Answer engine optimization services, written around the questions your buyers ask

Who are answer engine optimization services for?

AEO suits a business whose buyers ask questions before they get in touch: what a job involves, what it costs, which option fits them and who can do it nearby. Many of those questions now get an answer at the top of the results, in a featured snippet, an AI Overview or a ChatGPT reply. AEO services sit on top of SEO, so they fit a business with SEO running or about to start. The fit changes with the size of the business:

• An established local business: the questions are few and asked often, about price, timing and who serves your area, and plain answers on your service pages can cover most of them.
• A multi-location or regional business: the same questions come up in every area, so the answers have to stay in step from one location to the next.
• A national business: the list runs longer across services and buyer types, and the answers need updating as offers and prices change.

If buyers reach you mostly through referrals and rarely search, AEO isn’t the place to start, and on the 15-minute call we’ll say what is.

What counts as an answer engine?

Any search tool that replies to a question with an answer of its own, usually with links to where the answer came from. Google has two forms. The featured snippet is a box that, in Google’s words, reverses the usual result, showing a passage from one page ahead of its link, and it can also appear inside the “People Also Ask” questions. AI Overviews and AI Mode write a longer answer from several pages and link the ones they drew on. ChatGPT search and Perplexity do the same, and each runs its own search crawler: OAI-SearchBot and PerplexityBot.

No business gets to pick the answer. Asked how to mark a page as a featured snippet, Google’s reply is that you can’t, since its systems decide whether a page would make a good one. The same holds for the AI tools. What you control is whether your pages can be found, read and quoted, and that is the work. Our guide to GEO vs SEO sets AEO, GEO and SEO side by side.

How do answer engines pick the part of a page they use?

Passage by passage. Google says its systems can tell the different topics on one page apart and show the piece that fits the question, so there’s no need to cut pages into small chunks, and no ideal page length. Click a featured snippet and Google takes you to the section it quoted.

One question can also turn into several searches. Google’s AI features use what it calls query fan-out: related searches run at the same time, across subtopics, to build one response. Ask about replacing a roof, and the searches behind the answer might cover permits, materials and how long the job takes, each with its own sources.

So each answer has to make sense on its own, wherever it sits: the question as a heading, in the buyer’s words, the direct answer in the first sentence or two under it, then the detail. One page can answer a question and the few that follow from it. Google’s guide treats separate pages for every wording or fan-out query, made mainly to manipulate its rankings or AI responses, as scaled content abuse under its spam policies.

Which questions should your pages answer first?

Start from a list of what buyers actually ask: the “People Also Ask” questions on your main searches, the searches in your Search Console, the questions your team hears on the phone and by email, and the follow-ups that AI answers raise. A useful list covers four kinds of question:

• Buyers’ problems: “My water heater is leaking. Who should I call?”
• Local questions: the “best in” and “near me” searches for your town.
• Questions that name you or a rival: what buyers ask when they’re comparing.
• Service details: what a job includes, how it’s done and what it needs from the buyer.

Put the questions that come just before a sale ahead of popular ones that rarely lead to work. Then give each question one home on the site and point the other pages that touch it there, so two of your pages never compete for the same answer.

What makes an answer worth quoting?

The plain answer comes first, in text, then something only your business can add. Google’s guide asks for a point of view of your own, and gives a first-hand review as its example, set against a summary that restates what is already out there. It also says content built on common knowledge adds little. For a service business, that usually means:

• Ranges with their conditions: when the honest answer is “it depends”, say on what.
• Your process, step by step: what happens first, what the buyer prepares, what can change the plan.
• The mistakes buyers make: the ones you see on jobs, and how to avoid them.
• Facts that agree: prices, areas and hours the same on every page and in its structured data.

Google adds that AI systems understand synonyms and general meaning, so there’s no need to cover every wording of a question. Write for the buyer, in the buyer’s words.

Does answer engine optimization need FAQ schema or special markup?

Not for Google. Its page on AI features says no AI text files or special schema.org markup are needed to appear in them. Structured data still earns its place, since it helps a page qualify for rich results, and Google asks that it match the visible text.

FAQ markup changed in 2026. Google stopped showing the FAQ rich result on May 7, 2026, and removed the feature’s documentation in June. Markup added only to win those expandable questions in the results no longer earns them. Questions answered in plain text on the page still serve readers, featured snippets and AI answers alike.

