An AEO strategy you can run in 90 days

An AEO strategy is a sequenced plan to make your pages easy for answer engines to reach, understand and quote. Run it in 90 days: confirm crawlers can fetch your pages, map the questions buyers ask, rewrite key pages so each section answers first, add structured data that matches what is visible, then measure against a frozen baseline. No new file formats or tricks required.

What is an AEO strategy, and why run it as a sequence?

An AEO strategy is the order in which you fix things so that answer engines such as Google AI Overviews, ChatGPT, Perplexity and Claude can find and lift your content. The order matters because each step depends on the one before it: a perfectly written answer on a page no crawler can fetch is invisible.

Start from what Google itself says. Its guide to optimising for generative AI features defines the terms directly: "AEO" stands for answer engine optimization and "GEO" for generative engine optimization. The same guide states that, from Google Search's perspective, optimising for generative AI search is optimising for the search experience, and thus still SEO. That is the foundation of a sensible geo strategy too: you are not building a parallel website for robots, you are making your existing site easier to retrieve and quote.

Five steps, three months, each with a checkable output. If you want the underlying concepts first, our explainer on what AEO is covers the vocabulary.

Month one: can answer engines reach your pages?

Access comes first, because nothing else counts until crawlers can fetch your pages and read the text. In month one you prove access rather than assume it.

Google's AI optimisation guide sets the bar for its own features: to be eligible for generative AI features, a page must be indexed and eligible to be shown in Google Search with a snippet, fulfilling the Search technical requirements. The guide adds that a site must be included in Search generative AI features in Search Console to be eligible for display. Check that setting before anything else. Snippet controls matter here too: Google's featured snippets documentation explains that the nosnippet rule blocks all snippets for a page, and that text marked with data-nosnippet will not appear in snippets either. A stray nosnippet left on a template quietly removes a page from contention.

For the other answer engines, we test access by sending real requests, not by reading rules. Robots rules describe intent; a firewall or CDN decides what actually happens. Our free AI visibility checker reads robots.txt rules for 11 AI crawlers and requests your homepage once with each crawler's user agent. That is the checker's own request, not the real crawler's, and it covers the homepage only, so treat a pass as a good sign rather than proof, and a failure as something to investigate in your firewall and server logs.

• Confirm key pages are indexed and show a normal snippet in Google.
• Confirm the site is included in Search generative AI features in Search Console.
• Search templates for nosnippet and data-nosnippet that you did not mean to set.
• Request key pages with each AI crawler's user agent and record the HTTP status (200, 403, 429 or 503).
• Check that the main content is in the HTML, not only rendered after scripts run.

How do you map the questions your buyers ask?

Build a question map: a list of the exact questions your buyers type or speak, each assigned to one page that will answer it. This is the planning core of AEO marketing, and it is where most of the thinking in month one should go.

Gather questions from places you already own: sales call notes, support tickets, the queries in Search Console, and the questions prospects send by email. Phrase each one the way a person would ask it.

Then assign each question to a single page. If two pages answer the same question, pick one and point the other to it. If no page answers it, you have found a gap. Track each question, its page, the state of the answer (missing, buried or answer-first) and the date you fixed it; that becomes your measurement plan.

How do you write answer-first passages?

Open every section with a one- or two-sentence direct answer to the question in its heading, then give the detail. Answer engines lift passages, so each section should make sense on its own if it is quoted without the rest of the page.

This is the most visible of the AEO best practices, and it has a familiar precedent. Google's featured snippets page describes featured snippets as boxes where the format of a regular result is reversed, showing the descriptive snippet first, and notes that clicking one takes the user directly to the section of the page that appeared. You cannot mark a page as a featured snippet; Google's systems decide. The same applies to AI answers. You can make a passage easy to quote; you cannot make anyone quote it.

A working pattern for each priority page in month two:

1. Rewrite the H2s as the questions from your map, word for word where it reads naturally.
2. Put the direct answer in the first sentence under each H2, with the key term in it.
3. Follow with the mechanism, the steps or the comparison, in plain language.
4. Use a table or a numbered list where the answer is genuinely a comparison or a procedure.
5. Add a short FAQ for the follow-up questions that do not deserve their own section.
6. Check the page has exactly one H1 and that headings run in order.

Write for people, not for a word count. Google's guidance on helpful, people-first content asks whether you are writing to a particular word count because you have heard Google prefers one, and answers plainly: no, it does not. Your own data, process and examples are the passages worth quoting.

Should your AEO strategy include structured data and llms.txt?

Add structured data that matches what is visible on the page, because it helps Google understand the page and supports rich results. Do not expect it, or any special file, to unlock AI answers on its own.

Google is direct about this. Its AI optimisation guide lists overfocusing on structured data as a mistake: structured data is not required for generative AI search and there is no special schema.org markup to add, but it remains a good idea as part of your overall SEO strategy. The same guide says you do not need new machine-readable files, AI text files or Markdown to appear in Google Search, and that maintaining an llms.txt file for other services will neither harm nor help your visibility in Google Search, because Google Search ignores it.

When you do add markup, follow the rules in Google's intro to structured data: the markup describes the content of the page it sits on, you should not add structured data about information that is not visible to the user, and you must include all required properties to be eligible for enhanced display. Google recommends using the format that is easiest to implement and maintain, which in most cases is JSON-LD.

How do you measure an AEO strategy without invented numbers?

Freeze a baseline before you change anything, change one layer at a time, and re-measure with the same method. Month three is for measuring, fixing what broke and deciding what to repeat.

Measure what you control directly: how many priority pages return a 200 to each crawler you care about, how many have exactly one H1, how many sections in your question map are now answer-first, and how many pages carry valid structured data that matches the visible text. Then watch the outcome signals in Search Console for the pages you rewrote, and log any AI answers that cite you, with the date and the question asked. Answer engines choose what they quote, so do not promise a citation rate.

What we have seen

On one SEO Autopilot engagement, a US healthcare analytics company on WordPress behind a firewall, we froze the baseline on 15 July 2026 and re-measured on 18-19 August 2026 by fetching all 39 sitemap URLs. AI Visibility moved from 18.35 to 66.35 and SEO Health from 46.55 to 67.45. Schema rose from 5 to 78, AI crawler access from 45 to 82 and llms.txt from 0 to 85.

The access step mattered most. The firewall returned 403 to OpenAI's and Anthropic's search crawlers on every page, triggered by a generic bad-bot rule matching the substring "searchbot", while eleven other crawlers passed. The client's monitoring tool reported all sixteen AI crawlers as accessible because it read robots rules instead of sending requests. On structure, 21 of 39 pages had no H1 because the post template rendered the title as an H2; one template change took pages with exactly one H1 from 9 of 39 to 30.

Not everything improved. Performance fell from 48 to 28, because removing the firewall rule took the CDN edge cache with it (median time-to-first-byte 0.42s to 2.39s). AI review signals did not move, 18 to 18. That is why measurement sits inside the plan, not after it. Our GEO and AEO audit follows the same sequence and scoring, and the earlier post on how to optimise for AI search goes deeper on the access and content layers.

Frequently Asked Questions