Measure AI Visibility

How to Check If Your Business Shows Up in ChatGPT

By the Ask and Be Found team 7 min read
Short answer

To check if your business shows up in ChatGPT, open ChatGPT and ask the exact questions your buyers ask, such as “Who are the best [your service] providers in [your city]?” Run each prompt in a fresh chat with memory and personalization off, then note whether you appear, where you rank, and what ChatGPT says about you. At Ask and Be Found we run dozens of these prompts on a schedule so the read is reliable instead of a one-off.

If you want to know whether ChatGPT recommends your business, you do not need a special tool to get started. You need to act like one of your own customers and ask ChatGPT the questions they would ask before they buy. The catch is that a single chat tells you almost nothing. ChatGPT is probabilistic, it personalizes, and it sometimes browses the live web, so the same question can return different answers from one session to the next. A real check means running the right prompts, the right way, enough times to see a pattern.

This guide walks through exactly how to check your ChatGPT ranking yourself, how to read the results without fooling yourself, and what to do when the honest answer is “ChatGPT has never heard of me.” It is the same process our team runs at the start of every engagement, just translated into steps you can do today.

Why checking your ChatGPT visibility is harder than a Google search

On Google, you can search a keyword, scroll, and find your exact rank. ChatGPT does not work that way. It does not return a ranked list of ten blue links; it returns a short, opinionated recommendation that names a few businesses and skips the rest. If you are not one of the two or three names it mentions, you are effectively invisible, even if your website ranks on page one of Google.

Three things make this tricky to measure:

  • It varies. The same prompt can name different businesses across sessions because the model is probabilistic.
  • It personalizes. If you are signed in, ChatGPT may lean on your saved memory and past chats, which can make it mention you simply because you mention yourself a lot.
  • It sometimes browses. For local or current questions, ChatGPT often searches the live web and cites sources, so your reviews and listings feed the answer in real time.

That is why a clean test matters. Done casually, you will either flatter yourself or scare yourself. Done carefully, you get the truth.

Set up a clean test before you check your ChatGPT ranking

Before you type a single prompt, remove the things that bias the result. The goal is to see what a stranger sees, not what your own account has been trained to say back to you.

  1. Turn off memory and personalization. In ChatGPT settings, disable saved memory and chat history reference. If you skip this, ChatGPT may recommend you only because it remembers you.
  2. Use a fresh chat for every prompt. Earlier messages in a thread color later answers. Start a new chat each time so results do not contaminate each other.
  3. Test as a logged-out or incognito buyer when you can. A neutral session is the closest thing to how a real prospect experiences ChatGPT.
  4. Check more than one model. Free and paid tiers, and browsing versus non-browsing modes, can answer differently. Note which one you used.

Write down your setup once and keep it identical every time you check. Consistency is what turns a guess into a measurement.

The prompts that reveal whether you show up

The single biggest mistake is testing your own name. Asking “Tell me about [my company]” almost always returns something, because the model will summarize whatever it can find. That feels good and tells you nothing about whether ChatGPT recommends you when a buyer is deciding who to hire.

Run two kinds of prompts, and weight the first far more heavily.

Discovery prompts (you are not named)

These mirror how customers actually search. If you do not appear here, you have a visibility problem, full stop.

  • “Who are the best [your service] providers in [your city]?”
  • “I need help with [the problem you solve]. Who should I hire near [your area]?”
  • “What are the top-rated [your category] companies for [specific customer type]?”
  • “Who are good alternatives to [a competitor]?”

Brand prompts (you are named)

These tell you what ChatGPT says about you once it does find you, which is just as important as whether it finds you.

  • “Is [your company] reputable? What do people say about them?”
  • “What does [your company] specialize in, and who is it best for?”
  • “What are the pros and cons of working with [your company]?”

Run each prompt at least three to five times in fresh chats. You are looking for how often you appear and how you are positioned, not for one perfect answer.

How to read the results without fooling yourself

Once you have run your prompts, score them simply. A lightweight scorecard keeps you honest and gives you a baseline to improve against.

What to recordWhy it matters
Did you appear at all?The pass/fail question. No mention means no recommendation.
Position in the answerFirst name mentioned wins most of the attention; being fourth rarely converts.
Frequency across runsShowing up once in five tries is luck; four in five is real visibility.
Accuracy of the descriptionWrong services, old location, or stale claims cost you trust even when you appear.
Cited sourcesIf browsing is on, the links reveal which pages and reviews ChatGPT trusts about you.

When ChatGPT browses and shows citations, click them. Those sources are a map of exactly which directories, review profiles, and pages are shaping your reputation. If the citations are all third parties and none are your own site, that is a signal to publish clearer, answer-first content of your own. Our guide to measuring your AI search visibility goes deeper on turning these observations into a repeatable score.

