What AEO means for a local business

Somebody in your town is going to ask ChatGPT, Perplexity or Gemini who to hire for the thing you do. They will not get a page of ten links. They will get a paragraph that names two or three businesses and gives a reason for each. Answer engine optimization — AEO — is the work of making it more likely that yours is one of the names.

You have probably already been pitched on it. This guide is the version without the urgency: what it actually is, how the engines seem to pick, what is within your control, and what is not.

What changes when the answer is one paragraph

Traditional search gave you a ranking. If you were fifth, you were still on the page, and a buyer who scrolled could still find you. An assistant's answer has no fifth place. You are either named or you are not, and there is no scroll.

The second change is harder to see. When a buyer clicks through from Google, you can find the visit in your analytics. When an assistant recommends your competitor instead of you, nothing in your reporting records it. There is no impression, no lost click, no referrer. The only way to know what the engines say about your category is to ask them.

How an engine decides who to name

Nobody outside the companies that build these systems can see exactly how an answer is assembled, and anyone who tells you otherwise is guessing with confidence. What can be observed is what the answers draw on. Engines that search the web while they answer, such as Perplexity and ChatGPT with search, often show their sources, and the same kinds of source come up again and again for local questions:

• Directories and review platforms — Google Business Profile, Yelp, the Better Business Bureau, and the directories specific to your trade.
• Lists and roundups — “best of” articles from local publications, city guides and industry sites.
• Local news — coverage of a business, a community project, an award, an owner quoted as an expert.
• Forums and community sites — Reddit threads and neighborhood forums where people ask for recommendations and other people answer them.
• Your own website — but mostly when it states plainly what you do, where you do it, and for whom.

The pattern is that an engine is more comfortable recommending a business that many independent sources agree exists, does a specific thing, in a specific place, and is well regarded. It is less a single ranking factor than a weight of corroboration.

The distinction that matters: asking with and without your name

The first thing most owners do is ask an assistant about themselves by name. The answer is usually flattering, and it tells you almost nothing. A question that contains your name will nearly always produce your name back.

The question that matters is the one a stranger asks: “best roofer in Dayton”, “who should I call about a leaking water heater”, “family law attorney near me who handles custody”. None of those contain your name. The share of answers to questions like that in which you are named is your organic mention rate, and it is the number that says whether AI is sending you new customers.

Questions that do use your name are still worth asking, for a different reason: they show whether the engines describe you correctly — right services, right area, right phone number. That is accuracy, not visibility. Keep the two apart. Blend them and you get a number that looks healthy precisely because it includes the questions you were always going to win.

What you can actually do

Most of what helps is not new. It is the unglamorous local-marketing work that was always worth doing, now with a second reason to do it.

1. Make your name, address and phone number identical everywhere. Your website, Google Business Profile, Yelp, Facebook, the BBB and every trade directory should say the same thing in the same way. Conflicting details give an engine a reason to hesitate.
2. Finish your Google Business Profile. Correct categories, a service area that matches reality, hours, services listed individually, photos, and answers to the common questions people ask there.
3. Keep reviews coming. Volume matters, and so does recency: a steady trickle of recent reviews says you are still operating and still good. Ask every satisfied customer, and reply to reviews, including the bad ones.
4. Get listed where your category is discussed. Find the directories, roundups and local articles that already come up when you ask about your trade, and work out how to be on them — a listing, a correction, a pitch to the writer.
5. Write service pages that answer the question. One page per core service, saying what you do, where, for whom, and roughly what it involves. “We provide quality solutions” helps no one. “We repair and replace residential water heaters in Dayton, Kettering and Beavercreek” does.
6. Answer buyer questions on your site. An FAQ built from the questions customers actually ask you on the phone — cost, timing, what to check first — in plain sentences an engine can quote.
7. Add LocalBusiness schema markup. Structured data that states your name, address, phone, hours and service area in a form machines read without guessing.
8. Check you are not blocking AI crawlers. Some site builders and security plugins block crawlers such as GPTBot, OAI-SearchBot, PerplexityBot and ClaudeBot by default. Look at your robots.txt, or ask whoever runs your site.

Where it is genuine, community participation helps too: answering questions in a local subreddit or neighborhood group under your own name, as someone who knows the trade. Thinly disguised advertising tends to be removed, and it does not read well when it is not.

What you cannot do

You cannot control what an engine says. Nobody can. There is no ad slot to buy inside an organic answer, no form to submit, and no one who can guarantee you a place in one. The same question can produce a different answer tomorrow, a different answer on a different engine, and a different answer for someone asking from the next town over.

That is not a reason to ignore it. It is a reason to be suspicious of anyone promising a position, and to judge the work by measurement rather than by promises.

How to tell whether it is working

Because answers vary from run to run, a single check tells you very little. What tells you something is the same set of buyer questions, asked the same way, across several engines, repeated on a schedule — and a record of which answers named you, who they named instead, and which sources they drew on. Over months, that record shows whether the work is moving anything, and which sources are worth your time.

Two more things are worth tracking alongside it. The first is accuracy: when you are named, is the description right? A wrong phone number or a service you stopped offering costs you the customer the recommendation was supposed to bring. The second is the difference between “not mentioned” and “not tested”. If an engine errored or could not be reached, that is a gap in the measurement, not a miss on your record, and it should never be counted as one.

You can do a version of this by hand, and our guide to checking whether ChatGPT recommends your business walks through how. It is worth an afternoon even if you never pay anyone for anything.

Find out where you stand

If you would rather see a real result first, the free audit asks eleven buyer questions about your category — nine that never use your name and two that do — across two AI engines, and emails you what came back: whether you were named, who was named instead, and the sources behind the answers. No account and no card.