The AI-search playbook,written down as we learn it.

The practical definition: the share of AI answers that mention you. Why that number beats the alternatives, and the four signals that move it.

Why AEO isn't 'SEO with extra steps'. The signals are different, the playbook is different, the metrics are different.

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.

Five claims, each with a paragraph of expansion, a practical example, and an action you can take this week.

The complete guide: crawler access, source selection, extractable content, entity clarity, independent proof, and measuring whether a fix changed the answer.

The prioritised implementation checklist, in pipeline order: access, content, entity, independent proof, measurement. Each step with a way to verify it.

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.

What public documentation supports about source selection: query fan-out, relevance to sub-questions, and why competitors get cited instead of you.

The page formula: direct answer, evidence, example, qualification, next step. Plus honest structured data and how to verify the page after publishing.

Consistent naming, About pages, Organization schema, sameAs, Business Profile and directories: making every record tell the same story.

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.

Baselines, mentioned vs recommended vs cited, prompt-level tracking, and a before-and-after test design that survives AI answer variance.

Solid Google traffic, invisible in AI answers. The six verifiable failure classes behind the gap, and the order to test them in.

Backlinks and AI citations answer different questions. What each graph measures, when you need each, and what only the citation graph can tell you.

Tools built for Google rankings score the wrong target for AI answers. The honest tool landscape, split by what you're actually optimising for.

How the visibility, sentiment, share-of-voice and citation scores are computed, and the guardrails that keep small samples from lying. Nothing hidden.

vs Otterly

Both track your visibility in AI answers. Only one writes the fixes and proves they worked.

vs AIclicks

They cover the most AI platforms. We tie every fix to a question you are losing.

vs Peec AI

Their strength is analytics. Ours is doing the fixes and proving they worked.