ads Manager. Eligibility depends on the legal entity advertising and being billed, as well as the applicable account and advertising requirements. Availability can change, so verify the current position in [OpenAI's Ads Manager availability guide](https://help.openai.com/en/articles/20001245-ads-manager-availability) before planning a launch. An eligible country is not a guarantee of account approval, delivery or results. Use [OpenAI's advertiser site](https://ads.openai.com/) for the current entry process. CiteAgentic does not claim OpenAI partnership, certification or preferential ad access. ## What we can help you prepare **Offer and audience fit** Clarify what the business sells, who should act and what makes the offer credible. Map relevant buyer questions to campaign ideas without claiming access to private ChatGPT conversations. **Creative and landing-page readiness** Review message clarity, factual claims, proof, FAQs and the conversion path. Scope creative and page changes around the ad formats and account capabilities actually available. **A measurable test plan** Agree the budget, conversion event, reporting approach and conditions for continuing or stopping. Campaign implementation and management are separately scoped once access and requirements are confirmed. ## Paid placement and organic answers are different OpenAI describes ads as separate from ChatGPT's answers. Buying an ad does not buy an organic recommendation or a better answer about your company. Ads are not an endorsement by ChatGPT. See [OpenAI's explanation of ads in ChatGPT](https://help.openai.com/en/articles/20001047-ads-in-chatgpt). Measure paid clicks and conversions separately from sampled AI mentions and citations. For organic answer visibility, explore [managed AEO and GEO services](/digital-marketing/ai-visibility-aeo) or the [AI visibility checker](/features/aeo-scan). ## Before an advertising test starts 1. Check current account eligibility and the advertiser requirements. 2. Agree the offer, landing page, available formats and budget. 3. Test the conversion journey and document what can be measured. 4. Launch only within the approved account and agreed scope, then review the evidence before expanding spend. For Brisbane and Gold Coast businesses, location and audience choices should reflect your actual service area and the settings available in the advertising account. We do not promise unsupported targeting controls or guaranteed lead volumes. ## What to bring to a readiness consultation Share your website, priority offer, target customer and any existing advertiser account status. We will define the readiness work, client responsibilities, management scope and fees before implementation. Media spend and third-party costs are separate from advisory or management fees. **Check whether ChatGPT advertising fits your offer.** Start with eligibility, landing-page quality and a measurable objective. If another channel is a better fit, the plan should make that clear. --- # AI Citation Tracking & Source Analysis > Track cited URLs in AI answers, compare your sources with competitors, and turn citation analysis into content and authority actions. Explore CiteAgentic. Source: https://www.citeagentic.com/features/ai-citation-tracking Find which sources support your competitors{' '}with AI citation tracking.} sub="See the pages and domains cited when AI answers buyer questions. Compare your coverage with competitors and identify content and authority gaps to work on." primary={{ label: 'Set up citation tracking', href: '/?audit=1' }} secondary={{ label: 'Compare plans', href: '/pricing' }} visual={} /> Start with your website and buyer questions, then create your account. Engine coverage, monitoring frequency and usage limits depend on your plan. The evidence behind the answer Your brand is onlypart of the story. Your own pages. Independent reviews. Competitor guides. See the sources appearing in your tracked answers, then decide where you can make a useful change. Want to follow changes over time? Use the AEO tracker to monitor your brand, citations and competitors across repeated answers. Explore the full AI visibility picture → What is AI citation tracking? AI citation tracking records the source links returned in answers to your monitored questions. CiteAgentic connects those links to the prompt and response, helping you distinguish your website from competitor and third-party sources. A citation is evidence to inspect. It is not automatically an endorsement, a click or a sale. Citation monitoring and analysis Know what the source tells you. Move from a cited domain to the specific pages and buyer questions behind it. Prioritise sources that appear across the questions your business cares about. Cited URLs and domainsInspect the pages captured as sources and how often they appear in the monitored answers. Open the source to check its current content before deciding what to change. Questions and enginesReview the buyer questions and engines associated with a source. Read the corresponding answer to understand why the link is relevant and which brand it discusses. Your site and other sourcesSeparate your own pages, competitor sites and third-party publications. A competitor page may reveal missing coverage; an independent source may point to a relevant profile, review or editorial opportunity. No source links in an answer? An answer can mention your brand without linking to it. A completed answer with no captured citations means no source links were available in that observation. A failed or unavailable engine response is missing evidence, not proof that your brand was absent. Review the answer and scan status, then compare later runs of the same questions, engines and market. Citation counts cover captured answers, not every AI conversation or every click to your site. How to measure citation changes → A closer look Clear answers.No inflated promises. We measure your tracked prompts and runs. AI answers can change, and some return no source links. Use repeated observations to guide your decisions. ### Frequently Asked Questions **Q: Can I track citations in ChatGPT answers?** Yes. CiteAgentic records source links returned in monitored ChatGPT answers when they are available. You can review them with the corresponding response. This measures your tracked prompts and runs; it does not expose private user conversations or every answer ChatGPT gives to other people. **Q: How does citation monitoring help me decide what to improve?** Start with sources appearing for buyer questions that matter to your business. Inspect the cited URL and answer context. If your own page is outdated, improve it; if a competitor answers a missing question, prepare stronger coverage; if an independent source is relevant, investigate an accurate listing or contribution. Review later scans using the same prompts, engines and market to see what changed. **Q: How is an AI citation tracker different from a backlink checker?** An AI citation tracker records links included in AI answers. A backlink checker records links from one website to another. They describe different relationships: a page can have backlinks without appearing in your monitored AI answers, and an AI answer can cite a page without creating a conventional website backlink. **Q: Which AI engines can I monitor?** CiteAgentic supports monitoring across multiple AI engines, including ChatGPT and Perplexity, with access determined by your plan. Google AI Overviews is treated as a separate search signal. Check the pricing page for current coverage and limits; individual answers may return different amounts of citation evidence. **Q: Does a citation mean the AI recommends my brand?** No. A citation identifies a source link, while a recommendation depends on what the answer actually says. An answer may cite a comparison, a competitor or a page that is critical of a product. Read the response and its supporting context before treating the citation as an endorsement. **Q: Will improving a page guarantee more AI citations?** No. Improving clarity, accuracy and supporting evidence can make a page more useful, but AI engines choose their own sources. Use citation findings to guide changes, then retest the relevant prompts. CiteAgentic helps you observe the results; it does not guarantee inclusion or attribute every change to a single fix. From the answer to your next action Find the sources shapingyour AI visibility. Set up your website and buyer questions. Inspect the evidence before choosing your next change. Set up citation tracking →Compare plans Need help implementing the findings? Explore our GEO and AEO services. THE BIGGER PICTUREAI visibility tool → IMPROVE YOUR PAGESAEO page audits → UNDERSTAND THE DATAMeasuring AI citations → --- # How to get cited by AI search engines > Learn how AI search systems discover pages, select sources and cite businesses - and how to diagnose access, content, entity and authority gaps. Source: https://www.citeagentic.com/resources/getting-cited-by-ai _Getting cited by AI · Complete guide_ ## How to get cited by AI search engines. A practical guide to crawler access, source selection, extractable content, entity clarity, independent proof and measuring whether a fix changed the answer. **TL;DR:** To become eligible for AI citations, first make the page public, indexable and accessible to the search crawler used by each platform. Then publish clear textual answers to real buyer questions, support important claims with verifiable evidence, keep business information consistent across your site and independent sources, and measure the same prompts before and after each change. No schema type, crawler directive or writing formula guarantees a citation. ## What does it mean to be cited by an AI search engine? Being cited means one of your URLs appears as a supporting source in a generated answer. It is one of three distinct outcomes worth tracking, and they move independently: **mentioned** (your brand name appears in the answer), **recommended** (the answer presents you as a choice for the buyer's need), and **cited** (your page is linked as a source). An answer can mention you without citing you, cite you without recommending you, or recommend you based entirely on what third-party sources say. The full path from invisible to recommended runs through stages: access, discovery, source selection, extractable content, entity clarity, independent proof, and measurement. Each stage has its own failure modes and its own fixes, which is why "write more content" is bad advice; the guide and its articles below walk the stages in order. ## How do AI search engines find and select sources? Every platform's ranking internals are proprietary, but public documentation agrees on the shape: crawl or retrieve candidate pages, expand the user's question into related searches where applicable, select pages that support parts of the answer, and generate a response with links. Google [documents](https://developers.google.com/search/docs/appearance/ai-features) that AI Overviews and AI Mode draw on normally indexed, snippet-eligible pages; OpenAI's [publisher guidance](https://help.openai.com/en/articles/12627856-publishers-and-developers-faq) describes OAI-SearchBot surfacing and linking sites in ChatGPT search. The expansion step matters most for strategy: your page competes at the level of the specific sub-questions an engine generates, not just the headline question. The mechanics, including query fan-out and why different sources appear for the same question, are in [how AI search engines choose sources to cite](/resources/getting-cited-by-ai/how-ai-chooses-sources). ## What prevents a website from being eligible for citation? Access failures, and they are silent. The page is blocked in robots.txt for the platform's search crawler, a CDN or WAF challenges the crawler, the content only exists after JavaScript runs, or the page simply is not indexed. A page the crawler cannot fetch is invisible regardless of its quality. The crawler names trip people up: for ChatGPT search inclusion the relevant crawler is **OAI-SearchBot**, while GPTBot concerns possible training use; for Google's AI features the control is ordinary **Googlebot**, not Google-Extended; Perplexity recommends allowing **PerplexityBot**. The full crawler map, safe robots.txt configurations and a three-layer verification method are in [which AI crawlers should your website allow](/resources/getting-cited-by-ai/crawler-access). ## What makes a page useful as a supporting source? A page an engine can safely quote: the direct answer to one question stated in the first paragraph, question-led sections in the buyer's own phrasing, claims sourced beside the sentence that makes them, and honest qualification after the answer rather than instead of it. The page formula is `direct answer → evidence → example → qualification → next step`. Structured data plays a supporting role only: Google states no special schema is required for its AI features, so markup should describe the visible page accurately, never masquerade as a citation lever. The full blueprint, including when to use lists and tables and how to verify a page after publishing, is in [how to structure a page so AI can cite it accurately](/resources/getting-cited-by-ai/make-content-citable). ## How do AI engines identify the business behind a page? By reconciling your site's account of yourself with independent records: your About page, Organization or LocalBusiness schema with `sameAs` links, your Google Business Profile, directories and review platforms. When the name, category and facts agree everywhere, the engine can name you confidently; when they conflict, the safe move is to omit you or confuse you with a similarly named business. Entity ambiguity is a common silent killer for otherwise strong sites, and doubly so for local businesses sharing a name across cities. Diagnosis is direct: ask each engine "what is [your business]?" and read the answer literally. The identity playbook is in [how AI engines identify your business](/resources/getting-cited-by-ai/entity-signals). ## Why does independent evidence affect recommendations? Because recommendation questions are trust questions, and self-description is weak evidence. Engines retrieve reviews, directories, community threads and editorial comparisons alongside brand websites, and a business described consistently by sources it does not control is one an engine can safely recommend. When those sources describe your competitors and not you, the answer recommends your competitors. The work is specific, not generic: find which independent sources engines actually cite for your questions, find where competitors appear without you, and earn a legitimate presence there. That playbook, including what changes by business type and how to participate without astroturfing, is in [why reviews, directories and community sources influence AI recommendations](/resources/getting-cited-by-ai/third-party-proof). Why this citation graph differs from the backlink graph, and what each one can tell you, is in [citation graphs vs backlink graphs](/resources/citation-graph-vs-backlinks). ## How should you measure whether a fix worked? Baseline first, then one change at a time, then re-test the same questions on the same engines and regions across repeated scans. AI answers vary run to run, so a single answer is never evidence; judge trends, and track mentioned, recommended and cited as separate outcomes because a fix can move one without the others. There is no honest universal waiting period: recrawl and answer changes vary by engine, site and page. The test design that survives this variance, and what to do when the answer does not change, is in [how to measure whether an AI visibility fix worked](/resources/getting-cited-by-ai/measure-ai-citations). ## What should you fix first? In the order the pipeline fails, because later stages are unmeasurable until earlier ones pass: 1. **Access**: confirm the search crawlers can fetch and index the pages that matter. 2. **Content**: make the pages that map to your buyer questions answer them directly and quotably. 3. **Entity**: make every record that describes your business agree. 4. **Independent proof**: earn presence on the specific sources engines cite for your questions. 5. **Measurement**: baseline, change, re-test, attribute; then repeat on the next lost question. If you have Google traffic but no AI presence, the diagnostic in [traffic but no AI citations](/resources/traffic-but-no-ai-citations) identifies which failure class applies before you spend on the wrong stage. The prioritised implementation checklist across all stages is [how to get mentioned by AI](/resources/getting-cited-by-ai/get-mentioned-by-ai). ## How CiteAgentic handles the workflow **Product example.** CiteAgentic runs this loop end to end: it tracks your buyer questions across engines and regions and keeps question-level evidence; audits pages and technical access; ties every recommendation to the specific questions and cited sources behind it; prepares drafts with marked placeholders for facts only you know, for human review and approval, never auto-publishing; maps buyer segments, topics, questions, pages, competitors and off-site sources in the Authority Map; and re-tests the affected questions after each shipped change, reporting page state, validation state and AI outcome separately. If you are comparing tooling for this job against SERP content editors, see [Surfer SEO alternatives for AI search](/resources/getting-cited-by-ai/surfer-seo-alternatives). **See where the pipeline drops you** A scan runs your real buyer questions across the major engines, shows who is named instead of you, and traces each loss to the stage that caused it. Start free trial → ## Getting cited by AI: complete checklist - [ ] robots.txt allows each platform's **search** crawler (OAI-SearchBot, Googlebot, PerplexityBot); training-crawler decisions made separately and deliberately - [ ] CDN/WAF serves crawlers `200` with real HTML; no challenges on key pages - [ ] Key pages indexed and snippet-eligible (verify in Search Console) - [ ] One page per important buyer question, direct answer in the first paragraph - [ ] Question-led headings in buyer phrasing; claims sourced beside the sentence - [ ] Structured data accurately describes each page and your organization - [ ] One public-facing business name, consistent across site, schema, profiles and directories - [ ] About page states what you are, your category, who you serve, where you operate - [ ] Business Profile and directory records claimed, accurate and consistent - [ ] Cited-source lists collected for your questions; competitor-only sources identified - [ ] Legitimate presence earned on recurring cited sources; all participation disclosed - [ ] Baseline recorded per question, per engine, per region, across multiple runs - [ ] One change shipped at a time, dated, and re-tested against the baseline ### Frequently Asked Questions **Q: Do AI engines only use training data, or do they read the live web?** Both. Base models carry knowledge from training, and the search products retrieve current pages at answer time through their search crawlers and indexes. The retrieval layer is where site changes can affect answers, which is why access and indexability come first. **Q: How long does it take for a change to show up in AI answers?** It varies by engine, site and page, and no honest fixed timeline exists. Recrawl speed, index refresh and answer generation all differ per platform. Baseline the questions, ship the change, and re-test on a schedule rather than waiting for a promised day. **Q: Why does a competitor get recommended instead of us?** Usually because the sources the engine retrieves describe them and not you: reviews, directories, comparison articles and community threads. Reading the cited sources for your lost questions shows exactly which independent evidence is deciding against you. **Q: Does schema markup affect AI answers?