Getting cited by ChatGPT and Google AI: a guide for local businesses
AI assistants recommend local businesses based on five things: a consistent identity across the web, third-party validation from directories and reviews, pages that answer specific questions directly, structured data that tells the machine what your business is, and crawler access so the AI can read any of it. None of these require tricks. They require your business information to be clear, consistent, and findable in the places AI systems already look. Most local businesses get the first four partly right and the fifth wrong, and the result is that AI describes them from stale third-party data or not at all.
We learned this the hard way. Earlier this year we ran 30 customer-style queries across ChatGPT, Gemini, Perplexity, Copilot, and Google AI Overviews. Our agency, in business since 1986 with 80 Google reviews, appeared zero times. The businesses that did appear were not better than us. They were easier for a machine to understand. This guide is what we did about it, written for any local business that wants the same result.
How AI assistants pick the businesses they name
When someone asks ChatGPT “who is a good family dentist in Rockville,” the assistant does not consult a ranking. It searches the live web, reads a handful of pages, and composes an answer from what it can verify. The businesses it names are the ones that appear in multiple trustworthy sources with consistent details.
Three things are different from classic Google search:
There is no page two. An AI answer names three to six businesses. Everyone else is invisible for that question.
The sources are narrower. For most local-business queries, AI assistants cite directories (Yelp, Clutch, Healthgrades, Avvo, and industry equivalents), Google Business Profile data, review sites, and, increasingly, the businesses’ own pages when those pages are specific enough. Your homepage is rarely one of them.
Accuracy beats authority. A big site with a vague About page loses to a small site that states plainly what it does, where, for whom, and at what price. AI systems prefer sources they can quote without guessing.
The five things AI looks for
1. A consistent identity everywhere
Your business name, address (or service area), phone number, website, hours, and services should be identical on your website, Google Business Profile, every directory, LinkedIn, Facebook, and the data aggregators that feed all of them. When these disagree, AI assistants either pick one at random or decide you are unreliable.
This is where most established businesses fail, usually because of history. Old offices, old phone numbers, a name change, a merger. We found two data aggregators still publishing office addresses we had closed years earlier, and Copilot was repeating them. If your business has moved or changed in the last decade, assume something is wrong somewhere.
What to do: list every place your business appears online, correct each one, and fix the aggregators (Data Axle, Foursquare, and Neustar Localeze feed most of the others). If you no longer have a public office, convert your Google Business Profile to a service-area business and remove the street address from every listing rather than leaving it to rot.
2. Third-party validation
AI assistants are cautious about recommending a business on that business’s own say-so. They look for outside confirmation: reviews, directory listings, press, and industry sites. For agencies, the sources most often cited are Clutch, Expertise.com, and DesignRush. For medical practices, Healthgrades and Zocdoc. For contractors, Houzz, Angi, and the BBB. For restaurants, Yelp and Google. For law firms, Avvo and Justia.
Find yours by running the 30-query audit and recording which sources each answer cites. Then claim, complete, and get reviews on those specific profiles. An unclaimed or half-finished listing is worse than none because it may carry outdated data.
Reviews matter in two ways. Volume and rating are obvious. Less obvious is content: reviews that mention specific services, locations, and outcomes give AI systems language to work with. “Great job” tells the AI nothing. “Rebuilt our WooCommerce site and fixed the checkout problems we had for two years” tells it exactly what you do.
3. Pages that answer specific questions
AI assistants extract answers. They favor pages where the answer to a question sits at the top in plain language, followed by supporting detail. A page titled “Our Services” with three paragraphs of positioning gives them nothing to extract. A page titled “Web design in Frederick, MD” that opens with what you do, who it is for, what it costs, and how long it takes gives them everything.
What works:
- One page per service and one per service area, each opening with a two-to-four-sentence direct answer
- Pricing or price ranges stated in text. “How much does X cost” is among the most common AI queries, and businesses that publish ranges get cited for it almost by default
- Question-form headings that match how people actually ask
- A short FAQ on every substantial page, with answers that make sense on their own
- Specifics over adjectives. Numbers, timelines, process steps, named tools. “Award-winning” is unquotable. “Eight to fourteen weeks from kickoff to launch” is not
This is also where AI search and classic SEO converge. Pages built this way rank better on Google too, because Google’s AI Overviews use the same extraction logic.
