What an answer engine does with your page
An answer engine is any system that responds to a question with a direct reply instead of only a list of links. Google's AI Overviews and AI Mode, ChatGPT search, Perplexity and Microsoft Copilot all work this way to some degree. Each one reads pages, picks passages it considers useful and either quotes them or summarises them.
Google describes its approach in plain terms. AI Overviews and AI Mode show links to supporting pages, and they may run several related searches behind the scenes, which Google calls query fan-out. Google adds that an AI Overview is shown only when the system decides it adds something beyond classic results, which means many queries get none.
That matters for planning. A page that answers one question well can be useful in several places, but you cannot assume any single engine will show it. AEO is therefore a way of removing obstacles, not a way of buying a position.
Where your answers can appear
These are the surfaces we test for most clients. Each has its own crawler and its own reporting, covered in more detail on the related pages.
Google AI features
AI Overviews and AI Mode draw on pages that are indexed and eligible for a snippet in Search, on sites that have not been excluded in Search Console.
ChatGPT search
OpenAI says OAI-SearchBot is the crawler that surfaces sites in ChatGPT search features.
Perplexity
Perplexity's documentation says PerplexityBot is the crawler behind links in its answers.
Bing and Copilot
Bing Webmaster Tools has a public-preview report on how often your content is cited in Copilot and Bing AI summaries.
What we do, step by step
The work is deliberately practical. Each step produces something you can read and check.
- List the questionsWe gather the questions your customers ask before they call, using your enquiries, your staff and the wording people use in search.
- Match questions to pagesEach question is assigned to one page, or to a gap where no page answers it yet.
- Rewrite for answersWe draft or edit so the direct reply sits near the top in plain text, followed by the detail, conditions and limits.
- Check accessWe review robots.txt, meta robots rules and CDN or firewall settings so the right crawlers can read the page.
- Test and recordWe ask the same questions on each platform, record what appears, and repeat on a schedule.
Answer-first writing in practice
Good answer pages read like a knowledgeable person explaining something, not like a list of keywords.
Worth doing
- State the answer in the first one or two sentences under the heading.
- Name the conditions that change the answer, such as licence type, location or timing.
- Keep key facts in text, not only inside images or scripts.
- Write headings that sound like the question a customer would ask.
Not worth doing
- Splitting one topic across dozens of thin pages.
- Padding an answer with a long preamble.
- Stating certainty you do not have.
- Repeating the same answer on every page of the site.
What we check on each page
This is the review checklist we run before and after changes.
| Check | Why it matters | How we test it |
|---|---|---|
| Indexed and snippet-eligible | Google says a page must be both to appear in its AI features | URL Inspection in Search Console and a review of robots directives |
| Answer near the top | Systems can quote a short passage more easily than a buried one | Read the first screen of the page without scrolling |
| Facts match elsewhere | Conflicting details make a source harder to trust and to summarise | Compare the page with listings, profiles and other pages |
| Structured data matches text | Google says markup must match the visible page | Compare markup with the visible content |
| Crawler access | A blocked crawler cannot read the page at all | Review robots.txt and firewall or CDN rules |
What we cannot control
No provider publishes a complete list of what makes a page get chosen. Google's documentation says its AI features need no special schema.org markup or machine-readable file. Its generative AI guide adds that, from Google's view, optimising for AI search is still SEO.
Responses also change between users, locations and days. When we report, we describe what we saw on a given date and how we asked, not what will happen next month.
What to prepare before we start
Answer work goes faster when the questions come from real conversations. These are worth gathering first.
- Your most common enquiry questions: Pull them from emails, call notes and messages, including the awkward ones about price, timing and suitability.
- The pages you already rely on: Note which service and pricing pages bring enquiries today, so we improve those first and leave working pages alone.
- Access to Search Console: Verified access shows which pages are indexed, and the AI performance report shows where Google already displays your links.
- Facts only you can confirm: Licences, service limits, response times and what you do not do must come from you, because we will not guess them.
How this differs from nearby services
AEO sits between several services. On-page SEO covers titles, headings and internal links across the whole site, while AEO concentrates on how individual answers are phrased and placed. SEO content is about producing new pages, whereas AEO often starts by improving pages you already have.
It is also narrower than generative engine optimisation, which looks at how your business is represented across the web, not only on your own pages. The wider overview is on the AI search hub.
Frequently asked questions
What is the difference between AEO and SEO?
SEO is the wider job of helping search engines find, index and rank your pages. AEO is a focus within it: shaping pages so a system can extract a direct, correct answer. Most AEO work is good SEO with extra attention to how the first lines of each section read.
Can you guarantee my business will be quoted by ChatGPT or Google?
No. Providers do not publish a method that guarantees selection, and responses differ between people and days. We can make your pages easy to crawl and quote, then test and report what we observe. Any agency promising citations is promising something it does not control.
Do I need special markup or an AI file to be included?
For Google, no. Its documentation says no additional requirements, special schema.org markup or AI text files are needed for AI Overviews and AI Mode, and Search Console includes every site unless its owner opts out. Structured data still helps with rich results, but it must match what the page shows.
How do you measure whether it is working?
We ask a fixed set of real customer questions on each platform, record whether your pages appear, and repeat on a schedule. We also use the Generative AI performance report that Google rolled out to all sites by 31 August 2026, and the AI Performance report in Bing Webmaster Tools. The results are directional evidence, not a ranking.
Will AEO replace my normal SEO?
No. Google states its AI features are built on core Search systems, so pages that cannot be crawled or indexed will not appear in them. Normal SEO remains the base, and AEO adds careful work on how answers are written.
Which of my pages should be done first?
Start with the pages tied to revenue, such as your main services, pricing explanations and frequently asked questions. We rank them by how often customers ask about them and how weak the current answer is, then work down the list.
How long before I see a result?
It depends on how often your site is crawled and on each platform. OpenAI says a robots.txt change can take about 24 hours to reach ChatGPT search. Neither OpenAI nor Perplexity publishes a timeline for content changes, and Google says recrawling can take days to months. We do not give fixed timelines.
Sources and further reading
- Google Search Central: AI features and your website: Eligibility, query fan-out, no special markup or AI files, and when AI Overviews appear.
- Google: Optimizing for generative AI features on Google Search: Standard SEO applies; AEO and GEO are still SEO from Google's view; structured data is optional for AI features.
- OpenAI: Overview of OpenAI crawlers: OAI-SearchBot surfaces sites in ChatGPT search; about 24 hours for robots.txt changes to reach search.
- Perplexity: Perplexity crawlers: PerplexityBot surfaces and links sites in Perplexity results and is not used for model training; Perplexity-User generally ignores robots.txt.
- Microsoft Bing Blog: Introducing AI Performance in Bing Webmaster Tools: Published 10 February 2026 (page metadata). Public-preview citation reporting for Copilot and Bing AI summaries.
- Google Search Central Blog: Introducing Search Generative AI performance reports in Search Console: Reports on impressions, pages, countries, devices and dates for AI features; rolled out to all sites by 31 August 2026.
- Search Console Help: Search generative AI control: Include is the default; Exclude removes a site from AI Overviews and AI Mode.