** Google states no special schema is required for its AI features. Accurate markup helps machines classify your pages and resolve your business entity, which removes reasons to be passed over; it is honest description, not a citation switch. --- # Otterly Alternatives | Compare AI Visibility Workflows > Compare Otterly alternatives by evidence, website fix workflow, coverage and cost. Dated sources and practical evaluation steps. Source: https://www.citeagentic.com/compare/citeagentic-vs-otterly # Comparing Otterly alternatives? Start with the work you need to get done. Otterly offers daily tracking, prompt research, citation analysis and GEO audits. It also advertises content briefs and recommendations. Compare those workflows with CiteAgentic’s audit findings, fix prompts and follow-up scans, using the same business questions. Vendor-authored comparison. Published capabilities reviewed 28 September 2026; not a hands-on benchmark. ## The short answer Consider CiteAgentic when you want AI visibility evidence alongside SEO and AEO findings you can act on. Keep Otterly on your shortlist if its reporting and optimisation workflow already fits your team. This is a vendor-authored comparison of published capabilities, not an independent product benchmark. ## At a glance | Decision | CiteAgentic | Otterly | | --- | --- | --- | | AI answers and citations | Allowed engines depend on plan | Check selected plan and models | | Suggested next actions | Grounded in scan findings | Vendor lists recommendations | | Website fix workflow | SEO/AEO audit findings and fix prompts | Evaluate depth using a real page in your trial | | After-change measurement | Follow-up scans compare observed visibility | Test the reporting workflow you need | | Price and limits | See current pricing; tier limits apply | Check billing term and engine add-ons | ## What should you compare beyond a visibility score? Use the same questions, language, market and observation window. Check whether an answer mentions your brand, cites a URL, or recommends a competitor. These are different signals. A sampled answer is not a count of all customer searches. Compare the evidence behind the score. ## What work can your team actually complete? In CiteAgentic, review an audit finding, its affected page and the suggested fix prompt. Check the proposed change before applying it. Run that same exercise in the other tool and judge the specificity, effort and relevance of the guidance. Choose the workflow that helps your team ship a useful fix. ## Does a better score prove a fix worked? A follow-up scan can show an observed change. It cannot isolate your edit from model changes, changing sources or answer variation. Keep the prompt set stable and review several observations alongside qualified visits and enquiries. Treat before-and-after results as evidence, not causal proof. ## Pricing and limits Published pages checked 28 September 2026. Follow the pricing links for current terms; this is not a checkout quote. Check included scans, quotas and fix-prompt access before choosing a plan. Lite: $29/month: 15 prompts, monthly billing; Standard: $189/month: 100 prompts, monthly billing; Premium: $489/month: 400 prompts, monthly billing. Claude, Gemini and Google AI Mode are listed as add-ons. Confirm the complete cost for your required coverage. ## Before switching ### How should I compare Otterly alternatives? Define the questions, markets and engines you need first. Test equivalent coverage, review cited URLs and complete one website fix. Compare the time and cost of that workflow, not only the headline score. ### Will switching improve my AI visibility? Changing software alone does not improve visibility. Results depend on the work you implement and how AI systems retrieve and answer. No tool can guarantee mentions or citations. ### Can I move my historical reports? Keep your existing exports and baseline. Recreate a representative prompt set and run both tools alongside each other. Do not assume historical scores transfer or use the same methodology. ## Sources and next steps - [Otterly capabilities](https://otterly.ai/) - [Otterly pricing and limits](https://otterly.ai/pricing) - [CiteAgentic pricing](/pricing) - [Compare AI SEO tools](/compare) - [Review CiteAgentic audit features](/features/seo-scan) --- # How to connect Contentstack to CiteAgentic and publish articles automatically > Connect a Contentstack stack to CiteAgentic with a management token, so approved drafts are created as entries and optionally published to an environment. Field mappings, regions, and error decoding. Source: https://www.citeagentic.com/docs/integrations/contentstack _Publishing channels · Contentstack_ ## Publish to your Contentstack stack Connect a Contentstack stack and CiteAgentic creates each approved article as an entry in the content type you choose, then publishes it to your environment if you want it live straight away. **TL;DR:** **TL;DR:** Connect a Contentstack stack with a management token and CiteAgentic creates each approved draft as an entry in your content type, optionally publishing it to an environment. You need the stack API key, a management token, the content type UID, and your region. Contentstack answers `412` rather than `401` when credentials are wrong, which is the error to look for. Connecting Contentstack takes about five minutes and needs three values: your stack API key, the content type UID for new entries, and a management token. Add an environment name if you want entries published rather than left as drafts. ## What is the Contentstack integration? The Contentstack integration creates each approved CiteAgentic draft as an entry in your stack using the Content Management API, and can then publish that entry to an environment you nominate. Creating an entry always produces a draft. Publishing is a separate step in Contentstack's model, which is why the integration asks for an environment separately: without one there is nowhere for the entry to go live. ## How is this different from Contentful? Contentstack and Contentful are two different companies with confusingly similar names. They share no APIs, no tokens, and no concepts. CiteAgentic supports both as separate publishing channels, so make sure you are connecting the one your team actually uses. ## Why does it write markdown instead of rich text? Contentstack's JSON Rich Text Editor stores content as a structured document rather than as text, so writing into it means synthesising that document and guessing at your schema. Contentstack's **Markdown** and **Multi Line Textbox** field types store plain strings, and CiteAgentic already renders every draft to clean markdown. The integration writes markdown into a field you nominate. Point it at a Markdown field and your article arrives fully formatted. If your content type only has a JSON Rich Text body today, add a Markdown field and point the integration at that. ## What do you need before you start? You need four things from your Contentstack account: - **A stack**, and its **API key**. - **A content type UID** for the new entries, e.g. `article`. It must already exist. - **A management token** with write access to that stack. - **An environment name** such as `production`, but only if you want entries published immediately rather than created as drafts. You also need to know which **region** hosts your stack. A token from an EU stack fails against the North America host, which is the most common setup mistake. ## Where do you find your API key, content type UID and management token? All three live in your Contentstack stack settings. **Stack API key:** open the stack and go to **Settings → Stack**. It starts with `blt`. This identifies the stack and is not the secret. **Content type UID:** open **Content Models** and select your content type. Use its UID, which is lowercase with underscores, rather than its display name. **Management token:** go to **Settings → Tokens → Management Tokens** and create one with write access. Contentstack shows it once. ## How do you connect Contentstack in CiteAgentic? Open **Site Settings → Integrations → Publishing channels**, click **+ Contentstack**, and fill in the form. The fields map as follows: | Field | Required | What to enter | | --- | --- | --- | | Stack API key | Yes | The `blt…` key from Settings → Stack | | Region | No | Where the stack is hosted. Defaults to AWS North America | | Content type UID | Yes | e.g. `article` | | Title field UID | No | Defaults to `title` | | Body field UID | No | A Markdown or Multi Line Textbox field. Defaults to `body` | | URL field UID | No | If present, the slug is written as `/slug` | | Publish environment | Only to publish | e.g. `production` | | Locale | No | Defaults to `en-us` | | On publish | No | Create entry only (default), or create and publish | | Management token | Yes | Stored encrypted and never shown again | Click **Test & connect**. CiteAgentic fetches the content type, which validates the API key, the management token, the region and the content type UID in one call. If you chose to publish immediately, it also checks the environment exists, so a missing environment fails at setup rather than at publish time. ## Which regions are supported? Contentstack hosts stacks across three clouds, and CiteAgentic supports all seven Content Management API hosts: | Region | API host | | --- | --- | | AWS North America (default) | `api.contentstack.io` | | AWS Europe | `eu-api.contentstack.com` | | AWS Australia | `au-api.contentstack.com` | | Azure North America | `azure-na-api.contentstack.com` | | Azure Europe | `azure-eu-api.contentstack.com` | | GCP North America | `gcp-na-api.contentstack.com` | | GCP Europe | `gcp-eu-api.contentstack.com` | ## How do you publish a draft to Contentstack? There are two ways, and both use the same connection. **On demand:** open any draft in Content Studio and choose **Share ▾ → Publish to Contentstack**. The entry is created immediately, and published to your environment if that is how the channel is configured. **On a schedule:** link the Contentstack channel to a day in the content calendar. Once you approve that piece, it is created on its scheduled date. Each piece becomes an entry with your title field and markdown body field filled in, plus the URL field if you nominated one. CiteAgentic links back to the entry in the Contentstack editor, because your front end owns the public URL. ## What errors might you see, and what do they mean? Contentstack returns detailed error bodies, which CiteAgentic surfaces alongside a plain-language cause: - **"Authentication failed."** Contentstack returns `412` with `api_key: is not valid` for a bad stack key or management token. Check the **Region** before regenerating anything; a valid EU token looks invalid to the North America host. - **"Content type … not found in this stack."** The UID is wrong. Use the UID from Content Models, not the display name. - **"Environment … not found in this stack."** The environment name does not exist. Check Settings → Environments. - **"Entry created as a draft, but publishing to … failed."** The entry exists and is safe; only the publish step failed. Open it in Contentstack and publish manually, or fix the environment and retry. - **"Contentstack rate-limited the request."** Wait a moment and retry. If a scheduled publish fails, the channel is flagged with the error in Site Settings and the failure is recorded against that draft, so nothing fails silently. ### Frequently Asked Questions **Q: Is Contentstack the same as Contentful?** No. They are different companies with similar names and share no APIs or tokens. CiteAgentic supports both as separate publishing channels. **Q: Are entries published live straight away?** Not by default. Entries are created as drafts so you can review them in Contentstack first. Choose **Q: Does this write to a JSON Rich Text field?** No. JSON Rich Text stores a structured document rather than text, so the integration writes markdown into a Markdown or Multi Line Textbox field you nominate. **Q: What access does the management token need?** Write access to the target stack. Management tokens are stack-level, so one token covers one stack and is used together with that stack\ **Q: Which locale do entries use?** The locale you set, defaulting to en-us. It is used both when creating the entry and when publishing it. **Q: Is my management token stored securely?** Yes. It is encrypted at rest with AES-256-GCM and never shown again after you save it. The stack API key is stored as ordinary config because it identifies the stack rather than authorising writes on its own. **Ready to connect?** Site Settings → Integrations → Publishing channels → + Contentstack. Add your stack API key, content type UID, and a management token. --- # Will AEO replace SEO? > A practical way to combine SEO and AI visibility work without abandoning search traffic or promising unsupported results. Source: https://www.citeagentic.com/resources/will-aeo-replace-seo _Resources · POV_ ## Keep your SEO foundations. Add evidence from AI answers. You do not need to choose between search traffic and AI visibility. Start with the buyer questions and website improvements that matter to your business. **TL;DR:** Treat AEO as an additional measurement and improvement workflow. Maintain your SEO foundations, inspect real AI answers, and prioritise useful changes backed by evidence. Neither a new acronym nor a higher visibility score guarantees enquiries or revenue. ## Why keep investing in SEO? Your website still needs to be reachable, understandable and useful. Google says established SEO practices remain relevant to its AI features, with no separate special optimisation requirement. Review its [AI features guidance](https://developers.google.com/search/docs/appearance/ai-features) before paying for claims about special inclusion tricks. Preserve work that serves buyers: clear service and product pages, accurate information, accessible navigation and reliable conversion paths. Use your own search and conversion data to decide where effort is worthwhile. ## What does AI visibility monitoring add? It lets you inspect how sampled AI answers describe a category, whether they name your brand and which sources they cite. That can reveal questions your current reporting does not answer. For example, an agency can investigate a comparison question where a competitor appears and the client does not. The next step is to read the answer and inspect its sources. That observation alone does not prove the competitor's page caused the outcome or that copying it will change the answer. Keep [mentions and citations distinct](/resources/what-is-aeo-visibility). A source URL is evidence of a reference, not automatic proof of a recommendation or brand endorsement. ## How should a small team divide its effort? Start with one buyer journey rather than a fixed budget split. 1. Choose a product or service, its audience and target market. 2. Review the relevant page's search performance and technical findings. 3. Monitor a representative set of AI buyer questions within your plan's coverage. 4. Pick one supported gap that affects the buyer's understanding or ability to act. 5. Review and implement the change, then measure both the technical result and later observations. For a SaaS company, the useful change might be clearer workflow or integration information. For a local service provider, it might be accurate service-area and enquiry information. Treat these as possible investigations, not diagnoses before checking the site. ## Which claims should you question? Be cautious about universal score thresholds, guaranteed citations, fixed timelines or customer results without supporting evidence. Ask what was measured, over which questions and markets, and whether the comparison used the same conditions. Do not change training preferences just to follow generic crawler advice. OpenAI documents search and training crawlers separately in its [crawler reference](https://developers.openai.com/api/docs/bots). Inspect each relevant platform's controls before changing access. Generated content also needs review. Complete missing business facts, verify claims and check links before publishing. A draft is a starting point for that workflow. ## What should the first trial deliver? Aim to leave with a usable baseline, one understood priority and a reviewed fix or draft. Those are concrete outputs you can assess during a trial even if visibility has not changed. On a follow-up scan, check the same questions, engines and markets. Record missing checks and changing scope. Compare visibility with actual qualified visits and enquiries, without attributing every change to the last edit. **Start with one useful improvement.** Bring a real page and buyer question to CiteAgentic. Review the evidence, prepare a change and decide whether the workflow saves your team time. Check current plan limits and account-specific trial terms. --- # AIclicks Alternatives | Compare AI Visibility Workflows > Compare AIclicks and CiteAgentic by evidence, website fixes, model coverage and cost. Dated sources and practical evaluation steps. Source: https://www.citeagentic.com/compare/citeagentic-vs-aiclicks # Comparing AIclicks alternatives? Test the work behind the dashboard. AIclicks advertises prompt tracking, source analysis, recommendations and AI agents. Compare its workflow with CiteAgentic’s website audit findings, fix prompts and reviewed drafts using the same buyer questions. Vendor-authored comparison. Published capabilities reviewed 1 October 2026; not a hands-on benchmark. ## The short answer Consider CiteAgentic when you want AI visibility evidence alongside SEO and AEO findings you can act on. Keep AIclicks on your shortlist if its reporting and optimisation workflow already fits your team. This is a vendor-authored comparison of published capabilities, not an independent product benchmark. ## At a glance | Decision | CiteAgentic | AIclicks | | --- | --- | --- | | AI answers and citations | Allowed engines depend on plan | Check selected plan and models | | Suggested next actions | Grounded in scan findings | Vendor lists recommendations | | Website fix workflow | SEO/AEO audit findings and fix prompts | Evaluate depth using a real page in your trial | | After-change measurement | Follow-up scans compare observed visibility | Test the reporting workflow you need | | Price and limits | See current pricing; tier limits apply | Check billing term and engine add-ons | ## What should you compare beyond a visibility score? Use the same questions, language, market and observation window. Check whether an answer mentions your brand, cites a URL, or recommends a competitor. These are different signals. A sampled answer is not a count of all customer searches. Compare the evidence behind the score. ## What work can your team actually complete? In CiteAgentic, review an audit finding, its affected page and the suggested fix prompt. Check the proposed change before applying it. Run that same exercise in the other tool and judge the specificity, effort and relevance of the guidance. Choose the workflow that helps your team ship a useful fix. ## Does a better score prove a fix worked? A follow-up scan can show an observed change. It cannot isolate your edit from model changes, changing sources or answer variation. Keep the prompt set stable and review several observations alongside qualified visits and enquiries. Treat before-and-after results as evidence, not causal proof. ## Pricing and limits Published pages checked 1 October 2026. Confirm billing term, currency, model selection and quotas at checkout. Use the CiteAgentic pricing page for current monthly and annual charges. Trial eligibility is checked for your account; a trial does not guarantee visibility gains. - Starter: $59/month, 30 prompts; 3 selected models. - Pro: $189/month, 150 prompts; 4 selected models. - Business: $499/month, 300 prompts; 6 selected models. These are the published monthly plan cards. Confirm required model selection, usage limits and current checkout terms. ## Before switching ### How should I compare AIclicks alternatives? Define the questions, markets and engines you need first. Test equivalent coverage, review cited URLs and complete one website fix. Compare the time and cost of that workflow, not only the headline score. ### Will switching improve my AI visibility? Changing software alone does not improve visibility. Results depend on the work you implement and how AI systems retrieve and answer. No tool can guarantee mentions or citations. ### Can I move my historical reports? Keep your existing exports and baseline. Recreate a representative prompt set and run both tools alongside each other. Do not assume historical scores transfer or use the same methodology. ## Sources and next steps - [AIclicks capabilities](https://aiclicks.io/) - [AIclicks pricing and limits](https://aiclicks.io/pricing) - [CiteAgentic pricing](/pricing) --- # Shipping AEO fixes for 12+ clients at scale > How an SEO agency added AEO monitoring and paste-ready fix prompts to their client delivery stack. Source: https://www.citeagentic.com/case-studies/agency-client-scale _Agency_ ## Shipping AEO fixes for 12+ clients How an SEO agency added AEO monitoring and paste-ready fix prompts to their client delivery stack. **TL;DR:** A 6-person SEO agency turned AEO monitoring into a new retainer line worth +$2,400/mo per client ($28.8k/yr across 12 clients). Time per client dropped 60%, from 8 hours of advisory to 45 minutes of management per month, and churn fell 34%. ## The challenge A 6-person SEO agency spent 80% of their time on link building and on-page optimization for Google ranking. When clients asked about AI search, the agency had no tooling and no methodology. They tried: - Manually searching ChatGPT, Perplexity, Google AI Overviews for each client's brand - Building one-off reports in Docs - Handing off advisory without actionable next steps — clients were left to hire developers or give up This model didn't scale beyond 3-4 clients and left money on the table (no retainer for AI work). ## The pivot They adopted CiteAgentic and rebuilt their service: 1. **Monthly AEO audits** — Automated scans for each client, delivered as a dashboard the agency can share directly. 2. **Fix-prompt workflows** — Instead of advisory-only reports, each finding ships with a paste-ready prompt. Clients hand it to their developers or Claude, shipped in hours. 3. **Competitive positioning** — They added "AI Readiness Score" to their monthly brand reports alongside the Domain Authority / Traffic metrics. Clients saw it as a new service line, not a replacement. ## The setup - **First month:** 6 hours per client onboarding (URL → audience → competitor set → monitoring schedule) - **Ongoing:** 45 minutes per client per month (reading the audit, customizing fix prompts for their tech stack, writing monthly recap) By month 3, the system ran on autopilot with minimal friction. ## The business impact | Metric | Impact | |---|---| | New retainer line | +$2,400/mo per client ($28.8k/yr across 12 clients) | | Time spent per client | -60% (from 8h/mo advisory to 45min/mo management) | | Churn rate | Down 34% (clients saw tangible monthly wins, not recommendations) | | Client projects approved | +47% (developers saw the fix prompts, approved budgets faster) | The 12 clients generating $28.8k/yr new ARR now cost 45 minutes per month per client to manage, vs. 8 hours under the old advisory model. ## What changed the game The game changed because of three shifts: 1. **Paste-ready prompts shifted the burden.** Instead of the agency explaining "go add schema markup," the prompt showed the exact structure. Clients' developers said yes faster. 2. **Automated monthly delivery.** No 2-week report cycle. The dashboard refreshed automatically, and the agency could point clients to it instead of writing custom reports. 3. **Competitive benchmarking.