4. Structured data
Schema markup is the machine-readable layer that confirms what your pages say. For a local business, five types do the work: LocalBusiness (or its specific subtype), Organization, Service, FAQPage, and Person for the owners and authors. They tell AI systems your name, location or service area, services, the questions you answer, and who stands behind the content.
The critical rule is that schema must match everything else. A phone number in schema that differs from your Google Business Profile is a contradiction, not a signal. We cover implementation in detail in Schema markup for AI search.
5. Crawler access
None of the above matters if AI systems cannot read your site. Check three things:
- robots.txt does not block GPTBot, ClaudeBot, PerplexityBot, Google-Extended, or Bingbot. Some security plugins and CDN settings block these by default.
- Your pages render without JavaScript. Many AI crawlers do not execute scripts. If your content only appears after JavaScript runs, they see an empty page. Test by disabling JavaScript in your browser and loading key pages.
- No noindex headers on files that should be readable. We found our own llms.txt carrying an X-Robots-Tag: noindex header added by a plugin. Correct file, correctly placed, effectively invisible.
While you are there, add an llms.txt file. It is an hour of work, and it doubles as a forcing function for writing a clear one-paragraph description of your business.
A 12-step checklist
Work through these in order. The first six are where most of the improvement comes from.
- Run the 30-query audit across five platforms and record every cited source
- Correct your business information on Google Business Profile, then on every directory and aggregator
- Remove or update every reference to closed locations, old phone numbers, and former names
- Claim and complete the directories the audit showed being cited for your industry
- Ask recent customers for reviews that mention the specific service and location
- Rewrite your About page to state plainly what you do, where, for whom, and since when
- Build or rewrite one page per service with a direct answer at the top and a price range where possible
- Build one page per city or county you serve, each with real local detail, not a find-and-replace
- Add a five-to-eight-question FAQ to every service and location page
- Implement LocalBusiness, Organization, Service, FAQPage, and Person schema, and validate it
- Check robots.txt, JavaScript rendering, and response headers for crawler access
- Re-run the audit quarterly and track your mention rate
What to expect, and when
Fixing listings and getting reviews shows up in AI answers within four to eight weeks, because those sources update quickly. New pages and schema take two to six months, because AI search results depend on the web being re-crawled and re-indexed. A realistic goal for a business starting from zero is to appear in a quarter of relevant queries within six months and half within a year.
Track one number: the percentage of your audit queries that name you. Everything else is a means to that.
Mistakes that set businesses back
Publishing AI-generated filler. Thin pages written by AI to “cover keywords” are easy for AI systems to recognize and discount. Write less, and make it specific.
Hiding FAQs or stuffing schema. Schema that does not match visible content, or reviews in schema that are not on the page, violates guidelines and gets ignored.
Blocking AI crawlers to protect content. Understandable for publishers. Fatal for a business that wants to be recommended.
Treating the homepage as the answer. AI assistants cite the page that answers the question, and that is almost never the homepage.
Doing it once. AI answers shift constantly. A business that audits quarterly and keeps listings current will stay ahead of one that fixed everything in 2026 and stopped.
Frequently asked questions
Is AI search optimization the same as SEO?
It overlaps heavily. The technical foundation, clear content, and business listings serve both. The difference is emphasis: AI search rewards directness, consistency, and third-party validation more than link building and keyword volume.
Can I pay to appear in ChatGPT or Gemini answers?
No. There is no paid placement in AI assistant recommendations as of now. Google AI Overviews may show ads around the answer, but the businesses named in the answer itself are not sponsored.
Do I need a new website to do this?
Usually not. Most of the work is listings, reviews, rewritten pages, and schema, all of which fit on an existing WordPress site. A site that blocks crawlers or cannot render without JavaScript may need structural changes.
How is this different for a business with multiple locations?
Each location needs its own Google Business Profile, its own page, and its own consistent listings. Sharing one set of details across locations confuses AI systems as much as it confuses customers.
Which AI platform matters most?
Google AI Overviews reaches the most people because it appears inside normal Google searches. ChatGPT has the most direct usage. Optimize for both and the others follow, since they draw on overlapping sources.
What if AI is saying something wrong about my business?
Find the source. Ask the assistant where it got the information, or run the query on Perplexity, which shows citations. Correct the source, and the answer will follow within weeks.
We run this process for clients as a service, starting with the audit and ending with a quarterly mention-rate report. See how AI search optimization works, or request a free estimate and talk to the person who would do the work.
