A GEO strategy has seven steps: choose the buyer questions you want to win, baseline every AI engine your buyers use, make your site readable by AI crawlers, write answer-first pages for the gaps, earn mentions on the sources engines cite, correct what the engines get wrong, and measure share of voice every month.
Generative engine optimisation (GEO), also called answer engine optimisation (AEO), is the work of being named in AI answers rather than ranked in a list of links. If the terms are new, start with AEO vs SEO vs GEO. This playbook assumes you have the SEO basics in place and want a plan for 2026.
Why does GEO need its own strategy?
Three things make AI answers different from search results. First, an answer names a few brands or none, so being well known is not the same as being recommended: in our own count of 2,545 AI answers, engines named our company in 377 of 379 answers when the question included its name, and in 4 of 2,166 when it did not. Second, engines behave differently: in the same sample Perplexity listed sources in every answer, while Claude did in 246 of 625. Third, Google says its AI features may use query fan-out, issuing several related searches to build one answer, so a single page can be pulled in for questions it was never written for.
The seven-step GEO playbook
Step 1. Choose the buyer questions you want to win
List 25 to 100 questions buyers ask an AI assistant before they choose a product like yours, grouped by stage: learning about the problem, comparing options and deciding.
Think in questions, not keywords: “which CRM suits a five-person agency on a small budget” rather than “crm pricing”. Take them from sales calls, support tickets, search data and forums. The free AI prompt finder builds a starting list from Google search data, and our guide on which prompts to track covers how many and which kinds.
Step 2. Baseline every engine your buyers use
Ask each question in each engine and record whether you are named, your position, which competitors are named and which sources are cited, so every later change has a starting point.
Record the answers before you change anything, per engine. A brand can lead in one engine and be missing from another, so a single blended score hides the gaps. The guide to covering several engines at once explains how to keep the comparison fair.
Step 3. Make your site readable by AI crawlers
Check that robots.txt and your CDN let search and AI crawlers fetch your pages, that the main content is in the HTML, and that structured data matches the visible text.
An engine cannot cite a page it cannot fetch. Check every rule in robots.txt and any bot protection at your CDN; the free AI crawler checker shows which bots each rule lets through, and the AI SEO checker runs the wider set of checks.
Step 4. Write answer-first pages for the gaps
For each question where competitors are named and you are not, publish or rewrite a page that answers it in the first lines, then gives the detail, facts and dates an engine can cite.
Lead with a two- or three-sentence answer that stands on its own, then the evidence: tables, prices, steps, dates and named sources. Comparison pages, “best for” pages and clear pricing pages match many of the questions buyers ask before they choose. Our content strategy for AI search maps question types to page formats.
Step 5. Earn mentions on the sources engines cite
Find the review sites, lists, forums and publications the engines cite for your questions, and work to be listed on them and described accurately.
Look at which domains the engines cite for your questions. They are often review platforms, industry lists, forums and comparison articles. Being on those pages, described accurately, can matter as much as your own site. See how AI engines choose which sources to cite.
Step 6. Correct what the engines get wrong
Where answers misstate your prices, features or category, fix the source pages they draw on and publish a clear, dated statement of the facts on your own site.
Wrong prices, a retired product or the wrong category tend to come from a source the engine trusts. Find it, fix it where you can, and state the facts plainly on your own site with a date. The guide on what to do when AI gets your brand wrong walks through it.
Step 7. Measure monthly and report share of voice
Re-run the same questions on a schedule and report visibility, position, share of voice and citations per engine, alongside the visits that arrive from AI assistants.
Re-run the same questions on a schedule. Report visibility, position, share of voice against named competitors and the sources cited, per engine, with the visits that arrive from AI assistants as a supporting signal. Our guide to reporting AEO to a CFO shows how to present it.
What does a GEO strategy need in place?
| Area | What to have |
|---|---|
| Questions | A fixed, written list of buyer questions, grouped by stage and product line |
| Engines | The engines your buyers use, each measured separately |
| Site | Pages AI crawlers can fetch, answer-first copy, accurate structured data, visible dates |
| Sources | A list of the domains engines cite for your questions, and your status on each |
| Owner | One person responsible for the questions, the reporting and the fixes |
| Cadence | A monthly review of the answers and a quarterly review of the question list |
For larger teams, the guide to running AEO across teams and regions covers ownership and standards. For a shorter start, the 30-day AEO checklist condenses the first month.
Where Noma fits
Noma by Nurdd runs steps 2 and 7 for you, and finds the gaps for steps 4 and 5. It asks ChatGPT, Gemini, Perplexity and Claude your buyer questions every day, across 13 AI engines in all, and shows where you are named, where competitors are named instead and which sources each answer cites. Run a free brand audit to see your starting point.
See what AI engines say about your brand
Free · no account · or start a 3-day trial
Questions people ask
What is a GEO strategy?
A GEO (generative engine optimisation) strategy is a plan for getting your brand named, described accurately and cited in the answers AI engines such as ChatGPT, Gemini, Perplexity and Google AI Overviews write. It sets the buyer questions to win, the engines to cover, the pages and third-party mentions to work on, and how results are measured.
How long does a GEO strategy take to show results?
There is no fixed timeline. Engines that search the web while answering can reflect a change to a page they cite within weeks, while what a model learned in training changes on a much slower cycle. Judge the work on the trend across your tracked questions over several months, not on one day's answers.
Which AI engines should a GEO strategy cover?
Cover the engines your buyers use. That often means ChatGPT, Google AI Overviews and AI Mode, Gemini, Perplexity and Claude, and possibly Microsoft Copilot or Grok. Measure each separately, because engines often name different brands for the same question.
Is a GEO strategy different from an SEO strategy?
It builds on one. Google says SEO best practices apply to its AI features, so crawlable pages and clear answers serve both. GEO adds questions instead of keywords as the unit of work, per-engine measurement of answers, and more effort on the third-party pages engines cite.