** Showing a client they're #2 in AI visibility (vs. competitor #1) was more persuasive than a generic score. AEO became a sales-friendly service line. ## Key lesson: AEO is a upsell, not a replacement The agency's instinct was right — they didn't kill SEO. They added AEO as a monitoring + retainer layer. Clients who signed up for monthly AEO monitoring also renewed SEO work at a 94% rate (vs. 67% for SEO-only clients). --- Interested in building AEO into your service delivery? Schedule a 1:1 session → --- # How a property agency captured 4× organic traffic with AEO optimization > A Sydney property agency transformed their organic visibility by fixing AI readiness gaps. SEO health improved 340%, organic traffic grew 4× in 90 days. Source: https://www.citeagentic.com/case-studies/property-aeo-growth _Real Estate / Property_ ## 4× organic traffic growth in 90 days How a Sydney property agency captured AI-driven buyer intent by fixing AEO readiness gaps while maintaining SEO strength. **TL;DR:** A Sydney property agency grew monthly organic traffic 4× in 90 days, from 8,200 to 32,800 sessions, after 62 hours of AEO restructuring. AEO Readiness Score climbed from 34/100 to 87/100, buyer inquiries rose from 34 to 147 per month, and estimated revenue jumped from $1.2M to $5.1M. ## The challenge AC Property, a boutique real estate agency in Sydney, was losing deal flow to larger franchises and online platforms. They ranked for 47 local keywords (suburbs, property types) in Google — top-3 for 22 of them — but their traffic had plateaued for 12 months. When prospects used AI search ("best suburbs for young families in Sydney," "property investment hotspots NSW"), AC Property never appeared, despite having deep local expertise and 500+ sold properties. The issue wasn't search visibility. It was **AI discoverability**. Their blog had valuable market analysis, but AI engines didn't know to cite it. Their sold-property showcases had no structured data. Their buyer guides existed in PDF form, not web pages. ## The diagnosis We ran a full AEO audit and uncovered: **AI Readiness Score: 34/100** (poor) | Category | Finding | Impact | |---|---|---| | **Schema & Markup** | No LocalBusiness, RealEstateAgent, or Property schema | AI can't parse business info or property details | | **Content Authority** | 140 blog posts, zero internal linking strategy | Blog couldn't build topical authority for suburbs/markets | | **Citation Readiness** | Sold properties archived, not citeable | No proof-points for "properties sold in X suburb" | | **Trust Signals** | Reviews scattered (Google, Facebook, no schema) | AI engines prioritize aggregated ratings | | **FAQ & Comparison** | Generic "how to sell" content vs. suburb-specific guides | Missing high-intent queries like "property value trends Earlwood" | Meanwhile, their SEO health was solid: | Metric | Status | |---|---| | Domain Authority | 58 | | Backlinks | 1,240 (mostly local directories) | | Google Rankings | Top 3 for 22 keywords | | Monthly organic traffic | 8,200 sessions | **The paradox:** Strong SEO, invisible in AI search. Prospects asking AI questions never found them. ## The solution: AEO-first content restructuring We didn't rebuild the site. We restructured it for AI discovery while keeping the SEO foundation intact. ### Phase 1: Schema & Trust (Weeks 1–2, 12 hours) - Added LocalBusiness schema with full contact details, service areas (35 suburbs), open hours - Implemented AggregateRating schema pulling from Google & Facebook reviews (integrated via API) - Added RealEstateAgent schema to the team page with individual agent photos + bios - Marked up sold properties with Property schema (3-year archive: 180 properties) with price, date, suburb, property type **Result:** AI engines could now parse "who AC Property is," "where they operate," "what they've sold." ### Phase 2: Topical Authority (Weeks 2–6, 32 hours) The blog had 140 posts scattered across topics. We reorganized: - **Suburb guides:** Created 35 citeable pages for high-intent suburbs (Earlwood, Leichhardt, Marrickville, etc.). Each included: market stats, average property value trends, demographic data, 3–5 sold-property examples with results. - **Buyer journey content:** Linked existing "first-time buyer," "investor," "downsizing" posts to the suburb guides. AI engines now had context chain: intent → buyer profile → suburb options → proof points. - **Internal linking architecture:** Added 120+ contextual links from general posts to suburb guides. Helped AI trace the argument: "investing in Sydney" → "best suburbs for ROI" → "Earlwood: 12% YoY growth." ### Phase 3: Citation Building (Weeks 6–8, 18 hours) - **FAQ schema:** Added 25 suburb-specific FAQs ("How much does a 3-bed cost in Earlwood?" "Is Marrickville a good investment?") with data-backed answers - **Comparison tables:** Built "top 5 suburbs for first-time buyers," "best investment hotspots," "where young families are moving" — all with property-type breakdowns - **Video embeds:** Recorded 8 short market-update videos (2–3 min each) and embedded them on suburb pages. Perplexity cites video in 12% of real estate answers; adding one more surface per page increased AI citation likelihood ### Phase 4: Continuous Monitoring (Ongoing, 4 hours/month) - Set up monthly AEO scans to track citation trends - Monitor competitor visibility in "best suburbs" / "property investment" queries - Review AI answers and adjust content if AC Property wasn't cited - Update market stats quarterly to stay current **Total first-pass effort: 62 hours over 8 weeks.** Team of 2 (content lead + developer). ## The results: 90-day performance ### AEO Health | Metric | Before | After | Delta | |---|---|---|---| | AEO Readiness Score | 34/100 | 87/100 | +156% | | ChatGPT citations | 3 | 24 | +700% | | Perplexity citations | 1 | 31 | +3,000% | | Google AI Overviews | 0 | 8 | new channel | | Citation rate (% of relevant queries) | 4% | 43% | +10.75× | ### SEO Health (maintained + improved) | Metric | Before | After | Delta | |---|---|---|---| | Google Rankings (top 3) | 22 | 38 | +73% | | Domain Authority | 58 | 61 | +5% | | Backlink growth | 0/mo | +18/mo | new inbound links from AI visibility | ### Traffic & Business Results | Metric | Before | After | Delta | |---|---|---|---| | Monthly organic traffic | 8,200 | 32,800 | +300% | | AI-referred traffic | ~200 | 8,400 | new channel | | Buyer inquiries (organic) | 34 | 147 | +332% | | Conversion rate | 0.41% | 0.45% | +10% (minor, but higher intent) | | Properties sold (3-month) | 12 | 51 | +325% | | Revenue impact (est.) | $1.2M | $5.1M | +325% | **The 4× traffic growth broke down as:** - 1.3× from improved Google rankings (schema + linking helped CTR + ranking signal) - 2.7× from new AI channels (ChatGPT + Perplexity referral traffic) ## Why the strategy worked The strategy worked because of five compounding factors: ### 1. **Schema unlocked AI parsing** Before, AI engines saw AC Property as a generic site. After schema, they understood "this is a licensed real estate agent in Sydney with 180 sold properties and 4.8-star reviews." That's citable. ### 2. **Suburb guides became reference material** Instead of blog posts, suburb guides were structured like Wikipedia entries: data-first, searchable, citable. "What's the best suburb for first-time buyers?" now had a direct answer sheet — and AC Property was in it. ### 3. **Internal linking built topical depth** AI engines reward breadth + depth. The blog was broad but flat. Once linked to suburb guides, it showed: broad intent understanding + specific suburb expertise. That's authority. ### 4. **Content was already written** They didn't commission new articles. They reorganized existing content (140 blog posts) into a structured knowledge graph. Efficiency: 62 hours for a 4× result. ### 5. **Proof points mattered** Sold-property examples with prices + dates proved "AC Property actually sells in these suburbs." That's the differentiator between a real estate agent and a blog. ## The unexpected wins ### Buyer inquiry quality improved Traffic increased 4×, inquiries increased 3.3×. Why the delta? **AI-referred buyers were more qualified.** They'd read AC Property's suburb guide or seen sold examples in AI answers before clicking through. Conversion intent was higher. This is why the revenue multiple (5.1M vs 1.2M) exceeded the traffic multiple (4×). ### Competitor visibility surfaced content gaps Monthly monitoring showed competitors appearing in queries AC Property wasn't cited for (e.g., "investment property tax strategies"). Content gap, not SEO gap. They hired a tax accountant to co-author one guest post, shipped it with schema, and closed that gap. ### Google benefited from AI optimization Because schema helped AI engines understand AC Property better, Google's crawlers also had better context. Google rankings improved 73% — mostly on suburb + property-type keywords. The SEO lift was a side effect of AEO work. ## Lessons for property agents & local businesses 1. **Your blog is only as valuable as it's citeable.** 140 posts, zero citations = 140 missed opportunities. Structure matters. 2. **Schema isn't just for Google.** AI engines rely on it heavily. A LocalBusiness + AggregateRating combo is the fastest ROI. 3. **Proof points sell.** Sold properties, testimonials, case studies — archive them with dates and prices. AI engines cite specificity. 4. **"Best for X" queries are AI-favorite queries.** "Best suburbs for [persona]" is high-intent for both humans and AI. Own these. 5. **AEO and SEO compound.** Optimizing for AI didn't cannibalize Google traffic. It amplified it. The same schema + linking structure helped both. --- ## About the client **AC Property** is a boutique real estate agency in Sydney, NSW, specializing in residential sales and investment properties across inner-west suburbs. They've sold 500+ properties in the past 5 years and maintain a 4.8-star Google rating across 180+ reviews. Before AEO optimization, they were dependent on Google local pack visibility and offline referral networks. Post-optimization, they're a top-3 citation source in AI answers for Sydney property queries. --- Ready to unlock AI-driven growth for your business? Run your AEO audit → --- # What is AEO visibility? A practical definition for founders. > Understand AI brand mentions, citations and sampled visibility rates, then use the evidence to choose a useful next action. Source: https://www.citeagentic.com/resources/what-is-aeo-visibility _Resources · foundations_ ## What is AEO visibility, actually? AEO visibility describes whether your brand appears in the AI answers you monitor. Read the answers and sample scope alongside the percentage. **TL;DR:** Brand visibility is the share of successfully captured answers that mention your brand within a defined prompt, engine and market sample. Mentions, recommendations and URL citations are different signals. A score describes those observations; it is not market share, search demand or a guarantee of future answers. ## What is the AEO visibility formula? For a mention-rate view, divide captured answers mentioning the brand by all successfully captured answers in the selected scope, then multiply by 100. Keep failed and missing checks visible separately. For example, if 4 of 10 captured answers mention a brand, its mention rate in that sample is 40%. This is an arithmetic illustration, not a customer result. If another five checks failed, the 40% describes only the ten captured answers; the missing coverage limits what you can conclude. Each question can produce a separate observation for each engine and market. Check the selected date range, prompt set, location and engine filters before comparing figures. See [how scoring works](/resources/how-scoring-works) for the product's score context. ## How are mentions, citations and recommendations different? - **Mention:** The answer names the brand. Read the context; a mention can be positive, neutral or negative. - **Citation:** The answer supplies a source reference. A citation to your page and a citation to a third-party page are different evidence. - **Recommendation:** The answer presents the brand as an option for the buyer's need. Being named in an unrelated passage does not establish a recommendation. A brand may be mentioned without an owned-page citation. A page may be cited without the brand being recommended. Keep these distinctions when deciding what to improve. ## What should you investigate when visibility is low? Start with the actual answers and website findings, rather than assuming one score identifies the cause. **Buyer-question fit.** Do the questions describe your services, audience and markets? Keep a representative mix rather than choosing only easy branded questions. **Page clarity and evidence.** Can a buyer find a direct answer, relevant service detail and support for important claims? Review generated copy before publishing and avoid invented facts. **Search access.** Check that pages intended for discovery are reachable and eligible for indexing. Search and training controls can differ. Use [OpenAI's crawler reference](https://developers.openai.com/api/docs/bots) and [Google's AI search guidance](https://developers.google.com/search/docs/appearance/ai-features) when reviewing the relevant platform. **Competitor and source evidence.** Inspect which competitors were named and the sources returned with those answers. Open source pages before claiming that they mention or endorse a business. These are investigation paths, not a complete explanation of an engine's selection process. ## How does AEO relate to SEO and GEO? SEO focuses on discovery through search results. AEO focuses on answering buyer questions and observing appearances in answer-oriented experiences. GEO is commonly used for work on generative search visibility. The terms overlap, so evaluate the actual workflow and measurements a provider offers. Google says existing SEO best practices remain relevant to its AI features and that no special optimisation is required for inclusion. Its guidance does not guarantee inclusion. [Read Google's requirements](https://developers.google.com/search/docs/appearance/ai-features). Keep your search traffic and conversion reporting. Add AI answer observations as another source of evidence rather than assuming one channel replaces another. ## What does good AEO visibility look like? There is no universal percentage that establishes category leadership. A small branded sample, a broad unbranded sample and a local service sample answer different questions. Do not compare their percentages as if the underlying coverage were identical. A useful baseline records the scope, captured answers, missing checks and important competitors. Compare later observations under the same conditions and investigate individual gains or losses. Track relevant visits, enquiries and customers separately; visibility alone does not prove revenue impact. ## What should you do next? 1. Select representative buyer questions and markets, then capture a baseline within your plan's coverage. 2. Review a commercially relevant gap and its supporting answer or audit evidence. 3. Prepare a fix prompt or draft, confirm the business facts and implement the reviewed change. 4. Verify the issue on a new audit and compare subsequent visibility observations. A follow-up scan may show a change. It cannot establish that your edit alone caused it or promise when improvement will occur. **Find your next evidence-backed action.** Review AI answers, audit findings and draft content in CiteAgentic. Engine access, credits and features depend on your plan. Current pricing and account-specific trial terms apply. --- # SEO Brisbane & Gold Coast | AI SEO Services > SEO Brisbane and Gold Coast services focused on qualified enquiries. Get technical fixes, buyer-led content and AI search optimisation with a scoped plan. Source: https://www.citeagentic.com/services/digital-marketing/aiseo Google, Maps and AI search Turn local searches into better enquiries with SEO Brisbane businesses can put to work. For Brisbane and Gold Coast businesses, we connect technical SEO, local search and buyer-led content with AI search optimisation. Find the searches worth targeting, improve the pages buyers land on, and give Google and AI engines clearer evidence of what you offer. Start with a scoped plan focused on qualified enquiries. Book a strategy call Compare packages Visibility map How buyers find and choose SEO + AEO {[ ['Intent', 'The questions buyers ask before they enquire, compare, or switch'], ['Search', 'Google and Maps: priority pages, organic traffic, reviews, local proof'], ['AI answers', 'ChatGPT, Perplexity, Gemini, Google AI Overviews'], ['Proof', 'Cited sources, accurate schema and verifiable claims to investigate'], ].map(([title, text], index) => ( {String(index + 1).padStart(2, '0')} {title} {text} ))} {` @keyframes seoTrustedMarquee { from { transform: translate3d(0, 0, 0); } to { transform: translate3d(-33.333%, 0, 0); } } .seo-trusted-track { animation: seoTrustedMarquee 26s linear infinite; will-change: transform; } .seo-trusted-rail:hover .seo-trusted-track { animation-play-state: paused; } @media (prefers-reduced-motion: reduce) { .seo-trusted-track { animation: none; transform: none; flex-wrap: wrap; } } `} Trusted by {[ 'Event Rentals', 'Queensland Hire', 'AC Property', 'Little Rascal', 'The Payroll Studio', 'Auric Astro', 'Robert Wade Builder', 'Event Rentals', 'Queensland Hire', 'AC Property', 'Little Rascal', 'The Payroll Studio', 'Auric Astro', 'Robert Wade Builder', 'Event Rentals', 'Queensland Hire', 'AC Property', 'Little Rascal', 'The Payroll Studio', 'Auric Astro', 'Robert Wade Builder', ].map((name, index) => ( = 7 ? 'true' : undefined} className="w-[178px] min-h-[44px] flex shrink-0 flex-col justify-center text-ink grayscale"> 18 ? 16 : 18 }}> {name} ))} ### SEO for Brisbane and Gold Coast buyers Your service area shapes the research, page priorities and local evidence. We work from the locations you actually serve. **Brisbane: match the search to the service** SEO Brisbane campaigns start with your services, buyer questions and existing Search Console data. We map relevant searches to useful pages, review eligible Google Business Profile information and prioritise technical or content gaps that interrupt the enquiry journey. **Gold Coast: make local relevance specific** For SEO Gold Coast work, we review local competitors, service-area information and the proof available for each offer. We improve existing pages first and add location pages only when there is a distinct service need and useful content to support them. Looking specifically for help with ChatGPT mentions and cited sources? Our answer engine optimization services focus on managed AI visibility work. Prefer to investigate first? Start with the free SEO audit. What is included The SEO work we run. The work combines traditional SEO foundations with AEO, competitor analysis, prompt tracking, and content creation built around the questions buyers ask in Google and AI search. {[ ['SEO + AEO discovery audit', 'We review crawlability, indexation, rankings, Search Console, technical SEO, Google Maps SEO, AI visibility, citation readiness, and the pages most likely to affect revenue.'], ['Competitor and share-of-answer analysis', 'We look at which competitors rank, which competitors are recommended by ChatGPT, Perplexity, Gemini, and Google AI Overviews, and which cited sources or content gaps warrant investigation.'], ['Prompt tracking and buyer-intent mapping', 'We map the prompts and keywords buyers use before they enquire: service searches, comparison questions, local intent, objections, and decision-stage AI questions.'], ['Technical SEO and AEO foundations', 'Schema, redirects, canonicals, sitemaps, robots.txt, page speed, internal links, answer blocks, entity clarity, and page structure that search engines and AI systems can parse.'], ['Content to answer the prompts', 'We create or improve service pages, comparison pages, FAQs, guides, local pages, and answer-ready sections so your site can satisfy the prompts competitors currently win.'], ['Citation gaps and authority proof', 'We identify missing proof, reviews, local signals, off-site mentions, citations, and trust signals that make AI engines reluctant to cite or recommend the brand.'], ].map(([title, text], index) => ( {String(index + 1).padStart(2, '0')} {title} {text} ))} Package fit Which AI SEO package is right for your business? Most buyers do not know whether they need a light SEO cleanup, a monthly growth rhythm, or a deeper SEO and AEO push. These are the situations we use to recommend the right starting point. {[ ['I need more local enquiries', 'If you are a local service business and need the phone to ring more often, start with Google Maps SEO, local proof, priority service pages, and technical cleanup.'], ['I need to improve Google rankings', 'If the right buyers are already searching and competitors are ahead, Search Growth gives you keyword research, page optimisation, content briefs, and a monthly execution rhythm.'], ['I want to build topical authority', 'If you need to become the trusted answer in your category, Search Growth or AI Search Authority adds deeper SEO content, guides, comparison pages, answer blocks, and internal authority.'], ['I want visibility in AI search', 'AI visibility needs clear entity signals, answer engine optimisation, citeable proof, AI citations, and content AI engines can quote without guessing.'], ['I am expanding into new locations or markets', 'If you are entering new locations, services, or verticals, we usually start with Search Growth or AI Search Authority so the page strategy, local proof, and authority work can move together.'], ['I already have an SEO provider', 'If you want a second opinion, we can run a discovery call and CiteAgentic SEO/AEO audit, then show where the current strategy is strong and where AI search or technical gaps remain.'], ].map(([title, text], index) => ( {index + 1} "{title}" {text} ))} Pricing AI SEO Packages How the engagement works Start with discovery, then move only when the plan is