What goes in an llms.txt file?

The proposal’s current version, v2, was last revised on August 10, 2026. The file sits at a site’s root as /llms.txt, or in a folder such as /docs/llms.txt, and follows a fixed order:

• A heading with the site’s name: the only part the proposal requires.
• A short summary in a quote block: what the site is, and what a reader needs to make sense of the rest.
• Notes, if needed: a few paragraphs or lists on how to read the links.
• Lists of links under headings: each one a link with an optional note, plus an “Optional” list a tool can skip when space is short.

The proposal says the file is used most for software documentation, where coding assistants follow it to API references and tutorials, and that the same structure suits a business setting out its structure and policies. We add one to the sites we build.

What can keep answer engines away from your pages?

A few settings, each easy to miss:

• Firewalls and bot rules: a site can welcome a crawler in robots.txt and still turn it away at the firewall. OpenAI and Perplexity both publish the IP ranges their search crawlers use and recommend allowing them, and Perplexity notes that a web application firewall may need an explicit rule. After a robots.txt change, OpenAI says its search systems can take about a day to adjust.
• Visits made on a user’s request: when someone asks ChatGPT or Perplexity a question, it may fetch a page as ChatGPT-User or Perplexity-User, and robots rules may not apply to those visits. Whether a site shows up in ChatGPT’s search answers turns on OAI-SearchBot.
• Snippet rules: nosnippet, data-nosnippet and max-snippet limit what Google shows from a page, and its AI features use pages that can be shown with a snippet. A rule meant for one block of text can reach further than intended.
• The AI control in Search Console: by August 31, 2026, Google had given every site the Search generative AI control, and include is its default. Switched to exclude, it stops a site’s pages from showing in AI Overviews, AI Mode and Discover’s generative AI features, or from grounding their answers. Check that yours is still on include.
• Training crawlers, a separate choice: OpenAI sets GPTBot, which gathers content for training, apart from OAI-SearchBot, and Google-Extended limits training and grounding in some of Google’s other systems while Search follows Googlebot’s rules. Those choices are yours.

For Google’s preview controls in more depth, see our guide to Google AI Overviews and SEO.

How do you measure answer engine optimization?

By counting answers. A fixed list of buyer questions is put to AI tools on a regular schedule, and every answer is scored: does it name you, does it link to your site, and who does it name in your place. The share of answers that name you is the rate, and it means little without the number of answers behind it.

Two tools add numbers of their own. Search Console’s Generative AI performance report counts how often links to your site were shown in Google’s AI Overviews and AI Mode, and Ahrefs counts the links to your site in AI answers, platform by platform. Neither counts an answer that names you without a link, which is why the question list matters. Our AI visibility audit shows where answers put you today.

How do you judge an AEO agency’s advice?

Google’s guidance on third-party SEO advice names AEO and GEO outright, and gives two marks of good advice: it presents itself as opinion drawn from data or experience, or it cites Google’s official guidance. Its hiring guide, Do you need an SEO?, adds questions to put to any agency. Taken together:

• Is each claim labelled? Ask whether a recommendation rests on the agency’s own data, its experience or Google’s documentation, and which page.
• Does it claim Google’s approval? Google says it doesn’t evaluate third-party services, so treat “approved by Google” as a warning sign.
• Will you see every change? Ask whether the agency will share each change to your site, with the reasoning behind it.
• What sits behind a forecast? Google says third-party tools don’t have access to its internal ranking data, and that their predictions are their own and may not happen.

Put the same questions to us on the call.

What would answer engine work look like for your business?

For an established local business, what we would build is a service page for each main job that opens with the answer to the question buyers ask most about it, and articles on what moves the price and on the other questions that come up before a call. For a multi-location or regional business, we would add a page for each area where the searches and local proof back one, with the same answers and facts on each. For a national business, we would work through a longer question list by service and by buyer type, and keep the answers current as offers change.

Every version starts with a 15-minute call with Nicolas Perez or Noah Alvarez, our co-founders. From our office in Valencia, we work with businesses across the US. To find out which of your buyers’ questions are worth answering first, book a 15-minute call.