How do I check if ChatGPT mentions my brand?
Why one check is not a measurement
Answers vary by session, account history, region, and model version. Ask the same question twice and you can get two different brand lists. If you check once, see your name, and conclude you are visible, you have learned nothing.
Control what you can:
- Use a temporary or logged-out chat so personalisation and memory do not influence the answer.
- Use the same phrasing every time.
- Note the date and the model version.
- Run enough questions that one variable answer does not dominate.
One question on the panel“Best Korean sunscreen for oily skin?”
- ChatGPTnamed
- Perplexitynot named
- Claudenamed
- Gemininot named
- AI Overviewsnot named
2 of 5 — one question’s contribution to a named rate measured over the set you fix.
The questions to ask
Do not ask “what do you know about [your brand]”. A model will produce something for almost any brand name, and being named when the user already typed your name is not the commercial question.
Ask what a buyer asks before they know who you are:
- Best [product category] for [specific need]
- Best [category] for [audience segment]
- [Competitor] vs alternatives
- Is [attribute you compete on] worth paying more for
- Best [category] brands from [your country]
What to record
For each answer: were you named, in what position among brands mentioned, the exact sentence about you, which competitors were named, and which sources the assistant cited. The source list is the most actionable column — it tells you which domains you need to be present on.
When to stop doing this manually
At five to ten questions on one engine, manual checking is genuinely useful and costs twenty minutes a month. Past about forty questions across five engines it becomes a job, and the variance means you need the volume to read the trend. That is the point at which it makes sense to either build a light internal process or have someone run the panel for you.