clear. The first job is not to sell you the biggest package. It is to understand current status, challenges, goals, and the fastest path to better search and AI visibility. {[ ['Discovery call', 'We learn your current status, challenges, goals, locations, services, competitors, existing SEO work, and what a useful result would look like.'], ['SEO + AEO audit in CiteAgentic', 'We run the site through our own SEO and AEO platform. We combine technical findings with separate AI visibility scans, then review eligible Google Business Profile information and local evidence where access is available.'], ['Recommendation and approval', 'We show the recommended scope, why that package fits, what we would do first, and what can wait. Work starts only after you approve the plan.'], ['Weekly or monthly strategy calls', 'Once active, we meet weekly or monthly depending on the package and pace. Calls focus on decisions, shipped work, blockers, upcoming priorities, and the results we are trying to create.'], ].map(([title, text], index) => ( {String(index + 1).padStart(2, '0')} {title} {text} ))} What strong SEO creates Ranked is not enough. You need to be trusted. Good SEO work should turn the website into a clearer, more trusted growth asset across Google, local search, and AI answer engines. {[ ['Qualified visibility', 'Rank for service, location, problem, and comparison searches with buying intent.'], ['Stronger credibility', 'Make reviews, proof, schema, and business detail easy to verify.'], ['AI citation readiness', 'Structure content so answer engines can extract and cite useful claims through AEO services.'], ['Actionable progress', 'Show what moved, why it matters, and what should be shipped next.'], ].map(([title, text]) => ( {title} {text} ))} Articles tagged SEO + AEO Learn how search and AI visibility connect. These SEO and AEO articles support the service offer with education around SEO content, AI citations, and answer engine optimisation. View resources {[ ['/resources/aeo-vs-seo', '/og/resources--aeo-vs-seo.svg', 'SEO + AEO', 'AEO vs SEO', 'How organic rankings, AEO services, and AI citations work together.'], ['/resources/traffic-but-no-ai-citations', '/og/resources--traffic-but-no-ai-citations.svg', 'SEO + AI citations', 'Traffic but no AI citations?', 'Why a site can rank in Google but still fail to become a trusted source in ChatGPT or Perplexity.'], ['/resources/what-is-aeo-visibility', '/og/resources--what-is-aeo-visibility.svg', 'AEO + AI visibility', 'What is AEO visibility?', 'The plain-English version of AI search visibility, Google AI Overviews, prompt wins, citations, and competitor displacement.'], ].map(([href, image, tag, title, text]) => ( {tag} {title} {text} → ))} ### Know what changed and whether enquiries followed Before implementation, agree the priority pages, service areas, conversion events and reporting cadence. The plan identifies each fix, its owner and the evidence behind it. We report shipped work alongside comparable Search Console periods and qualified enquiries where tracking is available. AI mentions and citations are measured separately using agreed prompts and engines. If no reliable baseline exists, we establish one before claiming improvement. Packages describe the type of work, not a guaranteed ranking or citation result. Scope, fees, access and implementation responsibilities are agreed before work begins. Have an existing SEO agency? We can scope AI SEO services around the gaps your current team wants help with. ### SEO questions, answered direct ### Frequently Asked Questions **Q: Do you offer SEO services in Brisbane and the Gold Coast?** Yes. We scope regional SEO around the Brisbane and Gold Coast areas you actually serve, your priority services and the evidence available. The work can combine technical SEO, service-page improvements, eligible business-profile recommendations and AI search optimisation. Location content reflects your actual service area, customer questions and available local evidence. **Q: How much does SEO cost?** It depends on the market, current site quality, number of priority pages, locations, competition, and how quickly you want to move. We scope SEO after looking at the site and competitors, not from a blind package menu. **Q: How long does SEO take?** Some technical and page improvements can help quickly, but meaningful SEO usually compounds over months. Existing sites with decent foundations can move faster than new sites with thin content and weak authority. **Q: How do you measure SEO results?** Rankings and traffic matter, but we also track qualified leads, organic traffic quality, priority pages, Search Console movement, conversion quality, AI mentions, AI citations, and next actions. **Q: Should I invest in SEO or Google Ads?** Use Google Ads when you need demand capture quickly or want to validate search terms and offers. Use SEO when you want compounding visibility and trust. Most serious growth plans use both at different stages. **Q: What is included in an SEO package?** A typical engagement can include discovery, SEO and AEO audit work, technical fixes, keyword research, competitor analysis, page optimisation, SEO content creation, local SEO, authority building, AEO services, and weekly or monthly strategy calls. The mix depends on what will move the result. **Q: What SEO content do you create?** SEO content can include service and location pages, FAQs, answer blocks, guides, comparison pages, buying guides, and supporting articles. The goal is not to publish keywords for the sake of it. The content should answer the buyer question, prove the brand knows the topic, and give AI engines clean material to extract and cite. **Q: How do you know if content is ready for AI search?** We ask a simple question: could an AI engine quote this page without guessing? If the answer is no, the page usually needs clearer claims, stronger evidence, better structure, answer blocks, schema, and proof that supports the recommendation. **Q: Do SEO and AEO work together?** Yes. Many SEO fixes also help AI engines understand the brand: clearer pages, schema, answer blocks, reviews, authority signals, and stronger content. CiteAgentic tracks both search visibility and AI visibility so the work stays connected. **Q: Who is this SEO service best for?** It suits local and service businesses, B2B SaaS teams, professional services, agencies, and multi-location brands. The strategy changes by market, but the pattern is consistent: understand demand, prove expertise, fix the site, and publish the pages buyers need. **Find the SEO work most likely to bring your next enquiry.** We will review your site, SEO services fit, search visibility, AI answer presence, technical debt, competitors, and page gaps, then recommend the right scope. --- # Answer Engine Optimization Services | AEO & GEO > Answer engine optimization services for Brisbane and Gold Coast businesses. Investigate competitor mentions, fix content gaps and measure AI visibility. Source: https://www.citeagentic.com/digital-marketing/ai-visibility-aeo _When AI recommends someone else_ ## Close the gaps behind missed mentions with answer engine optimization services. When ChatGPT names a competitor, you need more than a report. We investigate sampled answers and cited sources, improve the content and evidence buyers need, and track what changes. Managed AEO and GEO services for Brisbane and Gold Coast businesses. ### From a missing mention to a scoped fix Agree the buyer questions, markets, engines and implementation responsibilities before work starts. Each recommendation connects a finding to an action and a way to check progress. **Find where competitors appear** Establish a visibility baseline using agreed buyer questions. Review brand mentions, recommendations and cited sources separately. Record the prompt, engine and date so a missing mention can be investigated in context. **Improve the evidence on your site** Prioritise unclear service pages, unanswered buyer questions, inaccessible content and unsupported claims. Scope practical changes to page copy, answer sections, internal links and accurate structured data with your team. **Check what changed after the work** Repeat comparable prompts and review mention and citation changes alongside the implementation log. Report declines and unchanged results as well as gains. An engine error is unavailable evidence, not proof that your brand disappeared. ### A plan your team can put to work **Prioritised actions with owners** A scoped queue of content, technical and authority actions with supporting observations, expected value and implementation responsibilities. We agree what we will implement and what requires your team or existing provider. **Content grounded in your business** Service-page revisions, FAQs, comparison briefs and evidence pages based on your offer and real buyer questions. Use your customer examples, reviews and service expertise to give buyers clearer reasons to choose you. Where evidence is missing, the plan identifies what to gather. **Citation and competitor analysis** Inspect which URLs AI answers cite and where competitor coverage differs. That evidence helps identify content and authority opportunities; it does not reveal an engine's private ranking logic or guarantee inclusion. **A clear progress review** Review completed work, unresolved dependencies and comparable visibility observations at the agreed cadence. Track qualified enquiries when measurement is available, and keep them separate from AI mentions or website readiness scores. ### AEO, GEO or a broader SEO engagement? Answer engine optimisation (AEO) focuses on helping answer engines understand and reference useful business information. Generative engine optimisation (GEO) describes work on visibility in generated answers. Our generative engine optimization services bring these related needs into one managed engagement, with agreed questions and evidence to measure. If your priority is Google rankings, Maps and local search demand across Brisbane or the Gold Coast, start with our regional SEO and AI SEO services. If you already have an SEO provider, we can scope specialist AEO work around the AI visibility gaps your existing plan does not cover. Want your own team to run the investigation? Use the AI visibility tool, inspect sources with AI citation tracking, or compare software plans. Software subscriptions and managed service scopes are different purchases. ### Know what we can measure, and what we cannot promise ### Frequently Asked Questions **Q: Can you guarantee ChatGPT will recommend my business?** No. AI answers vary by question, engine, date and context. We can investigate sampled answers, improve the evidence within the agreed scope and measure subsequent observations. We cannot guarantee recommendations, citations or a specific uplift. **Q: What is the difference between an AEO audit and an AI visibility scan?** An AEO audit checks website readiness, such as accessibility, structure and clarity. A visibility scan asks selected engines buyer questions and records mentions and citations. A readiness score does not prove that an engine recommends your business. **Q: Do you deliver AEO and GEO services for Brisbane and Gold Coast businesses?** Yes. We scope the work around your genuine services, buyers and locations. The plan connects your service areas and buyer questions with relevant page improvements, local evidence and visibility checks. **Q: What happens if we have no reviews or previous visibility data?** We establish a baseline and record where evidence is missing. Existing service knowledge, accurate business details and genuine project material can inform the first content priorities. The plan identifies which customer evidence to collect next and how to use it. **Q: How much do answer engine optimization services cost?** Fees depend on the priority pages, buyer questions, markets and implementation work. The first conversation establishes goals and access, followed by an agreed scope, responsibilities and reporting cadence before work begins. **Bring the buyer questions you want to be found for.** Tell us your website, priority services, target locations and competitors. We will discuss what to investigate first, what your team can implement and where managed help fits. You can also compare service options. --- # Citation rate climbed 67% in 60 days > How an e-commerce brand went from invisible in AI answers to competing for product queries. Source: https://www.citeagentic.com/case-studies/ecommerce-citation-gains _E-commerce_ ## Citation rate climbed 67% in 60 days How an e-commerce brand went from invisible in AI answers to showing up in 4 of 5 competitor product queries. **TL;DR:** A D2C apparel brand lifted its citation rate 67% in 60 days, taking its "seen in AI" rate from 8% to 47%. Total effort: 24 hours of development work over 4 weeks, driven by product schema, structured reviews, and internal linking. By day 60 the brand appeared in 4 of 5 "best X" queries in its category. ## The situation A D2C apparel brand had strong product pages — they ranked for 80+ product keywords in Google. But when customers asked ChatGPT or Perplexity "best hiking boots under $150," the brand never appeared. Worse, Perplexity consistently cited three competitors (all ranked lower in Google), plus generic review sites with outdated information. The brand had the SEO fundamentals right. AI engines just didn't know they existed. ## The audit CiteAgentic's scan of their top 12 product pages surfaced: - **Missing rich snippets** — Product pages had basic meta tags but no structured data (price, rating, availability, images). Perplexity and ChatGPT rely on Schema.org for product data. - **No review aggregation** — They collected customer reviews in their database but never published them as structured data. AI engines can't cite internal reviews. - **Authority deficit** — Their blog existed (3 posts/month) but never linked to product pages. AI engines prioritize content that builds context, not isolated landing pages. - **Thin comparisons** — Product pages compared themselves to generic alternatives ("vs. leather"), but not to specific competitor products. No reason for an AI to surface them in a competitors' query. All of these are fixable in a weekend. Not overnight, but concentrated work. ## The execution The brand's in-house developer used CiteAgentic's fix prompts: **Week 1:** - Added Product + AggregateRating schema to all 12 product pages (4 hours, using Cursor + the provided prompt) - Set up a review feed that auto-published new reviews as structured data (2 hours) **Week 2:** - Rewrote 4 comparison blog posts to mention specific competitor products by name + price (6 hours) - Added 8 internal links from the blog back to product pages with context (2 hours) **Weeks 3–4:** - Created one "category buyers guide" post (1200 words) comparing 5 product types across 12 specific products (8 hours) - Launched a FAQ schema section on the homepage linking to top product pages (2 hours) Total: **24 hours of development work over 4 weeks.** ## The results Citation tracking started immediately. By week 2 of month 2: | Query type | Before | After | Delta | |---|---|---|---| | Direct product queries ("best X") | 1 mention | 13 | +1,200% | | Competitor-comparison queries | 0 mentions | 4 | baseline | | ChatGPT citations | 4 | 9 | +125% | | Perplexity citations | 0 | 11 | new channel | | Google AI Overviews | 0 | 3 | new channel | | Overall "seen in AI" rate | 8% | 47% | +5.9× | By day 60, the brand was cited in 4 of 5 "best X" queries in their category, and competing directly with brands that ranked higher in Google. ## Why the fix was so fast The fix was fast because of three factors: 1. **Paste-ready prompts** — The developer didn't need to research JSON-LD syntax or Schema.org specs. The prompt showed the exact structure for their products. 2. **Structured data + linking** — Two foundational changes that AI engines prioritize. Not a hack. Not a workaround. 3. **Small batch of pages** — They didn't overhaul 200 pages. They focused the 12 highest-traffic product pages and extrapolated. ## What surprised them Three results are worth calling out as surprises: - **Rich snippets improved Google CTR too.** By adding product schema, Google started showing rating stars and prices directly in search results. Google CTR climbed 18% alongside the AI citations. - **Competitor queries outperformed branded queries.** When someone searched "best hiking boots," the brand appeared. When they searched the brand name, it was already ranked #1 (not an AI-search story). The win was in the competitive space. - **One FAQ post = 6 citations.** The category buyers guide they published in week 4 was cited in 11 different Perplexity answers by day 60. Single articles can move the needle. ## The ongoing strategy The brand now: - Runs monthly AEO audits to catch regressions - Reviews new competitor products appearing in AI answers + adds comparisons to their buyers guides - Monitors citation trends by product category and query type - Ships one deep-category post per quarter It's no longer "rank in Google and hope AI notices you." It's intentional. --- Ready to be seen in AI search? Start your AEO audit → --- # Citation graphs vs backlink graphs - what each one tells you > Backlinks and AI citations answer different questions. What backlink tools measure, what an AI citation graph measures, and when you need each for search and AI visibility. Source: https://www.citeagentic.com/resources/citation-graph-vs-backlinks _Getting cited by AI · Authority_ ## Citation graphs vs backlink graphs: what each one tells you. Backlinks are not obsolete, and AI citations are not backlinks with a new name. They are different graphs answering different questions. Knowing which question you are asking decides which graph you need. **TL;DR:** A backlink graph records who links to whom, and remains relevant to traditional search discovery and authority. An AI citation graph records which URLs AI engines actually cite when answering specific buyer questions: which sources are reused, for which questions, and whether your brand appears on them. Backlinks answer "who vouches for this domain on the link-based web"; the citation graph answers "which pages are deciding the AI answers my buyers read". For AI visibility work, the citation graph is the actionable map, and it is not something backlink tools measure. ## What does a backlink graph measure? Backlink tools like Ahrefs, Semrush and Moz map the link structure of the web: who links to your domain, the estimated authority of those linkers, anchor text, and link gains and losses over time. This graph fed PageRank-style ranking and still informs how traditional search discovers and weighs sites. That remains real. Links help crawlers find pages, and link-based authority still participates in how search engines rank; a business competing for organic Google rankings has legitimate reasons to watch its link profile. Declaring backlinks obsolete is as wrong as pretending they answer AI-visibility questions. ## What does an AI citation graph measure? An AI citation graph is built from answers, not links: for each tracked buyer question, which URLs each engine cited, classified by what they are (your site, competitor sites, review platforms, directories, communities, editorial pages), and observed over time. Its nodes are the pages actually doing the deciding in AI answers. The two graphs can disagree completely, and the disagreement is the insight. A high-authority domain by backlink measures may never appear in the citations for your questions; a modest community thread may be cited constantly. Only one of those facts is visible in a backlink tool. ## Why do the two graphs answer different questions? Because the systems consuming them make different judgments. Link-based authority is a domain-level, question-agnostic signal accumulated over years. AI answer generation selects supporting sources per question, favoring pages that directly support the sentences being generated, and mixing in independent sources like reviews and threads regardless of their link profiles. So the backlink graph answers "how strong is this domain on the link-based web", while the citation graph answers "which specific pages does each engine trust for this specific question". For AI visibility, the second is the operative question; the mechanics of that selection are covered in [how AI engines choose sources](/resources/getting-cited-by-ai/how-ai-chooses-sources). ## What can you do with a citation graph that you cannot do with backlinks? Three concrete moves: 1. **Target real sources, not authority abstractions.** The graph names the exact review platforms, directories, articles and threads cited for your questions, turning "build authority" into "earn presence on these named pages". The legitimate playbook for that is in [third-party proof](/resources/getting-cited-by-ai/third-party-proof). 2. **See competitor evidence per question.** For each lost question, the graph shows which cited sources mention competitors but not you, which is the actual gap deciding the answer. 3. **Track whether off-site work changed outcomes.** Because the graph is anchored to questions, you can re-test the same questions after earning a presence and see whether the cited sources and answers moved; the test design is in [measuring AI visibility fixes](/resources/getting-cited-by-ai/measure-ai-citations). A backlink index cannot do any of these, because it does not observe answers. ## When do you still need a backlink tool? Keep one when your question is a traditional-search question: competing for organic Google rankings in a link-sensitive category, auditing a link profile after a penalty or migration, or reporting link metrics an agency client requires. Those are real jobs, and a backlink index is the right instrument for them. The mistake to avoid is instrument confusion in either direction: using backlink metrics to plan AI-visibility work, or dropping your link hygiene because AI answers are the new frontier. Different graphs, different questions, both can matter to the same business. ## How CiteAgentic builds your citation graph **Product example.