You can check whether ChatGPT can see your business in about ten minutes. Run five recommendation prompts across ChatGPT, Perplexity, and Gemini, then read whether you show up invisible, wrong, or cited. No login, no paid tools, no agency required. Here is the exact audit, written to copy.
Updated August 2026.
I run Answer Engine Optimization for contractor and service-business clients at Current Digital Co, and the first thing I do on any new account is exactly this audit. Not a report full of scores, not a dashboard. I open ChatGPT and ask it to recommend a business in the client's trade and city, and I watch what it says. Nine times out of ten the client had never done it, and nine times out of ten the answer is worse than they expected.
The reason this matters is not theoretical. A 2026 study of roughly 350,000 business locations (SOCi's Local Visibility Index, reported by Search Engine Land) found that AI assistants recommend only about 1.2% of businesses on ChatGPT and 7.4% on Perplexity, while nearly 36% show up in Google's local map pack. In plain terms, being found in AI answers is roughly thirty times harder than being found in Google's local results, and almost nobody is checking whether they made the cut. This post is how you check. If you want the full playbook for fixing what you find, we wrote that separately in how to get your business found in ChatGPT and AI search.
Why does your Google ranking tell you nothing about this?
Because Google and ChatGPT read completely different data. Your Google position comes from Google's own crawl and its ranking system. ChatGPT does not use that. It routes its web search through Bing, then leans on structured business listings, the schema markup on your site, and third-party citations to decide who to name. Those are four separate data sources, and you can be strong in Google's while being absent from every one of ChatGPT's.
I have seen a roofer sitting at position two on Google for his main town and completely missing from ChatGPT's answer to "best roofer near me" in that same town, because he had never claimed Bing Places, his phone number was listed three different ways across directories, and his site had no LocalBusiness schema. Google forgave all of that. The AI did not. So the first mental shift is this: your Google ranking and your AI visibility are two different scores, and the only way to know the second one is to test it directly.
How do you check your business visibility in ChatGPT?
You run five specific prompts and read what comes back. Do this in a fresh chat with web search turned on, and do it signed out or in a private window if you can, so ChatGPT's memory of you does not color the answer. Type these exactly, swapping in your trade, your city, and your real business name:
- The existence check. "What do you know about [your exact business name] in [your city]?" This tells you whether the AI has any record of you at all, and whether the record is right.
- The recommendation prompt. "Who are the best [your trade] in [your city]?" This is the one that matters most, because it is the question a real customer asks. Watch whether you are named, and where in the list.
- The comparison prompt. "I need a [your trade] in [your city]. Give me three companies to call." This forces a shortlist. If you are not on it, you are losing that customer to whoever is.
- The specialty prompt. "Who does [your specific service, e.g. metal roof repair / recurring office cleaning / heated bathroom floor tile] in [your city]?" This checks whether the AI understands what you actually do, not just your category.
- The reputation prompt. "Which [your trade] in [your city] have the best reviews?" This tells you whether your reviews are feeding the AI's description of you, or whether they are invisible to it.
Save a screenshot of each answer. You are going to run the same five again in thirty days, and the only way to know if a fix worked is to compare against today's baseline. Ten minutes, five prompts, one screenshot each. That is the whole check.
Why run the same prompts in Perplexity and Gemini?
Because each engine pulls from a different mix of sources, so being visible in one does not mean you are visible in all three. Running the same five prompts in Perplexity and Gemini takes another few minutes and triples what you learn.
Perplexity is the most useful of the three for this audit, because it shows its sources inline. When it answers "best [trade] in [city]," it lists the exact pages it pulled from. If it cites a directory you are not on, that is your next listing to claim. If it cites a competitor's site and not yours, you are looking at the gap in plain sight. Gemini leans on Google's own index and Google Business Profile data, so it is the closest thing to a read on how Google's AI Overviews see you. ChatGPT, routing through Bing, is the one most sensitive to your Bing Places listing. Three engines, three different blind spots, same five prompts.
How do you read the result: invisible, wrong, or cited?
Every answer you get lands in one of three buckets. Sort each result into one before you do anything else:
- Invisible. The AI does not name you at all. On the existence check it says it has no information about your business, and on the recommendation prompt it lists competitors and skips you entirely. This is the most common result for businesses that have never done AEO work, and it is the one that costs the most, because the customer never learns you exist.
- Wrong. The AI names you but gets you materially wrong: wrong city, wrong phone, wrong services, an old business name, or it merges you with a similarly named company. Wrong is sometimes worse than invisible, because a customer who calls the wrong number or drives to the wrong town blames you, not the AI. If this is where you landed, the full repair sequence is in what to do when ChatGPT confuses your business with a competitor.
- Cited. The AI names you correctly, describes your services accurately, and includes you in the shortlist when a customer asks for a recommendation. This is the goal. If you are already cited in one engine, your job is to reproduce it in the other two.
Write down which bucket each of the fifteen answers (five prompts times three engines) fell into. The pattern across them is your real AI-visibility score, and it is far more honest than any number a tool hands you.
What does each failure mode actually point at?
Each bucket points at a specific, fixable cause. Invisible almost always means the AI has nothing structured to read: no Bing Places listing, thin or missing Google Business Profile, and no LocalBusiness or Service schema on your website, so there is simply no clean record for the engine to pull. Wrong almost always means an inconsistency: your name, address, and phone disagree across directories, or a larger business with a similar name is drowning out your signal, so the AI hedges or merges. Cited-in-one-but-not-another means the source that engine trusts is missing you specifically. ChatGPT missing you while Gemini has you usually points straight at an unclaimed Bing Places listing.