What it means when you do not show up

Drawing a blank is common, and it is rarely about your website being “bad.” Across the audits we run, the businesses that AI ignores almost always share the same gaps. ChatGPT recommends names it can verify, and it cannot verify what it cannot find clearly stated and corroborated in more than one place.

The usual culprits:

  • Vague website copy. Pages that describe a vibe instead of plainly stating what you do, where, and for whom give the model nothing concrete to repeat.
  • Inconsistent business data. A name, address, or phone number that differs across your site, Google Business Profile, and directories makes AI unsure you are one trustworthy entity.
  • Too few real reviews. Answer engines lean heavily on review volume and recency as proof you are real and well regarded.
  • Missing from the sources AI trusts. If you are absent from the directories and roundups ChatGPT cites, you are absent from its answers.
  • No structured data. Without schema markup, the model has to guess at facts a machine could otherwise read with certainty.

This is the core work of answer engine optimization: giving AI the structured, consistent, verifiable evidence it needs to name you with confidence. It is also why showing up is fixable. When we took a Seattle mortgage broker, Keith Akada, from invisible to the number-one AI-recommended broker in his market, the change came from exactly these fundamentals, and it produced roughly 30 leads and four closed deals inside six weeks.

When to graduate from manual checks to tracking tools

A manual check is the right way to start, and you should do it before you spend a dollar. But manual checks have a ceiling. You can realistically test a handful of prompts a few times before fatigue sets in, and a one-time snapshot cannot tell you whether you are trending up or down.

Once you have a baseline, AI visibility tracking tools take over the grind. They run dozens or hundreds of prompts on a schedule, across ChatGPT and other engines, log every time you appear and where, and chart the trend so you can see whether your work is moving the needle. That is also how you tell improvement apart from the day-to-day randomness of a probabilistic model. If you are weighing your options here, our overview of auditing your brand across AI platforms covers how to extend the same discipline beyond ChatGPT to Gemini, Perplexity, and Copilot.

Turn the check into a plan

Checking your ChatGPT visibility is the easy part once you know the setup: clean session, buyer-style prompts, several runs, an honest scorecard. The harder and more valuable part is what you do with the answer. If you already show up, protect it by keeping your information accurate and your reviews fresh. If you do not, treat it as a clear, fixable diagnosis rather than a verdict, and start closing the gaps the model is telling you about. The businesses that win in AI search are not the loudest; they are the easiest for an answer engine to trust.

Want to see if AI is recommending you? Get a free AI visibility report.

Run My Report →

Frequently asked questions

How do I check if my business shows up in ChatGPT?
Open ChatGPT and ask the buying questions your customers ask, such as “Who are the best [your service] providers in [your city]?” Run each prompt in a fresh chat, with memory and personalization turned off, and note whether your name appears, where it ranks, and what ChatGPT says about you. Repeat across a handful of phrasings to get a reliable read instead of a one-off result.
Why doesn’t my business show up in ChatGPT?
Usually it is because ChatGPT cannot find enough clear, consistent, third-party evidence that you exist and that you are credible. Thin or vague website copy, inconsistent name-address-phone data, few reviews, and no presence in the directories AI trusts all make you invisible. The fix is to give answer engines the structured, verifiable signals they rely on.
Does ChatGPT search the live web or use training data?
It does both, depending on the question and the model. For local or current recommendations, ChatGPT often browses the live web and cites sources, so your reviews, directory listings, and recent content matter. For general questions it may rely on what it learned during training, which is why building durable, widely referenced authority pays off over time.
Why does ChatGPT give me different answers each time I check?
ChatGPT is probabilistic, so the same prompt can produce slightly different answers across sessions. Personalization, saved memory, and live browsing add more variation. To get a trustworthy read, run each prompt several times in fresh chats with memory off and look for patterns rather than treating any single answer as final.
What prompts should I use to test my ChatGPT visibility?
Use the questions a real buyer would type: “Best [service] in [city],” “Who should I hire for [problem]?”, “Is [your company] any good?”, and “Who are [a competitor]’s top alternatives?” Mix discovery prompts, where you are not named, with brand prompts, where you are, so you learn both whether AI finds you and what it says about you.
Are AI visibility tracking tools worth it?
Yes, once you are past a quick manual spot-check. Tracking tools run many prompts on a schedule across ChatGPT and other engines, log when you appear, and show trends over time. That removes the noise of probabilistic answers and tells you whether your visibility is actually improving instead of relying on a single lucky or unlucky chat.

Keep reading

Become the answer AI recommends

We make your business the name ChatGPT, Claude, Gemini, and Google AI give when buyers ask.

Book a call