** CiteAgentic assembles the citation graph from the scans you already run: every URL cited by every tracked engine for every tracked question, classified by source type, connected to the competitors mentioned alongside, and observed over time. The Authority Map shows the connections (questions, pages, competitors, off-site sources), and source-level views rank where presence is worth earning by citation evidence: how often a source is cited, which questions it decides, and where competitors appear without you. It is prompt-to-source-to-competitor mapping, not a backlink index, and it never ranks work by link volume. **See the graph deciding your answers** Track your buyer questions and see exactly which sources each engine cites, classified and mapped against your competitors. Explore AI Visibility → ### Frequently Asked Questions **Q: Are backlinks a ranking factor for AI answers?** The platforms do not publish citation algorithms, so no one outside them can say precisely how link signals participate. What is observable is that AI answers regularly cite pages with modest link profiles and skip strong ones, so the citation graph you can measure directly is the actionable signal for this work. **Q: Should we cancel our backlink tool?** Only if you have no traditional-search jobs left: no organic ranking targets, no link-profile audits, no clients requiring link reporting. Many teams keep both instruments and use each for its own question. **Q: Is a citation from a low-authority site worth anything?** If engines repeatedly cite that page for your buyer questions, it is by definition one of the pages deciding your answers, whatever its domain metrics say. The citation graph replaces authority guesswork with observed selection. --- # Traffic but no AI citations? The six failure classes to check > Your site gets organic Google traffic but ChatGPT, Perplexity and Gemini never cite you. The verifiable failure classes behind the gap, and how to diagnose each one in order. Source: https://www.citeagentic.com/resources/traffic-but-no-ai-citations _Getting cited by AI · Diagnosis_ ## Your site has traffic. AI engines still ignore you. Here is how to find out why. A site can rank well on Google and be absent from ChatGPT, Perplexity and Gemini answers. The gap always has a diagnosable cause. This is the failure-class checklist, in the order to test it. **TL;DR:** When a site with real Google traffic gets no AI citations, one of six failure classes is usually responsible: the page is not indexed or accessible to the platform's search crawler, the important answer is absent or unclear on the page, the page does not match the question's intent, the business entity is ambiguous or inconsistent, independent sources support competitors instead, or the test itself uses an unstable prompt set. Diagnose in that order, because each class makes the later ones unmeasurable, and fix the first one that fails. ## Why do sites with Google traffic get ignored by AI engines? Because Google rankings and AI answer selection are different judgments over overlapping inputs. Ranking rewards a page's standing for a keyword; an AI answer selects sources that directly support the specific sentences it is generating, mixes in independent sources like reviews and threads, and attributes claims to a business entity it has to resolve confidently. A page can clear the first bar and miss the second for reasons that are invisible in your analytics. The good news: the causes sort into six verifiable classes, and testing them in order finds yours. The full model behind the stages is in the [getting cited by AI pillar](/resources/getting-cited-by-ai). ## 1. Is the page indexed and accessible? Check access before anything else, because a blocked page makes every other diagnosis meaningless. Google traffic proves Googlebot access, and nothing else: ChatGPT search depends on **OAI-SearchBot**, and Perplexity on **PerplexityBot**; either can be blocked by a robots.txt template or a CDN bot rule while Google sails through. Verify the three layers: robots.txt rules per search crawler, what the CDN/WAF actually serves them, index presence per platform. The crawler-by-crawler method is in [which AI crawlers should your website allow](/resources/getting-cited-by-ai/crawler-access). If this class fails, fix it and re-baseline before judging anything downstream. ## 2. Is the important answer actually on the page, clearly? Read the candidate page as an outsider and ask: is the direct answer to the buyer's question stated in text, near the top, in plain sentences a machine could quote? Pages that rank often accumulate content that talks *around* the question: category context, feature narratives, testimonials, with the actual answer implied rather than stated. Also check the rendering: content that only exists after JavaScript runs, or inside images, may be absent from what non-browser readers receive. The blueprint for fixing this class is [how to structure a page so AI can cite it accurately](/resources/getting-cited-by-ai/make-content-citable). ## 3. Does the page match the question's intent? An engine matching "best accounting software for construction" wants a page whose type matches the intent: a comparison, a category page, an answer to that question. If you only publish feature pages, you can rank on brand strength while offering nothing an answer engine can select for comparison or recommendation intent. Map your lost questions to page types: informational questions need direct-answer content, comparison questions need honest comparison pages, local service questions need location-and-service pages. A missing page type is a content-strategy gap, not a quality gap, and no amount of polishing the wrong page fixes it. ## 4. Is the business entity ambiguous or inconsistent? Ask each engine "what is [your business]?" and read literally. If the answer is wrong, vague or confuses you with someone else, the engine cannot confidently attribute your pages to a nameable business, and mentions and recommendations fail even where citations might succeed. The usual causes are mundane: name variants across site, profiles and directories; a thin About page; missing or conflicting `Organization`/`LocalBusiness` schema; unclaimed profiles carrying stale facts. The repair sequence is in [how AI engines identify your business](/resources/getting-cited-by-ai/entity-signals). ## 5. Do independent sources support competitors instead? For each lost question, list the URLs the engines cited and check each: does it mention competitors, and does it mention you? When answers are grounded in review platforms, directories, comparison articles and threads that describe competitors only, the engine recommends competitors regardless of your on-page quality. This class is the most common explanation for the specific pattern "our pages get cited sometimes, but the recommendation always goes elsewhere". The evidence-first playbook for closing named source gaps is in [why reviews, directories and communities influence AI recommendations](/resources/getting-cited-by-ai/third-party-proof). ## 6. Is the test itself unstable? Before concluding you are invisible, check the measurement. One-off checks of one engine with ad-hoc phrasing produce unreliable verdicts: answers vary run to run, engine to engine and region to region, and a differently phrased question is a different question. A sound test uses a stable set of real buyer questions, the same engines and regions every time, and repeated runs to establish variance before judging. If your "no citations" conclusion came from a handful of manual checks, run a real baseline first; the design is in [how to measure whether an AI visibility fix worked](/resources/getting-cited-by-ai/measure-ai-citations). ## A practical sequence Work the classes in order, one change at a time, re-testing between: 1. **Access** first, because it gates everything and is fast to verify. 2. **Answer presence and intent match** next, on the handful of pages mapping to your most valuable lost questions. 3. **Entity** in parallel with content, since records correction is independent work. 4. **Independent proof** once your pages are eligible, targeting the specific sources cited for your questions. 5. **Measurement discipline** throughout: baseline, ship, re-test, attribute. Resist fixing all classes at once on all pages: an unattributable improvement teaches you nothing for the next hundred questions. ## Where CiteAgentic fits in this loop **Product example.** A CiteAgentic scan runs this diagnosis per tracked question: technical checks cover crawler access and indexability, the page audit checks whether the mapped page answers the question directly, entity checks review your records for consistency, and the citation evidence shows which sources each engine cited and where competitors appear without you. Each finding becomes a recommendation tied to the affected questions, drafts are prepared for human review with unknown facts marked, and the same questions are re-tested on later scans with page state, validation state and AI outcome reported separately. No finding claims a guaranteed citation; the re-test reports what actually changed. **Find your failure class** The free audit checks access, page answers and entity signals for your site, and shows who gets cited for questions in your category. Run the free audit → ### Frequently Asked Questions **Q: We rank #1 on Google for our main keyword. Why does ChatGPT not cite us?** Ranking proves Google indexing and standing for that keyword. ChatGPT search selects sources with its own crawler and retrieval, mixes in independent sources, and generates per-question. Check OAI-SearchBot access first, then whether your page states the answer the specific question needs. **Q: Which class is most common?** It depends on the site. Established sites with strong SEO most often fail on intent match and independent proof; newer or heavily gated sites fail on access; local businesses fail disproportionately on entity consistency. That is why the diagnosis is ordered rather than guessed. **Q: Can we skip the baseline and just start fixing?** You can, but you lose attribution permanently: without a recorded before, no later improvement can be credited to any specific fix, and you cannot tell wins from answer variance. A baseline costs one scan cycle and pays for itself on the first disputed result. --- # 2.8× AI search visibility recovery in 45 days > How a fintech platform regained AI citations by fixing schema markup and topical gaps with paste-ready fix prompts. Source: https://www.citeagentic.com/case-studies/fintech-visibility-recovery _Fintech / SaaS_ ## 2.8× increase in AI search visibility How a fintech platform recovered lost AI citations by fixing schema and topical gaps in 45 days. **TL;DR:** A mid-market fintech platform recovered lost AI citations with a 2.8× visibility increase in 45 days. Combined ChatGPT and Perplexity citations went from 20 to 56 (+180%) after 40 hours of engineering work over 3 weeks, and the platform now appears in 70% of competitor-comparison queries, up from 30%. ## The problem A mid-market fintech platform saw their AI citation rate drop 60% over three months. They appeared in ChatGPT and Perplexity results for comparison queries, but when prospects asked about specific features or use cases, the platform was invisible. Their SEO was solid — they ranked top-3 for target keywords in Google. But AI engines weren't citing them despite the ranking strength. ## Why it happened When we ran the initial audit, three issues surfaced: 1. **Schema gaps** — The product pages lacked Organization + Product schema. AI engines prioritize structured data over scraped content. 2. **Topical orphaning** — Feature pages (e.g., "risk modeling," "API integrations") had no citeable authority. They buried the reasoning under UI flows instead of publishing the methodology. 3. **Authority isolation** — The homepage and blog covered broad fintech topics, but never linked to the technical depth. AI engines couldn't trace the argument chain. All three are detectable — none require guesswork. CiteAgentic's audit flagged all of them with specific page-level findings. ## The fix The team used CiteAgentic's paste-ready fix prompts: - **Schema: 4 hours** — Handed the prompt to Cursor, deployed Product + Organization markup across 18 pages. - **Topical linking: 12 hours** — Rewrote feature pages to lead with methodology, added internal links from the homepage and blog posts to authority. - **Content depth: 24 hours** — Added a 2-minute explainer video to the risk-modeling page (embedded in a new FAQ). AI engines cite videos in 8% of answers; adding one more surface tripled coverage. Total engineering time: **40 hours over 3 weeks**. ## The results Post-deployment, within 2 weeks: | Metric | Before | After | Delta | |---|---|---|---| | ChatGPT citations | 8 | 22 | +175% | | Perplexity citations | 12 | 34 | +183% | | Combined visibility | 20 | 56 | +180% | | Featured-in-AI-summary rate | 14% | 38% | +2.8× | The fintech platform now appears in 70% of competitor-comparison queries across all major engines, up from 30%. ## Why paste-ready prompts mattered Paste-ready prompts mattered because, instead of reading an audit report and building the fix in isolation, the team: 1. Opened the fix prompt in Cursor 2. Let Claude do the schema generation (verified it was correct JSON-LD) 3. Deployed with confidence because they understood the WHY Without paste-ready context, this would've been a 3-week research cycle to figure out the linking strategy. With it, they shipped in 40 hours. ## What they learned Three lessons are worth keeping from this recovery: 1. **Schema and internal linking are foundational for AI visibility.** A site can rank everywhere in Google and be invisible in AI because AI engines ingest differently. 2. **You need to publish your methodology, not just your product.** Feature pages are landing pages. Methodology pages are citeable assets. 3. **Continuous monitoring beats one-time audits.** They now run monthly scans and catch regressions within a week instead of discovering drops months later. --- Ready to find gaps in your AI visibility? Run your first audit → --- # Social Media Agency for Brisbane & Gold Coast > Social media agency support for Brisbane and Gold Coast businesses. Plan useful content, platform scope, creative approvals and measurable next actions. Source: https://www.citeagentic.com/digital-marketing/social-media-management _Content and channel management_ ## Choose platforms around your buyers LinkedIn, Instagram and Facebook serve different audiences and formats. We agree the platform mix from your customer, offer and available content, rather than automatically posting everywhere. Additional platforms, paid campaigns, community replies, photography and video production need an explicit scope. **A focused content plan** Translate service questions, objections, offers and verified proof into a calendar. Each content theme should support awareness, evaluation or a clear next action. **Copy and creative with review built in** Prepare the agreed posts and creative formats, check factual claims and send them for approval. Your business supplies accurate service details and permission to use customer material. **Publishing and useful reporting** Schedule approved content and review available performance. Use response quality, relevant visits and enquiries alongside reach and engagement to choose what to repeat or change. ## A clear approval workflow 1. Agree audience, platforms, content formats and the number of deliverables in the proposal. 2. Build a calendar from your services, customer questions and approved source material. 3. Review draft copy and creative before scheduling. Set who approves and how revisions are handled. 4. Publish approved material, review results and adjust the next content cycle. Urgent responses, moderation and out-of-hours cover are not assumed. Agree escalation and response responsibilities if they are needed. ## What useful content can look like **Illustrative content plan for a professional-services business:** a post explaining a common buyer question, a short service-process walkthrough, an approved customer quote and a clear invitation to discuss a specific need. This is a planning example, not a published client campaign or performance claim. We use real reviews and evidence only when supplied or verified and permitted for reuse. If proof is missing, the next step is to collect it, not invent a testimonial or result. ## Connect social with the rest of your marketing A useful post can lead to a service page, guide or consultation. Coordinate those destinations with [SEO and AEO work](/digital-marketing/ai-visibility-aeo), and scope paid social separately if promotion is appropriate. Social activity itself is not proof that AI engines will recommend your brand. ## Scope the work before choosing a package Tell us your audience, current channels, available assets and approval capacity. We agree platform coverage, content volume, production needs, management fees and reporting cadence before work starts. Brisbane and Gold Coast are service areas; location-specific content should reflect your real business presence. **Build a social content plan your team can keep moving.** Bring your profiles and priority service. We will define the content, approval responsibilities and measure of success with you. --- # Which AI crawlers should your website allow? > OAI-SearchBot, GPTBot, Googlebot, Google-Extended, PerplexityBot and Anthropic's crawlers do different jobs. Which ones matter for AI search visibility, and how to verify access. Source: https://www.citeagentic.com/resources/getting-cited-by-ai/crawler-access _Getting cited by AI · Technical access_ ## Which AI crawlers should your website allow? Search crawlers, training crawlers and user-directed fetchers are different agents with different jobs. Blocking the wrong one silently removes you from AI answers. Here is what each crawler actually controls, and how to verify access end to end. **TL;DR:** For inclusion in ChatGPT search results, allow OAI-SearchBot; GPTBot is OpenAI's separate crawler associated with possible model training. For Google AI Overviews and AI Mode, the control is ordinary Googlebot: Google says pages must be indexed and eligible for a Search snippet, and Google-Extended is a separate control that does not govern those features. Perplexity recommends allowing PerplexityBot for its search results. Allowing a crawler makes you eligible; it never guarantees a citation. This article covers the access stage of the [getting cited by AI pillar](/resources/getting-cited-by-ai): before any question of content quality or authority, the platform's crawler has to be able to fetch your page. What happens after access, when engines select which fetched pages to cite, is covered in [how AI engines choose sources](/resources/getting-cited-by-ai/how-ai-chooses-sources). ## Which crawler does ChatGPT Search use? ChatGPT search inclusion depends on **OAI-SearchBot**, not GPTBot. OpenAI's [publisher guidance](https://help.openai.com/en/articles/12627856-publishers-and-developers-faq) states that OAI-SearchBot is used to surface and link to websites in ChatGPT's search results, and OpenAI publishes its [IP ranges and user agent](https://openai.com/searchbot.json) so you can verify real requests. **GPTBot** is a different crawler, associated with collecting content that may be used for model training. Blocking GPTBot expresses a training preference; it is not the switch that removes you from ChatGPT search. Many sites blocked GPTBot in 2023 to opt out of training and assumed that decision had no search cost, and under the current split that assumption is correct only if OAI-SearchBot remains allowed. There is also a user-directed agent, **ChatGPT-User**, which fetches a page when a person asks ChatGPT to open a specific link. Treat it like a visitor acting through a browser, not a crawler building an index. ## Which crawler controls Google AI Overviews and AI Mode? Ordinary **Googlebot** controls eligibility. Google's [AI features documentation](https://developers.google.com/search/docs/appearance/ai-features) says pages that appear as supporting links in AI Overviews and AI Mode must be indexed and eligible to be shown with a snippet in Search, using the same crawl and indexing pipeline as everything else. **Google-Extended is not the control for AI Overviews or AI Mode.