How to Check If ChatGPT Mentions Your Business (Free DIY Self-Check)

When a potential customer asks ChatGPT “who are the best companies for [what you do] near me?” — does your business come up? More buyers now run that exact query inside an AI assistant before they ever touch a search engine. If the model doesn't name you, names you incorrectly, or names a competitor instead, you lose deals you never even see in your analytics.

The good news: you can check this yourself, today, for free, in about fifteen minutes. You don't need a tool or a consultant to find out whether you have a problem. Here's the exact process.

Step 1: Write down the prompts a real buyer would use

Don't search for your own company name — of course an assistant can repeat a name you hand it. The point is to discover whether you surface unprompted, the way a stranger looking for a vendor would find you. Write 5–10 prompts in plain language, for example:

• Who are the best [your service] providers in [your city]?
• I need a [your service] for a small business — who should I look at?
• Recommend a few companies that do [the specific thing you sell].
• Compare the top [your category] options and their strengths.
• What's a reliable [your category] vendor for [your customer type]?

Mix in the phrasing your actual customers use, not internal jargon. These are the questions that decide whether you make the shortlist.

Step 2: Run them across more than one assistant

Different assistants pull from different sources, so test the ones your buyers actually use: ChatGPT, Perplexity, Google Gemini, and Microsoft Copilot. Run each prompt in a fresh chat with no prior context, and where the tool offers it, turn on web browsing — that's closer to how a buyer with a live question gets answered. Run each prompt once or twice; answers vary slightly between runs, and that variation is itself useful information.

Step 3: Record what actually happened

For each prompt and each assistant, note one of four outcomes:

1. Named and accurate — you're mentioned and the description is correct.
2. Named but wrong — you appear, but the services, location, or details are inaccurate or outdated.
3. Absent — you're not mentioned at all.
4. Competitors only — the assistant confidently lists rivals and not you.

A simple grid — prompts down the side, assistants across the top — makes the pattern obvious fast. If most cells are “absent” or “competitors only,” you have a real visibility gap. If you're “named but wrong,” that's often more urgent than being absent, because the assistant is actively telling buyers something false about you.

Step 4: Find out why — ask the assistant directly

The models will often tell you their reasoning if you ask. Follow up with prompts like What sources are you using to answer that? or Why didn't you mention [your business]? You'll usually see one of a few root causes: thin or missing information about you on the open web, no structured data the model can parse cleanly, inconsistent details across directories, or simply that competitors have far more citable content about the exact thing buyers are asking for. Treat these answers as leads to investigate, not gospel — assistants can be confidently wrong about their own reasoning too.

Step 5: Fix the highest-leverage gaps first

You don't need to fix everything at once. The common, high-impact fixes:

• Make the basics unambiguous and consistent — what you do, who you serve, where, and how to reach you — on your own site and across the directories assistants read.
• Add structured data (Organization, LocalBusiness, FAQ schema) so models can parse your business cleanly rather than guessing.
• Publish content that answers the exact buyer questions from Step 1, in plain language an assistant can quote.
• Correct the record wherever you found “named but wrong” — stale listings and outdated descriptions are frequently the source.

Re-run your prompt grid every few weeks. Visibility in AI answers shifts as models and their sources update, so this is a check you repeat, not a one-time task.

When to stop doing it yourself

The DIY check is genuinely enough to tell you whether you have a problem and roughly where it lives. Where people get stuck is turning a messy grid of results into a ranked, do-this-first plan — and confirming the fixes actually moved the needle. If you'd rather not assemble that yourself, that's exactly the gap a structured audit fills.

Part of our GEO guide: Which AI Engines Name Your Business? — compare how ChatGPT, Perplexity, Gemini and Copilot each pick who to name.