GUIDE / CHATGPT
How to get your local business recommended by ChatGPT
The straight answer
ChatGPT recommends a local business when it can find the business, understand exactly what it does and where, and see other sources agree. Being crawlable by OpenAI’s bots, keeping one consistent entity across your site, your Google Business Profile and the directories, being present on the sources it cites, and publishing specific, quotable facts are what move the answer — and asking it the same question every month is how you know.
Key takeaways
- ChatGPT with search on retrieves live web pages and cites them; without search it answers from what it learned in training. Both reward a business whose record is consistent and specific.
- Let it in: OAI-SearchBot fetches pages for search answers, GPTBot for training, ChatGPT-User for pages a user asks it to open. Blocking them removes you from the answer.
- One entity everywhere. Name, address, phone, services, service area — identical on the site, the Business Profile and every listing.
- The sources ChatGPT cites for your trade in your city are a list of where you need to be listed.
- Ask it the same questions monthly and record whether you are named. It is the only measurement that matters.
How ChatGPT answers “who is the best roofer in Franklin”
It depends on whether search is in play. With search — which ChatGPT now uses on its own for questions that need current information — it runs web searches, reads a handful of pages and map-style listings, writes an answer and shows citations. For many local questions it presents businesses with their ratings and a map. Without search, it answers from what it learned in training, which is older and thinner, and it tends to name the businesses that were most consistently described across the web when the model was trained.
Either way the winning business is the one the web describes clearly and consistently. The difference is that with search, changes you make show up as soon as its crawler has read them; without search, they show up in the next model.
Step 1: let it in
OpenAI uses several crawlers, and they are named in your robots.txt or they are not. OAI-SearchBot fetches pages to build search answers; GPTBot gathers pages for training; ChatGPT-User fetches a page when a person asks ChatGPT to open it. Many sites block all bots they do not recognize, or run a security plugin that does, and are then surprised to be absent from the answers. Allow the three by name, keep the private paths disallowed, and check that the page renders its text in the HTML rather than painting it in with a script the bot will not run.
The free scan reports whether a site’s robots.txt admits the AI agents and whether the important content is in the HTML.
Step 2: one entity
Pick the exact business name, the one description, the list of services and the service area, and repeat them identically: the site, the Google Business Profile, Apple and Bing maps, the directories, the review sites, the association pages. Then say the same thing in structured data — Organization or LocalBusiness with the same name, address, phone, area and a sameAs list pointing at your profiles — so a machine can connect the site to the listings without guessing.
The failure mode is common and quiet: “Smith Roofing LLC” on the site, “Smith Roofing & Construction” on the profile, an old phone number on two directories. To a model that is three businesses, and none of them is trusted enough to recommend.
Step 3: be on the sources it cites
Ask ChatGPT for your trade in your city and look at the citations. They are usually a mix of the Business Profile and map data, review sites, general directories, a trade-specific directory or two, a local news or community site, and one or two business websites. That list is your listings plan: every source it cited that you are not on is a place to be, with the same facts as everywhere else.
The free AI answer check reports the sources Google’s AI Mode drew on for the same question; the full report we prepare records them for five engines.
Step 4: publish what it can quote
A generative engine cites the page that gave it a fact. Give it facts: a page per service that says what the job involves, how long it takes, what it includes and what it does not; a page per city you want work from that says you work there and shows it; an about page that says who you are, how long you have done this, what you are licensed for. FAQ blocks with the questions you get on the phone. An llms.txt file that states the whole entity on one page. Vague pages are not cited because there is nothing in them to cite.
Step 5: reviews it can read
ChatGPT reads reviews the way it reads everything else — as text. A steady flow of Google reviews that describe the work and the town, answered by you, is evidence of exactly the kind it is looking for. Ask every customer; reply to every review; never pay for or fabricate one, which violates every platform’s terms and, when detected, removes the trust the rest of the work built.
Step 6: ask it, monthly
Write down five or six questions your customers would ask — “best roofing contractor in Franklin, TN,” “who should I call for a roof leak in Brentwood,” “roof replacement companies near Spring Hill with good reviews” — and ask ChatGPT each of them on the first of the month. Record whether you were named, whether your site was cited, and who else was. Do the same in Gemini, Perplexity, Copilot and Google’s AI Mode. That table, month over month, is the only honest measure of whether the work is working.
This is what the AI SEO program does for a client every month, and what the report shows beside the leads.
What not to do
- Do not write “the best roofer in Franklin” about yourself on every page. Engines discount self-praise and look for what others say.
- Do not buy reviews, listings or links. Corroboration only counts when it is independent.
- Do not put hidden text or instructions for the model on your pages. Serious engines strip or penalize it.
- Do not build a page for every town with the name swapped. A model reads them as one page and as a reason to trust the site less.
- Do not block the crawlers and then wonder why you are absent.
Where this sits in the work
The layer that does this.
FAQ / Answers
How to get your local business recommended by ChatGPT — questions
Does ChatGPT use Google to find local businesses?
No. ChatGPT’s search draws on its own web index and partner data rather than Google’s results, and it reads map and profile data for local questions. That is why a business can rank on Google and still be missing from ChatGPT: the record it reads is assembled differently, and it rewards consistency across the whole web rather than a Google position.
How do I know if ChatGPT can read my website?
Check your robots.txt for rules that block OAI-SearchBot, GPTBot or ChatGPT-User, or a blanket rule that blocks all unrecognized bots. Then check that the page’s text is in the HTML rather than loaded by a script. The free scan at nashvilleswebdesign.com reports both.
How long until ChatGPT reflects changes to my site?
With search in play, as soon as its crawler has re-read the page and the sources it cites have updated, which is typically weeks rather than days. Answers given without search come from the model’s training and change only when the model does. The monthly check is how you see it move.
Can I pay to be recommended by ChatGPT?
No. There is no placement for sale in its answers. The businesses it recommends are the ones its sources describe most clearly and consistently, which is work rather than a purchase.
Read next
Related guides.
Guide / GEO
Generative engine optimization (GEO): how a business gets cited in AI answers
What generative engine optimization is, how ChatGPT, Gemini, Perplexity and AI Mode build and cite an answer, what gets a business named, how to measure it.
Read the guideGuide / Google AI
How a local business shows up in Google AI Overviews and AI Mode
What Google AI Overviews and AI Mode are, when they appear for local searches, what Google says helps, the local signals they read, how to check for your name.
Read the guideGuide / llms.txt
llms.txt: what it is, what it is not, and whether your business needs one
A plain guide to llms.txt, the proposed text file that tells language models what a site is: the format, what it does not do, why we ship one, how to write one.
Read the guideThree free checks / One a month each
See what the engines say about you — in a minute.
The same checks we run before quoting a job. Each is free, asks for no email, and the result is yours whether you hire us or not.
Free / Scan your website
85 checks on your site.
Technical SEO, speed on a phone, structured data, AI search readiness and whether the page turns a visitor into a call — every finding with its fix, a PDF and its own link.
Scan a siteFree / Ask Google’s AI
Does the AI name you?
Google’s AI Mode asked for the best business of your kind in your city, live, and read for your name — then the map at your own address and your profile against the pack.
Run the checkFree / Check your links
How many of your links fail the rules?
Every site linking to yours read by the backlink audit’s written rules — how many fail outright, how many are borderline, the reasons, five examples, and a PDF.
Check the linksNEXT STEP
Find out what the AI engines say about you — and whether your market is still open.
Tell us your trade and your city. Before we call, we check what the engines answer and who takes the calls today — so the first call is a straight answer on whether your market still has room.