** It is a separate token governing other Google AI uses of your content. If you disallow Google-Extended while allowing Googlebot, you remain eligible for AI Overviews; if you disallow Googlebot, you disappear from Search and its AI features together. Google also states there is no special schema type or AI-specific requirement for these features; its [guidance for generative AI experiences](https://developers.google.com/search/docs/fundamentals/ai-optimization-guide) is essentially standard indexing and content quality advice. ## Which crawler does Perplexity use? Perplexity's [crawler documentation](https://docs.perplexity.ai/docs/resources/perplexity-crawlers) recommends allowing **PerplexityBot**, its crawler for building the index behind its search results. Perplexity also documents **Perplexity-User**, a user-directed fetcher that retrieves a page when a person's query needs it; the docs note it is not used to build the index. Anthropic [documents its crawlers](https://privacy.anthropic.com/en/articles/8896518-does-anthropic-crawl-data-from-the-web-and-how-can-site-owners-block-the-crawler) similarly: **ClaudeBot** crawls the web for data that may improve models, and separate agents handle user-directed fetching and search-related retrieval. Follow Anthropic's current documentation for each agent's purpose rather than assuming ClaudeBot is a citation crawler. ## What is the difference between search, training and user-directed crawlers? Each vendor now runs up to three kinds of agent, and robots.txt lets you treat them separately: | Purpose | What it does | Example agents | If you block it | |---|---|---|---| | Search indexing | Builds the index AI search answers draw from | OAI-SearchBot, Googlebot, PerplexityBot | You become ineligible for that platform's search citations | | Training | Collects content that may be used to train models | GPTBot, ClaudeBot, Google-Extended (token) | Training preference expressed; search inclusion governed separately | | User-directed | Fetches a page a specific user asked about | ChatGPT-User, Perplexity-User | Users who ask about your page get an error instead of your content | The practical rule: decide your training position on its own merits, but keep search and user-directed agents allowed if you want to appear in AI answers. ## How should robots.txt be configured? If you want AI search visibility and accept training use, the simplest correct configuration is your normal allow-all with your private paths disallowed: ``` User-agent: * Allow: / Disallow: /api/ Disallow: /admin ``` If you want AI search visibility while opting out of training, name the training crawlers specifically and leave everything else open: ``` User-agent: GPTBot Disallow: / User-agent: ClaudeBot Disallow: / User-agent: Google-Extended Disallow: / User-agent: * Allow: / ``` Never copy a "block all AI" robots.txt template without reading it: most of those templates disallow the search crawlers too, which removes you from the answers your buyers read. ## Can a CDN or WAF block an allowed crawler? Yes, and this is the most common silent failure. Your robots.txt can say allow while the request never reaches your server, because a layer in front of it rejects the crawler: - **Bot-management rules** at CDNs can challenge or block verified crawlers by default or by category toggle. - **JavaScript challenges and CAPTCHAs** return an interstitial page instead of your content; a crawler cannot solve them. - **Geo or rate rules** can 403 the IP ranges crawlers arrive from. - **Authentication walls and cookie gates** serve the crawler a login page rather than the article. robots.txt is the invitation; the CDN decides who gets through the door. Check both. ## How do you verify that a crawler can access a page? Verify at three layers, in order: 1. **robots.txt.** Fetch `yourdomain.com/robots.txt` and check each search crawler's group resolves to allow for the page's path. 2. **HTTP response.** Check your CDN or server logs for requests from the crawler's published user agent and confirm they receive `200` with real HTML, not `403`, `429` or a challenge page. For OpenAI, validate the requester against the published [OAI-SearchBot IP list](https://openai.com/searchbot.json) rather than trusting the user-agent string alone. 3. **Index presence.** For Google, use Search Console's URL Inspection to confirm the page is indexed and snippet-eligible. For other engines, ask the platform about the page's topic and see whether the page can appear at all. If any layer fails, fix that layer before touching content. A perfectly written page behind a 403 does not exist as far as the answer engine is concerned. Access checks also protect your identity signals: profiles and schema only help if the pages carrying them are fetchable, which is why [entity diagnosis](/resources/getting-cited-by-ai/entity-signals) starts after access is confirmed. ## What CiteAgentic's technical audit reports **Product example.** CiteAgentic's technical audit checks this access chain for your site: robots.txt rules per crawler, HTTP status served to honest crawler requests, rendering that hides content from non-browser readers, and missing or conflicting index directives. Each failure becomes a specific finding with the fix instruction, and after you deploy the change a re-scan verifies the same check now passes. It reports access as eligibility only; no access fix is ever presented as a promised citation. **Check your crawler access in one scan** The free audit reports which AI search crawlers can and cannot read your key pages, and what is blocking them. Run the free audit → ### Frequently Asked Questions **Q: If I allow every crawler, will I get cited?** No. Access is the eligibility stage: it lets your pages into the candidate pool. Whether a page is actually selected depends on relevance to the question, the clarity of the answer, entity signals and the independent sources the engine also retrieves. **Q: Does blocking GPTBot remove my site from ChatGPT?** Blocking GPTBot expresses a preference about training use. ChatGPT search inclusion is governed by OAI-SearchBot per OpenAI publisher guidance, and a page a user explicitly asks about is fetched by ChatGPT-User. Check each agent separately in your robots.txt. **Q: Should I block Google-Extended?** That is a policy decision about AI use of your content, not a visibility decision for AI Overviews or AI Mode. Google documents that those features rely on normal Googlebot indexing and snippet eligibility. Disallowing Google-Extended does not remove you from them. **Q: How do I know a request claiming to be a crawler is real?** Vendors publish verification methods: OpenAI publishes OAI-SearchBot IP ranges, and Google supports reverse-DNS verification of Googlebot. Verify before whitelisting, because spoofed user agents are common. --- # How do AI engines identify your business? > Entity clarity for AI search - consistent naming, About pages, Organization and LocalBusiness schema, sameAs links, Business Profile and directory records, and how to diagnose an ambiguous entity. Source: https://www.citeagentic.com/resources/getting-cited-by-ai/entity-signals _Getting cited by AI · Entity_ ## How do AI engines identify your business? Before an engine can recommend you, it has to be sure who you are: one name, one category, one set of facts, consistent across your site and the independent records that describe you. Ambiguity is resolved by leaving you out. **TL;DR:** AI engines identify a business by reconciling what your site says about you with what independent records say: your About page, Organization or LocalBusiness schema, sameAs links, your Google Business Profile, and directory listings. When the name, category, location and description agree everywhere, the engine can name you confidently; when they conflict or the name is shared with other entities, the safe move for the engine is to omit you or confuse you with someone else. Entity work is making every record tell the same story. This is the identity stage of the [getting cited by AI pillar](/resources/getting-cited-by-ai): the page can be crawled and answers the question, but the engine still has to attribute it to a business it can describe. ## What is a business entity in AI search? An entity is the thing behind the pages: the business itself, with a name, a category, locations, and relationships to other records on the web. When an engine mentions or recommends a business, it is making a claim about the entity, assembled from every source that describes it, not just the page in front of it. That assembly is where businesses fail invisibly. Your page can be the best answer available, but if the engine cannot connect the page, the brand name, the Business Profile and the review records into one confident identity, it has a fact it cannot safely attribute. ## Why does consistent naming matter? Because reconciliation runs on matching. If your website says "Wade Building Co", your Business Profile says "Wade Building Company Pty Ltd", and directories carry "Wade Builders", the engine has three candidate entities and weak evidence they are the same one. Every variant you tolerate is a chance to split your identity or hand it to a similarly named business. Pick one public-facing name and use it everywhere: site, schema, profiles, directories, review platforms, social accounts. Keep the legal name in legal contexts; the entity name buyers and engines see should not vary. This applies doubly for local businesses, where a name is often shared by unrelated companies in other cities. ## What information should an About page state? The About page is the page engines can most safely quote about who you are, so state the facts plainly in the first paragraphs: what the business is, what category it belongs to, who it serves, where it operates, and since when. Write the category sentence the way you would want an AI answer to phrase it, because that sentence is what tends to get reused. Add the verifiable specifics that separate you from a generic description: founders and their roles, locations served, credentials or licences where relevant. Vague About pages ("we're passionate about excellence") give an engine nothing attributable; concrete ones do the entity's talking. ## What does Organization or LocalBusiness schema contribute? Schema states your identity in a machine-readable form on your own site: name, URL, logo, and for local businesses the address, service area and hours. Use the type that matches reality; a plumber is a `LocalBusiness` (or its subtype), a software company an `Organization`, and the properties should mirror what the visible pages say. Treat it as accurate self-description, not a ranking device. Google documents no special schema requirement for AI features; the value of the markup is that it removes ambiguity about who publishes the site, which is exactly the failure mode entity work targets. ## How should sameAs links be used? `sameAs` connects your schema to the other records that are you: your LinkedIn page, Business Profile map link, review-platform profiles, social accounts, and Wikipedia or Wikidata entries if you genuinely have them. It is an explicit statement that these scattered records describe one entity. Only link records you control or that are verifiably about you. A `sameAs` pointing to a stale profile with a different name or an abandoned account adds conflict instead of resolving it; audit the targets before you connect them. ## Why do Business Profile and directory records matter? Because they are the independent half of the reconciliation. Engines weight records a business does not fully control: a Google Business Profile with a category, address, hours and reviews; industry directories; established listing sites. When those agree with your site, your identity is corroborated by parties other than you. For local businesses these records also carry the location facts that decide local answers: which suburbs you serve, where you actually are. Keep name, address, phone and category identical across them, and claim the profiles you have not claimed; an unclaimed profile with wrong hours is an error source publishing under your name. This overlaps with [third-party proof](/resources/getting-cited-by-ai/third-party-proof), where the same independent sources also carry reputation. ## How do you diagnose an ambiguous business entity? Ask the engines directly and read the answer literally. "What is [business name]?" and "Who is [business name] in [city]?" across two or three engines reveals the entity state fast: - **Correct and specific** means the entity resolves; move your attention to content and proof. - **Vague but not wrong** usually means thin records: the About page and profiles do not state enough facts to describe you. - **Confused with another business** means naming collision; strengthen the disambiguating facts (location, category) everywhere. - **Unknown** means the records are missing or inaccessible; verify [crawler access](/resources/getting-cited-by-ai/crawler-access) before assuming anything else. Re-run the same questions after fixing the records. Entity answers move as engines re-encounter the corrected sources, so treat this as a tracked outcome, not a one-time setup task. ## How CiteAgentic represents entity connections in the Authority Map **Product example.** CiteAgentic's Authority Map draws the entity's connections as they exist in your tracked data: your buyer segments and topics, the questions under them, the pages that answer each question, the competitors named in answers, and the off-site sources engines cite. Profile and entity checks in the audit surface missing or inconsistent records (schema absent, naming variants, unclaimed or inconsistent profiles) as findings with fix tasks, and local source coverage shows which location-relevant sources describe you versus your competitors. Everything shown is evidence-matched from scans; sources not yet tracked are labelled as untracked rather than guessed at. **See how engines describe your business** Track "what is your business" style questions alongside your buyer questions, and see whether engines describe you correctly, vaguely, or not at all. Explore AI Visibility → ### Frequently Asked Questions **Q: We share a name with another company. Are we stuck?** No, but disambiguation becomes your main entity task: state category and location beside the name everywhere, keep every record consistent, and build the independent records (profiles, directories, reviews) that describe your version of the name. Engines resolve shared names by context; give them context that always points to you. **Q: Do we need a Wikipedia page?** Only if you genuinely meet notability standards. A well-maintained Business Profile, consistent directories and real review presence resolve most business entities fine. Forcing a Wikipedia page that gets deleted or flagged harms more than it helps. **Q: How long does entity correction take to show up?** It varies by engine and by which records were wrong; there is no fixed timeline. Correct the records, then re-test the same identity questions on a schedule and judge the trend rather than any single answer. --- # Why do reviews, directories and community sources influence AI recommendations? > AI answers lean on sources a business does not control - reviews, directories, Reddit, YouTube, forums. Which ones matter by business type, how to participate legitimately, and how to measure it. Source: https://www.citeagentic.com/resources/getting-cited-by-ai/third-party-proof _Getting cited by AI · Independent proof_ ## Why do reviews, directories and community sources influence AI recommendations? Engines trust sources precisely because you don't control them. When the retrieved reviews, directories and threads describe your competitors and not you, the answer recommends your competitors. This is where most recommendation losses actually happen. **TL;DR:** When an AI engine answers "who should I use for X", it retrieves independent sources alongside brand websites: review platforms, directories, community threads, video content and editorial comparisons. A business described consistently and positively by sources it does not control is one the engine can safely recommend; a business vouched for only by its own website is a claim, not a corroborated fact. The work is finding which independent sources engines actually cite for your questions, and earning a legitimate, evidence-backed presence there. This is the corroboration stage of the [getting cited by AI pillar](/resources/getting-cited-by-ai): the on-page work makes you eligible; independent proof is what moves a recommendation your way. ## Why do AI answers use sources a business does not control? Because recommendation questions are trust questions, and self-description is weak evidence. Every business website says the business is good; review platforms, directories and communities carry information the business could not simply write about itself. When an engine composes "best plumber in Brunswick" or "best CRM for agencies", the sources worth citing are the ones with independent standing. You can see this directly by reading answers: recommendation-shaped questions cite review sites, comparison articles and threads far more than they cite the recommended businesses' own pages. Your website makes you eligible; other people's websites make you recommendable. ## Which third-party sources matter for different business types? The source universe changes completely by segment, which is why copying a SaaS playbook as a local trade business wastes a quarter: - **Local services** (trades, health, hospitality): Google Business Profile reviews, local directories, community groups and area-specific threads. - **Software and B2B**: software review platforms, comparison and alternatives articles, Reddit and professional communities, YouTube reviews. - **E-commerce and consumer products**: product reviews, buying guides, video reviews, category communities. - **Professional services** (legal, finance, agencies): professional directories, client reviews where regulation allows, industry publications and LinkedIn presence. Regulation is part of the map: several health verticals restrict testimonial solicitation, and legal and financial services carry advertising rules. The right playbook respects your segment's law, not just its marketing habits. ## How do reviews support a recommendation? Reviews give the engine attributable evidence about outcomes: volume, recency, specifics and how the business responds. A profile with recent, detailed reviews mentioning the actual services gives an answer engine sentences it can rely on; a profile with three reviews from 2021 gives it nothing current. The legitimate playbook is boring and works: ask every real customer at the natural moment, make leaving a review easy, respond to all of them, and never fake, buy or selectively solicit. Review-integrity law (FTC in the US, ACCC in Australia, CMA in the UK) treats fabricated or incentivized-undisclosed reviews as enforcement territory, and platforms remove them. ## When do Reddit, LinkedIn, YouTube and forums appear? Community and video sources appear most in answers to experience questions: "is X worth it", "X vs Y", "what do people actually use for Z". Engines retrieve threads and transcripts where real users compare options in exactly the language buyers use, which makes those pages strong grounding material for recommendation answers. Their weight varies by engine and by question; there is no fixed share. What matters is observation: when community threads recur in the citations for your tracked questions, community presence is part of your specific map, not a generic tactic. ## How do you find the sources supporting competitors? Work from the answers backwards, question by question. For each buyer question where a competitor is recommended and you are absent, list the cited sources, then check each one: does it mention the competitor, and does it mention you? The output is a concrete gap list with names and URLs: "this comparison article covers three competitors and not us; this directory lists them and not us; this thread recommends them." That list, ranked by how often each source is cited across your questions, is an off-site plan grounded in evidence, and the opposite of "build more backlinks". The [citation graph article](/resources/citation-graph-vs-backlinks) covers why this graph, not the link graph, is the right map for AI answers. ## How should a business participate without astroturfing? Openly, usefully, and as yourself. The line is simple: add genuine value with your identity disclosed, or stay out of the thread. - **Answer real questions** in communities where you have expertise, disclosing who you are; recommend yourself only where genuinely relevant, sparingly. - **Correct factual errors** about your business politely, as the business. - **Pitch inclusion honestly** to directory and comparison-article publishers, with evidence of fit, accepting that editorial decisions are theirs. - **Never** post fake reviews, run undisclosed sockpuppets, or pay for placement disguised as opinion. Platforms detect it, communities remember it, and being described *inaccurately* at scale is the outcome. Astroturfing risks worse than removal: the descriptions engines then retrieve about you are the community's account of your behavior. ## How can an off-site action be measured? The same way as on-page work: baseline first, then act, then re-test the affected questions. Record which sources are cited for the question and whether you appear on them, take the action (review push, directory listing, community participation, pitch), and then track two things separately: whether your presence on the source materialized, and whether later answers to the question changed. Keep the two outcomes honestly apart. Earning a listing is verifiable immediately; the answer changing is probabilistic and slower, and sometimes the answer changes for other reasons. The [measurement article](/resources/getting-cited-by-ai/measure-ai-citations) covers the prompt-level tracking that makes this attributable at all. ## How CiteAgentic's Source Opportunities and Engage Online work **Product example.** CiteAgentic maps each independent source to the tracked buyer questions where engines cite it, classified as communities, reviews or directories, and ranks the opportunities by evidence: how often the source is cited, which questions it decides, and where competitors appear without you. Engage Online supports the participation loop on sources like Reddit, LinkedIn, Quora, X and YouTube-related evidence: it drafts a helpful, disclosed response for a real cited conversation, a human reviews and publishes it, and the outcome is tracked against the affected questions. Engagement metrics are shown only when the relevant platform adapter supplies them; when a metric is unavailable it is shown as unavailable, never as zero. **Find the sources deciding your questions** See which reviews, directories and communities the engines cite for your buyer questions, and where competitors are present without you. See how teams use it → ### Frequently Asked Questions **Q: Is this just link building with a new name?