To find out whether ChatGPT mentions your brand, ask it the questions your customers ask, ask more than once, and check two things separately: whether ChatGPT’s answer names you, and whether ChatGPT’s search results cite or mention you. One chat in your own account is a single sample. The SEO MCP returns both from inside Claude or ChatGPT. Its llm_responses_chatgpt tool returns ChatGPT’s answer to a prompt (18 credits at Growth Engine rates), and llm_scraper_chatgpt returns ChatGPT search results for a keyword, with cited sources and brand mentions (6 credits).
Last updated: November 2026
“Mentioned” means two different things
This is the part that changes how you check. Most people open ChatGPT, type their brand name, see a friendly paragraph and decide they’re covered.
That test answers the wrong question twice.
First, typing your brand name only tests whether the model has heard of you. Your customers don’t ask that. They ask “who’s a good bookkeeper for a small bakery?” or “best emergency plumber near me”, and the question you care about is whether you’re in that answer.
Second, there are two places a mention can show up, and they’re worth checking separately:
| What you’re checking | The SEO MCP tool | What it returns | Credits (Growth Engine) |
|---|---|---|---|
| ChatGPT’s answer | llm_responses_chatgpt |
ChatGPT’s answer to a prompt | 18 |
| ChatGPT search | llm_scraper_chatgpt |
ChatGPT search results for a keyword, with cited sources and brand mentions | 6 |
Nothing says the two will agree, so don’t assume they do. The answer tells you whether you’re recommended. The search results tell you which sources were cited, and that list is the part you can actually act on.
Why your own screenshot proves very little
Here’s my firm opinion on this: a screenshot of ChatGPT naming your business is close to worthless as evidence. It shows that it happened once.
AI answers aren’t fixed rankings. Ask the same question twice and the wording can change, and sometimes the businesses named change with it. Your own account’s chat history and memory settings can also shape what you see. So one good result in your own chat is a nice moment, not a measurement.
Local businesses already know this problem from Google Maps. One search from your desk shows one point, which is why a geo grid scan checks up to 49 points around a location. A ChatGPT check needs the same many-samples habit, only the samples are questions and repeat runs instead of map coordinates.
A one-afternoon brand check
You don’t need a tracking program to get a first answer. You need a short list of questions and a few runs of each.
- Write five to ten customer questions. Use the words a customer would type, not your brand name. Add a second phrasing of each.
- Ask ChatGPT each one twice through
llm_responses_chatgpt. - Run your main keywords through
llm_scraper_chatgptto see the cited sources and brand mentions in ChatGPT search results. - Write down three things per result: whether you’re named, who’s named instead, and which sources are cited.
You don’t call these tools by name. In Claude or ChatGPT, you’d type something like:
“Ask ChatGPT ‘who’s a good bookkeeper for a small bakery?’ and ‘which bookkeeping service do bakeries use?’, twice each. Then pull ChatGPT search results for ‘bakery bookkeeper’ and ‘bookkeeping for bakeries’. Make a table of which businesses are named in each and list the cited sources.”
At Growth Engine rates, ten questions asked twice is 360 credits, and ten search checks add 60, so the whole afternoon is 420 credits. Local Dominator uses fewer credits per call than those figures. New accounts get 1,000 free credits once they verify their email, and paid plans start at $99 a month.
Want a second opinion beyond ChatGPT? Matching tools return answers from Claude (18 credits) and Gemini (17), and Perplexity’s answer comes with its cited sources (12).
Reading what comes back
Most results fall into one of three patterns.
You’re named. Good. Note where you appear (first recommendation, one of several, a footnote) and whether the description is right. If ChatGPT describes services you don’t offer, look at the cited sources from the search check, since they’re the closest thing you’ll get to a reason.
Competitors are named and you aren’t. Those businesses are your real rivals in AI answers, and they aren’t always the ones you rank against in Google. Open the sources cited next to them. The on_page_content_parsing tool (44 credits) breaks down a page’s structure, so you can see what a cited page covers that yours doesn’t.
Nobody is named consistently. The answers shift from run to run. Sometimes the question is too vague; sometimes very few people ask it. Before you chase it, check demand: ai_keyword_data_search_volume estimates how often people ask ChatGPT, Gemini and Perplexity about a keyword, with a 12-month trend, at 29 credits per started block of 100 keywords. Our ChatGPT keyword research with real data post covers why measured numbers beat the ones a model guesses.
Where this check stops
Be clear about the limits before you plan around it.
No tool can tell you why a model picked one brand over another. You can see what it said and what it cited, and the reasons stay inside the model.
Every check is a live call that runs when you ask. Nothing in The SEO MCP’s catalog runs on a schedule, so if you want to know whether things are improving, save each round’s table and compare. Our LLM visibility tracking guide covers turning this one-off check into a repeat routine, with prompt sets and mention rates. Two further tools that count mentions across many LLM answers exist too, but they’re available on request only.
If you’d like to run the checks from inside ChatGPT itself, which is a little funny but works, the ChatGPT SEO MCP post covers connecting it through Developer mode. Setup everywhere is one hosted URL and a browser sign-in, with nothing to install.
Frequently asked questions
How do I check if ChatGPT mentions my brand?
Ask ChatGPT the questions your customers ask, not your brand name, and ask each one more than once. Record whether you’re named, who’s named instead and which sources are cited. With The SEO MCP, llm_responses_chatgpt returns ChatGPT’s answer to a prompt and llm_scraper_chatgpt returns ChatGPT search results with cited sources and brand mentions.
Why does ChatGPT mention my brand sometimes and not other times?
AI answers aren’t fixed rankings. The same question can produce different wording, and sometimes different businesses, from one run to the next. That’s why one result proves little. Ask several phrasings of each customer question, repeat each a few times, and look at how often you’re named across all the answers rather than in any single one.
Can The SEO MCP show which sources ChatGPT cites?
Yes, for ChatGPT search results. The llm_scraper_chatgpt tool returns ChatGPT search results for a keyword, with cited sources and brand mentions, for 6 credits at Growth Engine rates. A separate tool returns Perplexity’s answer to a prompt with its cited sources. Checks run when you ask; nothing in the catalog runs on a schedule.
Can I make ChatGPT mention my brand?
No tool can make a model name you. What you can do is study the sources it cites for your customers’ questions and the competitors it names instead. Then work on being covered by those sources, or publish a clearer answer to the question on your own site, and run the same check again later to see whether anything moved.