There is one more failure mode worth naming, because it is invisible in every other way: your site could be blocking the AI crawlers outright. A single line in your robots.txt, a Cloudflare setting, or a WordPress plugin can tell GPTBot and the others to stay out, and if that is happening, none of the fixes below will land until you undo it. That is a five-minute check of its own, and we walk through it in is your website blocking ChatGPT, and how to check in five minutes.
What do you fix first?
Fix in the order of what moves the AI's answer fastest. You do not do all of this at once, and you do not need to. Work down this list and re-run the five prompts after each step:
- Claim Bing Places for Business. ChatGPT routes web search through Bing, so this is the single highest-leverage move, and most of your competitors have not done it. It takes about twenty minutes and uses the data already in your Google Business Profile.
- Make every listing say the exact same thing. One business name, one address, one phone number, matched across Google Business Profile, Bing Places, Apple Business Connect, Yelp, and the trade directories. Inconsistency is what produces the "wrong" bucket.
- Add LocalBusiness and Service schema to your site. This gives the AI's crawler exact, parseable facts instead of prose it has to guess at. It is one of the highest-impact half-day changes a service site can make, and it is the backbone of what we mean by AEO. If the acronyms are new, start with what AEO and GEO actually are.
- Get reviews with specific language. AI engines read review text, not just the star count. Fifty reviews that describe the actual work get you described richly; ten that say "great" get you described as nothing.
- Write your key pages in question-and-answer form. The AI generates answers, so it quotes pages that already answer the question cleanly, with the answer in the first sentence.
None of these is exotic. All of them are tedious, and the order matters, which is exactly why they get skipped. The audit is ten minutes. The fixes are the work.
How often should you re-run the check?
Every thirty days, with the same five prompts and the same screenshots for comparison. AI answers are not stable the way a Google ranking is. They shift as the engines re-crawl, as your listings change, and as reviews come in, so a single check is a snapshot, not a trend. Running it monthly is enough to see whether a fix actually moved the needle, to catch a new competitor appearing in your recommendation prompt, and to notice when a correct citation quietly goes wrong again. Put it on the calendar the same way you would a monthly numbers review, because for the third of your customers who now start in an AI, it is a numbers review.
How does Current Digital handle AI visibility for clients?
This audit is the first thing we run on every new account, and rebuilding what it exposes is most of what our AEO service is. We build every client site with the LocalBusiness subtype, the areaServed city list, Service and FAQ schema, and question-and-answer content structure from day one, and the Bing Places claim and citation cleanup are part of the first ninety days, not an upsell later. The point is not to score well on a tool. It is to be the business the AI names when a customer asks it for a recommendation.
If you ran the five prompts and landed in the invisible or wrong bucket, that is fixable, and it is worth fixing before your competitors do it first. Run your site through our free AI-visibility and site grader for a baseline on AEO, schema, and technical health, see exactly what we handle on the SEO, AEO, and GEO service page, or book a call and I will tell you honestly whether your current site can be fixed in place or whether the visibility problem is a rebuild. Either way, do the ten-minute check first. You cannot fix a score you have never looked at.
FAQ
How do I check if ChatGPT knows my business exists?
Ask it directly. Open a fresh ChatGPT chat with web search on and type: "What do you know about [your exact business name] in [your city]?" If it returns your services, service area, and phone correctly, it can see you. If it says it has no information, guesses, or describes a different company, you are invisible or misread in the source data it pulls from. Then run a recommendation prompt like "best [your trade] in [your city]" to see whether it actually names you when a customer would be choosing.
Why doesn't ChatGPT recommend my business even though I rank on Google?
Because they read different data. Google ranks pages from its own crawl. ChatGPT routes web search through Bing and leans on structured listings, schema markup, and third-party citations. A business can sit on page one of Google and be missing from Bing Places, have inconsistent name-address-phone across directories, or have no schema on its site, and any one of those keeps it out of the AI's shortlist. A strong Google ranking tells you nothing about your AI visibility.
What does it mean when ChatGPT confuses my business with another company?
It means the AI cannot tell two entities apart, usually because your name-address-phone is inconsistent across the web or a larger namesake dominates the signal. The fix is entity disambiguation: standardize your listings to one exact name, address, and phone, add LocalBusiness schema with a sameAs link to your verified profiles, and earn a few strong third-party citations that describe you specifically. One clear citation beats ten weak ones.
Is checking my visibility in Perplexity and Gemini different from ChatGPT?
The method is identical, run the same prompts, but the answers differ because each engine pulls from a different mix of sources. ChatGPT leans on Bing, Perplexity blends several crawlers and cites its sources inline, and Gemini uses Google's index and Business Profile data. Running all three shows you which engines already cite you and which do not, and Perplexity's visible citations tell you exactly which pages it trusted, which is the fastest way to see what is working.
How often should I re-check my AI visibility?
Re-run the same prompts every 30 days and keep the screenshots. AI answers are not stable the way a Google ranking is, they shift as the engines re-crawl and as your listings and reviews change. A monthly check is enough to see whether a fix moved the needle and to catch a new competitor showing up in your recommendation prompt before it costs you calls.