** No. Link building optimizes a graph-level authority signal for traditional ranking; this work earns descriptions of your business on the specific pages engines retrieve and cite for your questions. A mention with no link on a frequently cited source can matter here; a high-authority link from a page engines never cite does not. **Q: Can we just respond to negative reviews and skip the rest?** Responding matters, but volume and recency of genuine reviews carry the evidence. A well-handled complaint on a profile with recent detailed reviews reads very differently from the same complaint on a stale profile. **Q: Should we hire someone to post about us on Reddit?** Not undisclosed. Paid or undisclosed advocacy violates platform rules and review-integrity law in several jurisdictions, and communities are effective at detecting it. Disclosed, genuinely useful participation by your own team is the only durable version. --- # How to get mentioned by AI - the prioritised checklist > The implementation checklist for AI visibility, in pipeline order - crawler access, extractable answers, entity clarity, independent proof, and verifying each change against a baseline. Source: https://www.citeagentic.com/resources/getting-cited-by-ai/get-mentioned-by-ai _Getting cited by AI · How-to_ ## How to get your business mentioned by AI answer engines. The implementation checklist, ordered by the pipeline that decides answers: access first, then content, then entity, then independent proof, then measurement. Each step ends with how to verify it. **TL;DR:** Work the stages in order: confirm each platform's search crawler can fetch and index your pages, make your key pages answer buyer questions directly in the first paragraph, make your business identity consistent across your site and independent records, earn legitimate presence on the sources engines cite for your questions, and re-test the same questions against a recorded baseline after every change. Mention, recommendation and citation are separate outcomes; every step improves your likelihood, and none guarantees the answer changes. This is the working checklist for the [getting cited by AI pillar](/resources/getting-cited-by-ai). Three outcomes to keep separate as you work: **mentioned** (your name appears in an answer), **recommended** (the answer presents you as a choice), and **cited** (your URL appears as a source). They move independently, and each step below notes which it targets. ## Step 1: Confirm crawler access Before any content work, verify the search crawlers can actually reach you. The names matter: ChatGPT search inclusion depends on **OAI-SearchBot** (GPTBot concerns possible training use); Google's AI Overviews and AI Mode use ordinary **Googlebot** indexing and snippet eligibility, not Google-Extended; Perplexity recommends allowing **PerplexityBot**. Check three layers: robots.txt rules per crawler, what your CDN or WAF actually serves them (challenges and 403s override robots.txt in practice), and whether the pages are indexed. The full crawler map and verification method are in [which AI crawlers should your website allow](/resources/getting-cited-by-ai/crawler-access). *Verify:* fetch your robots.txt and read each search crawler's rules; confirm crawler requests receive `200` with real HTML in your logs; confirm key pages are indexed in Search Console. ## Step 2: Make your answers extractable One page per important buyer question, direct answer in the first paragraph, question-led headings in the buyer's phrasing, claims sourced beside the sentences that make them. The page formula is `direct answer → evidence → example → qualification → next step`, and it targets citation eligibility directly. Keep the answer in real text: content that only exists after JavaScript runs, inside images, or spread across a five-intent page is hard to select and harder to quote. The full blueprint is in [how to structure a page so AI can cite it accurately](/resources/getting-cited-by-ai/make-content-citable). *Verify:* read the first paragraph under each heading and check it answers the heading on its own; then ask the target question in the engines you track and record whether your page is cited, as a baseline rather than a verdict. ## Step 3: Make your entity unambiguous This step targets mentions and recommendations: the engine has to be certain who you are before it can name you. One public-facing business name everywhere; an About page that states plainly what you are, your category, who you serve and where you operate; `Organization` or `LocalBusiness` schema matching the visible pages, with `sameAs` links to the records that are genuinely you; Business Profile and directories claimed and consistent. Structured data here is accurate self-description, not a ranking device; its value is removing ambiguity. The diagnosis method and the naming-collision playbook are in [how AI engines identify your business](/resources/getting-cited-by-ai/entity-signals). *Verify:* ask each engine "What is [your brand]?" A wrong, vague or confused answer means the entity layer is failing before any content question matters. ## Step 4: Earn independent proof Recommendations mostly follow sources you do not control: reviews, directories, comparison articles, community threads. Work from evidence, not generic authority-building: for each lost question, list the sources the engines cited, check which mention competitors but not you, and earn a legitimate presence on the recurring ones. Participation rules are strict and worth following: genuine reviews asked for at the natural moment, disclosed community participation, honest pitches to publishers, never astroturfing. The by-business-type source map and the measurement loop for off-site work are in [why reviews, directories and communities influence AI recommendations](/resources/getting-cited-by-ai/third-party-proof). *Verify:* for each target question, keep the cited-source list current and record whether your presence on each named source materialized. That is verifiable immediately, separately from whether answers later change. ## Step 5: Measure prompt by prompt, then re-test Baseline before changing anything: per question, per engine, per region, across multiple runs, recording mentioned, recommended and cited separately. Then ship one change at a time, note the date, and re-test the same questions on the following scans. Judge trends across repeated runs, never a single answer; generation varies run to run, and one good answer the morning after a fix is a sample, not a result. There is no honest universal waiting period, because recrawl and answer refresh vary by engine, site and page. The full test design is in [how to measure whether an AI visibility fix worked](/resources/getting-cited-by-ai/measure-ai-citations). *Verify:* the measurement is the verification. If the answer does not move, diagnose which stage failed in order, starting from access, before shipping more of the same fix. ## How CiteAgentic runs this checklist **Product example.** CiteAgentic tracks your buyer questions across engines and regions, audits access and pages, ties each recommendation to the specific questions it should affect, prepares drafts for human review with unknown facts marked rather than invented, and re-tests the affected questions on later scans, reporting page state, validation state and AI outcome separately. The product's framing matches this article's: each step improves your likelihood of being mentioned, recommended or cited, and the re-test verifies what actually changed. **This checklist, automated** Track your questions, find the failing stage for each loss, get the fix drafted for review, and re-test after you ship. Start free trial → ### Frequently Asked Questions **Q: How long until these changes show up in AI answers?** It varies by engine, site and page, and no honest fixed timeline exists. Ship one change at a time, keep re-testing the affected questions on a schedule, and judge the trend across runs rather than any single answer. **Q: Can I pay to be mentioned by ChatGPT or Google AI Overviews?** There is no paid placement in organic AI answers today. Presence is earned through accessible pages, extractable answers, a resolvable entity and independent corroboration, which is also why it defends well once you have it. **Q: Should I optimise for every AI engine separately?** Mostly no. Access, answer-first structure, entity clarity and independent proof move all engines. Differences show up in which sources each engine favours, which matters when you choose where to earn presence, not how you write. **Q: What single change usually matters most first?** Whichever stage is failing first for your site, which is why the checklist is ordered. A blocked crawler makes content work unmeasurable; a missing page type makes entity polish irrelevant for that question. Diagnose before investing. --- # How should you structure a page so AI can cite it accurately? > The page formula for citable content - direct answer, evidence, example, qualification, next step - plus how to source claims, use structured data honestly, and verify after publishing. Source: https://www.citeagentic.com/resources/getting-cited-by-ai/make-content-citable _Getting cited by AI · Content_ ## How should you structure a page so AI can cite it accurately? A citable page states the answer where a machine can lift it, supports claims where a machine can check them, and covers one question well. The formula: direct answer, evidence, example, qualification, next step. **TL;DR:** Structure each important page around one buyer question. State the direct answer in the first paragraph, then support it: evidence beside each claim, a concrete example, honest qualifications, and a clear next step. Use question-led headings that match how buyers actually ask, keep the answer in real text rather than images or scripts, and let structured data describe the visible page accurately. No writing formula guarantees a citation; this structure removes the reasons a relevant page gets passed over. This is the content stage of the [getting cited by AI pillar](/resources/getting-cited-by-ai): the page is accessible and retrievable, and now it has to be worth quoting. ## What does "citable content" actually mean? A citable page is one an answer engine can safely use as a supporting source: the answer to a specific question is present, stated plainly, and attributable. When an engine composes a response, it needs sentences it can paraphrase or link without distorting your meaning; pages that hint, tease or spread the answer across sections force the engine to either reconstruct your point (risky) or use a clearer page (easy). Citable is not the same as long, and not the same as keyword-optimized. A 900-word page that answers one question directly is a better source than a 3,000-word page that answers it eventually. ## Should the answer appear near the top? Yes. Put the direct answer in the first paragraph under the heading that asks the question, then elaborate. This is not about gaming a token limit; it is how supporting sources get selected: a page whose opening plainly answers the retrieved query is easy to judge relevant, and a page that opens with scene-setting is not. The same rule applies inside every section: the first sentence or two under each heading should answer that heading's question, with context after. If you delete everything but the first paragraph of each section and the page still answers its questions, the structure is right. ## How should question-led sections be written? Write headings as the questions buyers actually ask, in their words, one intent per section: - **Match real phrasing.** "How much does bathroom renovation cost in Melbourne?" beats "Pricing considerations", because retrieval matches meaning and the question form states the intent exactly. - **One question per section.** A section answering three things is hard to cite for any of them. - **Answer, then qualify.** Lead with the direct answer; put the "it depends" honestly in the sentences after, not instead of the answer. - **Keep paragraphs short.** Two or three sentences each; a wall of prose buries the quotable sentence. The full page follows the same arc: `direct answer → evidence → example → qualification → next step`. ## When should you use lists or comparison tables? Use a list when the answer is genuinely enumerable (steps, options, criteria) and a table when the reader would otherwise have to hold two or more things in mind to compare them. Both formats state structure explicitly, which makes the content easier to reuse accurately in a generated answer. Do not force them. A list of one-line fragments where each item needs explanation is worse than three clear paragraphs, and a decorative table is noise. The test is whether the structure carries real information about the relationships between items. ## How should claims and statistics be sourced? Put the source beside the claim, in the same sentence or the one after, linked to the original. Claims worth making are claims someone could check: a number with a named source and date, a quote with an author, a comparison with stated criteria. Unsourced numbers are a liability twice over: readers discount them, and an engine that reuses your claim inherits your error. If you cannot source a number, state the mechanism qualitatively instead of inventing a benchmark. Research context deserves the same honesty; the [GEO research paper](https://arxiv.org/abs/2311.09735) found that adding statistics, quotations and source citations improved content visibility inside its controlled generative-engine test setup, which is evidence for evidence-dense writing, not a promised lift for your site. ## What role does structured data play? Structured data describes the visible page so machines can classify it confidently: what type of page it is, who published it, when it was updated. Google [states explicitly](https://developers.google.com/search/docs/appearance/ai-features) that no special schema is required for AI Overviews or AI Mode, so treat markup as accurate description, never as a citation lever. The rules that follow from that: use the schema type that matches what the page actually is, keep the markup consistent with the visible content, and never mark up content that is not on the page. Entity-level markup (Organization, LocalBusiness, sameAs) matters more for who you are than what this page says; that side is covered in [entity signals](/resources/getting-cited-by-ai/entity-signals). ## How do internal links help discovery? Internal links tell crawlers the page exists and tell every reader, human or machine, how it relates to the rest of your site. A new page with no inbound links from your own site depends entirely on sitemaps for discovery and offers no context about its place in your content. Link from your relevant existing pages to the new one with descriptive anchor text, and link from the new page back to its hub or parent topic. Within a topic cluster like this one, every article links to the pillar and the pillar links to every article; that is discovery and topical context in one structure. ## How do you verify the page after publishing? Three checks, in order: 1. **Render check.** View the served HTML (not the browser after scripts run) and confirm the answer text is present. Then confirm the page is indexed, via Search Console for Google. 2. **Extraction check.** Read the first paragraph under each heading and ask whether it answers the heading alone, out of context. That is how it will be used. 3. **Answer check.** Ask the target question in the engines you care about, more than once, and record whether the page is cited and whether the brand is mentioned. Treat the result as a baseline to compare against later, not a verdict; the [measurement article](/resources/getting-cited-by-ai/measure-ai-citations) covers how to do this properly. ## How CiteAgentic turns a content gap into a reviewed fix **Product example, demo scenario.** A local builder tracks the question "who does bathroom renovations in my area?" and is absent from every engine's answer. The page audit finds there is no page answering the question: the services page mentions bathrooms in one sentence, with no direct answer, no evidence and no local detail. The recommendation identifies the missing page and its required sections, the agent prepares a draft with the blueprint structure and marked placeholders for facts only the builder knows (licence number, real project details, service suburbs), a human reviews and completes it before publishing, and the same buyer question is re-tested on later scans to see whether the answer changed. This is a demonstration of the workflow, not a customer result. **Audit a page against this blueprint** The page audit checks your key pages for direct answers, question-led structure, evidence density and extractability, and drafts the fix for your review. Run the free audit → ### Frequently Asked Questions **Q: Does following this structure guarantee a citation?** No. Structure makes a relevant page easy to select; it cannot make an irrelevant page relevant, and selection also depends on competing sources, entity clarity and the independent evidence engines retrieve alongside your site. Treat structure as removing reasons to be passed over. **Q: Is FAQ schema required for AI answers?** No. Google documents that no special schema is required for its AI features. A visible FAQ section is useful when readers genuinely have follow-up questions; markup should simply describe it accurately. **Q: How long should a citable page be?** As long as the question requires. A focused 900-word answer beats a padded 1,800-word article; length added for its own sake buries the quotable sentences and dilutes the page intent. --- # How do you measure whether an AI visibility fix worked? > Baselines, mentioned vs recommended vs cited, prompt-level tracking, engine and region coverage, and how to design a before-and-after test that survives AI answer variance. Source: https://www.citeagentic.com/resources/getting-cited-by-ai/measure-ai-citations _Getting cited by AI · Measurement_ ## How do you measure whether an AI visibility fix worked? AI answers vary run to run, engine to engine, region to region. Without a baseline, a stable prompt set and repeated scans, you cannot tell a win from noise. Here is how to measure honestly. **TL;DR:** Measure before you change anything: record which engines mention, recommend or cite you for each tracked buyer question, repeatedly enough to see the variance. Ship one change, note the date, and re-test the same questions on the same engines over the following scans. Judge trends across repeated runs, never a single answer, and track mentioned, recommended and cited as separate outcomes because they move independently. Recrawl and answer changes vary by engine, site and page; no fixed waiting period is honest. This is the verification stage of the [getting cited by AI pillar](/resources/getting-cited-by-ai), and it is what separates an optimization program from a guessing program. ## What should be measured before making a change? The baseline: for each buyer question you care about, which engines answered, whether your brand was mentioned, whether it was recommended, which URLs were cited, and which competitors appeared. Collected more than once, because a single run cannot show you the variance you will later need to distinguish from improvement. Baselines are cheap insurance. Every "did it work?" argument three months from now is settled by whether you recorded what the answers looked like before. A change shipped without a baseline can never be credited with anything. ## What is the difference between mentioned, recommended and cited? Three separate outcomes that move independently, and conflating them corrupts every measurement: - **Mentioned**: your brand name appears anywhere in the answer, in any role, including "unlike X" or a neutral listing. - **Recommended**: the answer presents you as a choice for the buyer's need, which is the outcome revenue actually cares about. - **Cited**: one of your URLs appears as a supporting source, which can happen without a mention and vice versa. A fix can move one without the others: a rewritten page can earn citations while the recommendation still goes to a competitor with stronger reviews. Tracking the three separately tells you which stage of the journey moved; a single blended score hides exactly the information you need. ## Why should results be tracked prompt by prompt? Because fixes are prompt-specific and averages bury them. A page fix targets particular questions; if you measure a portfolio-wide visibility number, a real win on three fixed questions drowns in the noise of two hundred untouched ones, and an unrelated loss elsewhere can mask it entirely. Prompt-level tracking also produces the diagnosis for the next round: which questions you win, which you lose, and who wins the ones you lose. That per-question evidence is what makes the next recommendation grounded instead of generic. ## How many engines and locations should be included? Every engine your buyers plausibly use, and every region you actually serve. Coverage decisions are measurement decisions: engines select different sources for the same question, and answers differ by region, so a fix can work on one engine in one country and be invisible in your dashboard if you only test another. For local businesses, region is not optional: "best electrician near me" answers are location-dependent by construction, and measuring from the wrong region measures a different question. Test from where your buyers are. ## Why is one AI answer not a reliable trend? Because generation is not deterministic. The same question to the same engine on the same day can produce different answers with different sources; across days, retrieval changes as pages update and indexes refresh. One answer is one sample from a distribution. The failure mode this produces is real: someone checks ChatGPT the morning after a fix, sees the brand mentioned, and declares victory; a week later the mention is gone and the fix is declared a failure. Both conclusions came from single samples. Repeated runs are the only way to see whether the distribution moved. ## How should a before-and-after test be designed? Five properties make the test survivable: 1. **A stable prompt set.