SEO and AEO playbooks

Ten plays for getting found by search engines and quoted by answer engines. The first five are table stakes. The last five are where the difference is.

Unblock the crawlers

Trigger: You have never checked your robots.txt against the AI user agents.
Move: Allow GPTBot, ClaudeBot, PerplexityBot, Google-Extended and OAI-SearchBot explicitly.
Why it works: A blocked crawler describes you from whatever it finds elsewhere. Usually a competitor's comparison page.

Low effort

Answer in the first paragraph

Trigger: Any page that opens with context before the answer.
Move: Put the complete answer in the first 60 words, then explain.
Why it works: Models lift self-contained paragraphs. A paragraph that depends on the one after it cannot be lifted.

Low effort

Headings as questions

Trigger: Any page with headings like Overview, Features, Benefits.
Move: Rewrite them as the question a buyer would type, then answer directly underneath.
Why it works: Question headings map onto the query. Noun headings do not.

Low effort

Ship FAQPage schema

Trigger: Any page with a genuine FAQ section on it.
Move: Add FAQPage JSON-LD that matches the visible questions exactly.
Why it works: It is the block answer engines quote most often, and the easiest one to get right.

Low effort

Date everything

Trigger: Undated pages, which is most pages.
Move: Put a visible updated date on the page and a dateModified in the schema.
Why it works: Undated pages get treated as stale regardless of how current they are.

Low effort

Own the comparison queries

Trigger: Buyers compare you to two or three named competitors.
Move: One page per competitor, honest, with a table. Include where they win.
Why it works: Comparison queries are the highest intent thing anyone types, and the pages that answer them get read by models.

Medium effort

Get onto the pages that decide

Trigger: The same handful of Reddit threads, review sites and roundups get cited for your category.
Move: Find them, then get accurately represented on each one.
Why it works: Models read a small number of sources per answer. Your own site is rarely one of them.

High effort

Name your entities

Trigger: Pages that describe a category without naming anything in it.
Move: Name the tools, standards, people and organisations a knowledgeable person would name.
Why it works: Entity coverage is how a model decides whether a page is written by someone who knows the field.

Medium effort

Publish the number nobody has

Trigger: You sit on usage data or can survey your customers.
Move: Publish it with the method, in a table, under a question heading.
Why it works: Original data is the only content type that reliably gets cited rather than paraphrased.

High effort

Measure the answer, not the ranking

Trigger: You are still reporting positions and impressions only.
Move: Track the questions buyers ask and whether you are named in the answer.
Why it works: The ranking and the answer have come apart. Only one of them is what the buyer sees.

Medium effort

Run the first five once. Run the last five continuously. The first five stop you losing; the last five are how you win.

Unblock the crawlers

Answer in the first paragraph

Headings as questions

Ship FAQPage schema

Date everything

Own the comparison queries

Get onto the pages that decide

Name your entities

Publish the number nobody has

Measure the answer, not the ranking

Every other tool tells you you’re losing.