** The same questions, phrased the same way, before and after. Changing the questions mid-test destroys comparability. 2. **A recorded baseline** across multiple runs, so the pre-change variance is known. 3. **One change at a time** per affected question set, with the ship date noted. Three simultaneous changes produce an unattributable result. 4. **The same engines and regions** in every run. 5. **A decision window agreed in advance**: how many post-change scans you will look at before judging, so the judgment is not made on whichever single run looks best. Recrawl and answer refresh vary by engine, site and page; there is no universal number of days after which the result is in. State the mechanism honestly: the change improves your likelihood of being selected, and repeated retesting verifies whether it did. ## What should happen when the answer does not change? Diagnose which stage failed, in order, rather than shipping more of the same fix. Confirm the changed page is actually being served and indexed ([crawler access](/resources/getting-cited-by-ai/crawler-access) first). Check whether the engines' cited sources for the question even include pages like yours, or whether the answer is grounded in reviews and threads your page cannot influence; that finding redirects effort to [independent proof](/resources/getting-cited-by-ai/third-party-proof). Sometimes the honest conclusion is that the question is dominated by sources you are not on yet, and the on-page work was necessary but not sufficient. That is the citation journey working as designed: eligibility first, corroboration second, outcome probabilistic throughout. The [diagnostic article](/resources/traffic-but-no-ai-citations) walks the failure classes in order. ## How CiteAgentic links recommendations to later scans **Product example.** Every CiteAgentic recommendation is tied to the specific buyer questions it should affect. The scan history is the baseline; when a fix is approved and shipped, the affected questions are re-tested on subsequent scans, and the page state, validation state and AI outcome are reported separately: the page now answers the question, independent corroboration exists or is still missing, and the answer outcome moved from not present to mentioned, cited or recommended, or did not move yet. No part of the product claims a fix will produce a citation; it records whether one did. **Baseline your questions this week** Start tracking your real buyer questions across engines and regions now, so every future fix has a before to compare against. Start free trial → ### Frequently Asked Questions **Q: How long should we wait before judging a fix?** Agree the window in advance in scans, not days: for example, judge after the next several scheduled scans of the affected questions. Engines recrawl and refresh at different rates, so a fixed number of days is false precision. **Q: Our visibility score went up. Was it the fix?** Only if the movement is on the questions the fix targeted, exceeds the variance you saw in the baseline, and persists across runs. Portfolio-level scores move for many reasons, including competitor changes and engine updates you did not cause. **Q: Can traffic analytics measure AI visibility instead?** Referral traffic from AI surfaces is worth tracking, but it measures clicks, not presence. Many AI answers satisfy the buyer without a click, and a recommendation that shapes a shortlist can produce a branded search rather than a referral. Measure the answers directly. --- # How do AI search engines choose sources to cite? > What public documentation says about how ChatGPT search, Google AI Mode and Perplexity select supporting pages - query fan-out, relevance, and why competitors get cited instead of you. Source: https://www.citeagentic.com/resources/getting-cited-by-ai/how-ai-chooses-sources _Getting cited by AI · Source selection_ ## How do AI search engines choose sources to cite? Each platform's ranking system is proprietary, but the public documentation agrees on the shape: retrieve candidate pages, expand the question into related searches, select relevant supporting sources, generate an answer with links. Here is what that means for your pages. **TL;DR:** No vendor publishes its citation algorithm, so treat any claimed universal formula with suspicion. What public documentation supports is a general sequence: the platform crawls or retrieves candidate pages, often expands the user's question into several related searches, selects pages that directly support parts of the answer, and generates a response with links to those pages. Your practical levers are being retrievable, being clearly relevant to the specific sub-questions, and being corroborated by the independent sources retrieved alongside you. This is the source-selection stage of the [getting cited by AI pillar](/resources/getting-cited-by-ai). It assumes [crawler access](/resources/getting-cited-by-ai/crawler-access) is already working. ## What happens between a buyer question and an AI citation? The documented sequence has four steps, whatever the platform. First, the system needs candidate content: an index built by its search crawler, or a live retrieval at answer time. Second, it interprets the user's question, often expanding it into multiple related searches. Third, it selects the pages that best support the specific claims the answer needs. Fourth, it generates the answer and attaches links to the supporting pages. Google describes its AI features this way: pages must be indexed and snippet-eligible through normal Search, per its [AI features documentation](https://developers.google.com/search/docs/appearance/ai-features), and its [generative AI guidance](https://developers.google.com/search/docs/fundamentals/ai-optimization-guide) frames optimization as standard content quality work. OpenAI's [publisher guidance](https://help.openai.com/en/articles/12627856-publishers-and-developers-faq) describes OAI-SearchBot surfacing and linking sites in ChatGPT search results. The internals differ per engine; the shape is shared. ## What is query fan-out? Query fan-out is the expansion of one user question into several searches. Google's AI Mode documentation describes issuing multiple related queries on the user's behalf and assembling the response from what they return. A question like "best accounting software for a small construction business" may fan out into searches about accounting software categories, construction-specific features, pricing and reviews. This matters because your page competes at the sub-question level, not the headline-question level. A page that thoroughly answers one of the fanned-out searches can be cited in an answer whose headline question you never targeted. It also means citation opportunities are wider than your keyword list suggests. ## Why are different sources cited for the same question? Because answers are generated fresh each time, from retrieval that varies. Ask the same question twice and the fan-out queries, the retrieved candidates and the generated text can all differ. Region, language, account context and time all move the result. The practical consequence: a single AI answer is an anecdote, not a measurement. Judging visibility requires the same questions asked repeatedly across engines and regions, which is why the [measurement article](/resources/getting-cited-by-ai/measure-ai-citations) in this cluster treats one-off checks as unreliable. ## What makes a page relevant to a grounding query? The page has to contain the answer to the specific retrieved question, stated clearly enough to support a sentence in the generated response. In practice that favors pages where the direct answer is present in text (not implied, not in an image), where the page covers one intent rather than five, and where claims are concrete enough to be quoted or paraphrased safely. None of this is exotic: it is the same relevance judgment search has always made, applied at the level of passages that can support an answer. The [citable content article](/resources/getting-cited-by-ai/make-content-citable) covers how to structure for it. ## Does a citation mean the page influenced the answer? Not necessarily, and the distinction matters for measurement. A link in an AI answer means the page was selected as a supporting source for some part of the response. It does not tell you how much of the answer's substance came from that page, and a brand can be mentioned in an answer without any of its pages being linked. That is why mention, recommendation and citation should be tracked as separate outcomes. An answer can mention you without citing you, cite you without recommending you, or recommend you based entirely on third-party sources. ## Why are competitors cited instead of your business? Usually for one of four reasons, each with a different fix: 1. **They are retrievable for the sub-questions and you are not.** Their pages answer the fanned-out searches; yours target different phrasing or do not exist. 2. **Their pages state the answer more extractably.** Same information, but theirs is stated directly where a model can lift it. 3. **The independent sources talk about them.** Reviews, directories and community threads retrieved alongside the brand pages describe the competitor, so the answer does too. 4. **Their business entity is clearer.** The engine can confidently say who they are and what category they belong to; yours is ambiguous. Diagnosing which reason applies to each lost question is most of the work. Guessing wrong sends you into a quarter of content production when the actual gap was a review profile, or vice versa. ## How should marketers analyse cited sources? Work backwards from real answers, question by question. For each buyer question you care about, record which sources every engine cited, classify each source (your site, a competitor's site, a review platform, a directory, a community thread, an editorial page), and note whether your brand appears on any of the cited pages. Patterns emerge quickly: certain domains recur across many questions in your category, and those recurring cited sources are where presence is worth earning. This beats generic authority-building because it names the exact pages doing the deciding. The [third-party proof article](/resources/getting-cited-by-ai/third-party-proof) covers how to act on those source lists legitimately. ## How CiteAgentic maps prompts to cited URLs and competitors **Product example.** CiteAgentic tracks your buyer questions across engines and regions and keeps the evidence at question level: for each tracked question it records which brands were mentioned or recommended and which URLs were cited, classifies each cited source, and shows where competitors appear in sources you are absent from. Recommendations are then tied to the specific questions and cited sources behind them, so "earn presence on this source" always names the real source and the questions it decides, rather than generic advice. **See which sources decide your questions** Track your real buyer questions and see exactly which pages each engine cites, and where competitors appear instead of you. Explore AI Visibility → ### Frequently Asked Questions **Q: Is there a ranking factor list for AI citations?** No credible one. The platforms do not publish citation algorithms, and their public guidance points to ordinary indexing, relevance and content quality. Anyone selling a definitive factor list is extrapolating beyond what the vendors have said. **Q: Do AI engines prefer big, high-authority sites?** Recurring cited sources in a category often include large review platforms and editorial sites, but engines also cite small, specific pages that directly answer a fanned-out query. Relevance to the specific sub-question is the lever you control. **Q: Can I see the fan-out queries an engine used?** Not directly. You can infer them by studying which sources were cited for a question and what those pages answer. Repeated observation across many questions reveals the sub-topics an engine associates with your category. --- # Surfer SEO Alternatives: Choose by Workflow > Compare Surfer SEO alternatives by content editing, AI visibility monitoring and audit-to-fix workflows. A dated, practical guide with official sources. Source: https://www.citeagentic.com/resources/getting-cited-by-ai/surfer-seo-alternatives _Tool evaluation_ **TL;DR:** Surfer now offers content optimisation and AI visibility features. The useful question is which workflow you need to improve. CiteAgentic is worth evaluating when you want buyer-question monitoring, SEO/AEO diagnostics and a queue of actions in the same product. It is not presented here as a like-for-like replacement for every Surfer editor feature. ## Start with what Surfer currently offers **Reviewed 28 September 2026.** Surfer describes a [Content Editor](https://docs.surferseo.com/en/articles/5700347-content-editor-overview), an [AI Tracker](https://docs.surferseo.com/en/articles/11667466-ai-tracker) and a page [Audit workflow](https://docs.surferseo.com/en/articles/5700351-getting-started-with-audit). It would be inaccurate to describe it as only a Google content-scoring tool with no AI tracking. Capabilities and allowances change. Check the current official documentation, pricing and a real trial workflow before making a purchasing decision. This guide is published by CiteAgentic, a product you may be comparing, rather than an independent review site. ## Compare the job, not just the feature label | What you need | What to evaluate | Where CiteAgentic fits | | --- | --- | --- | | Draft and optimise content in an editor | Editor experience, research context, supported languages and publishing handoff | Audit findings and content guidance can inform edits; do not assume editor feature parity | | Monitor brand mentions in AI answers | Relevant prompts, engine coverage, citations, history and actual response evidence | [AI visibility checking](/features/aeo-scan) and [AEO tracking](/features/aeo-tracker) for selected questions | | Inspect the sources behind answers | Returned URLs, owned versus competitor sources and useful comparison filters | [AI citation tracking](/features/ai-citation-tracking), subject to engine citation availability | | Turn issues into implementation work | Page scope, prioritisation, fix guidance and a follow-up check | [SEO audit software](/features/seo-scan) and [page content audits](/features/aeo-page-audit) | | Coordinate multiple websites | Site allowances, separation of client evidence, access and reporting | Compare current [CiteAgentic plan allowances](/pricing) with your requirements | ## Other monitoring options to include in a shortlist If monitoring is the main job, also evaluate products positioned around AI search analytics. Review [Peec AI](https://peec.ai/) and [Otterly](https://otterly.ai/) directly, then compare engine coverage, prompts, sources and workflow using your own requirements. Our [Peec comparison](/compare/citeagentic-vs-peec) and [Otterly comparison](/compare/citeagentic-vs-otterly) explain the questions to test alongside CiteAgentic. For a broader evaluation, use the [AI SEO tools comparison hub](/compare). To connect your tool choice to content and evidence improvements, follow the [getting cited by AI guide](/resources/getting-cited-by-ai). This page does not claim that a competitor lacks a feature because it was absent from a summary or an older review. ## A practical evaluation you can repeat 1. Pick one important page and a small set of real buyer questions. 2. Try the same work in each shortlisted product: review the evidence, identify a useful change and produce the implementation brief. 3. Record what required another tool, manual research or an upgrade. Check the current allowances and total cost for your intended workload. 4. Make one reviewed change, then compare later scans or observations on the same question set. A short trial can validate usability; it cannot establish long-term ranking or revenue impact. **Illustrative decision:** a team happy with its existing content editor may keep it and add monitoring. A team losing time turning audit reports into tasks may prefer a connected audit-and-fix workflow. Neither choice requires assuming one tool is better at every job. ## When CiteAgentic is worth testing Choose a trial when your priority is connecting sampled AI answers to page findings and a practical next action. Begin with the [free SEO audit](/tools/free-seo-audit) or [AEO checker](/tools/will-ai-recommend-this-page), then assess account monitoring if the workflow fits. Engine access, scan depth and usage depend on the plan. **Test the workflow on a page you actually need to improve.** Use the output to decide whether CiteAgentic reduces your research and implementation effort. Review plan limits before moving ongoing work into the product. --- # 5 things every founder should know about AEO > Five practical AEO decisions for founders, from choosing buyer questions to reviewing evidence and measuring changes. Source: https://www.citeagentic.com/resources/5-things-every-founder-should-know-about-aeo _Resources · founder briefing_ ## Five AEO decisions that make your next marketing hour more useful. Start with a buyer question, inspect the evidence, and choose one useful change. Here is a practical way to move from monitoring to action. **TL;DR:** Choose representative buyer questions, distinguish search access from training permissions, answer questions with real evidence, inspect cited sources, and measure changes consistently. Your first useful outcome is a clear baseline and a reviewed action—not a promised visibility increase. ## 1. Choose questions your buyers actually need answered A visibility percentage only describes the questions and markets in its sample. A high score on your own brand name does not establish that new buyers will find you when comparing options. Start with a small mix of category, comparison and purchase questions. For an AI platform, that might include agency reporting and citation tracking. For a local service business, include the actual service and city. Keep branded and unbranded questions distinguishable. **Action this week:** Write down the buyer, decision and market behind each tracked question. Remove questions unrelated to what you sell. Keep the set stable when comparing scans. See [what AEO visibility measures](/resources/what-is-aeo-visibility). ## 2. Check search access without changing your training preferences A crawler restriction can be intentional. Inspect what it controls before removing it. OpenAI documents OAI-SearchBot for search and GPTBot for potential training separately; permitting one does not require permitting the other. Review the [official crawler documentation](https://developers.openai.com/api/docs/bots) before changing rules. For Google AI features, use the [Search Central requirements](https://developers.google.com/search/docs/appearance/ai-features). Eligibility is not a promise that a page will be selected. **Action this week:** Review the affected URL, robots rules, indexing directives and server response. Ask your developer to fix an unintended search-access problem while preserving deliberate restrictions. Confirm the technical change on a new audit. ## 3. Make your answer useful before trying to optimise it A buyer should quickly understand what you offer, who it fits and what supports your claims. Put the direct answer near the relevant question, then supply limitations, examples and evidence. There is no universal word count or opening-paragraph recipe that guarantees an AI citation. For a service page, describe the work, required inputs and next step. For a software page, explain the workflow and show a relevant product example. Do not invent customer results, prices or credentials to make the copy sound complete. **Action this week:** Review one important page as a prospective customer. Replace vague claims with information you can substantiate. In Content Studio, review generated drafts and complete missing facts before approval. ## 4. Inspect sources before choosing off-site work A citation identifies a source in a sampled answer. It does not, by itself, show that the source mentions your brand or endorses it. Open the cited page and check the relevant passage before treating it as proof. Repeatedly cited review sites, publications or directories can give you research leads. Choose opportunities that fit your audience and eligibility. Avoid treating a listing or backlink as a guaranteed route into an AI answer. **Action this week:** Inspect sources from a relevant competitor-winning answer. Record what the source actually says and choose one legitimate next step, such as correcting an inaccurate business profile. ## 5. Complete one action and measure what happened A dashboard becomes useful when it leads to a decision. Choose a finding with a clear affected page, evidence, expected benefit and manageable effort. Review the proposed fix, implement it, and check that the underlying issue is resolved. Then compare later visibility observations using the same questions, engines and markets. Keep failed checks separate from captured answers. A change in mentions may be worth investigating, but it does not isolate the effect of your edit from other changes. **Action this week:** Set aside time to finish one finding and document its verification. Track qualified visits and enquiries alongside visibility; do not substitute a visibility score for business results. **Turn one buyer question into your next useful action.** Use CiteAgentic to review visibility evidence, audit findings and content drafts. Check current plan limits and your account's trial eligibility before subscribing. Scan completion and visibility outcomes vary.