How to check if ChatGPT recommends your product

More buyers now ask ChatGPT for a recommendation before they search Google or ask a friend. If your product isn't in that answer, you never get the visit. The good news is that checking takes minutes, as long as you check the right way.

The quick way: a free check

The fastest way is the free AI visibility check. Enter your domain, confirm what your product does, and it asks ChatGPT and Claude 5 buyer questions about your category. In about a minute you see a score, which questions you're named in, and which rivals are named instead. No sign-up needed.

If you want to understand what's happening, or you'd rather do it by hand, here's the manual method.

Step 1: Ask buyer questions, not your brand name

The most common mistake is typing "What is [your product]?" ChatGPT will usually describe you if you exist online, but that tells you nothing about whether it recommends you. Buyers don't ask about you by name. They ask questions like:

• "What's the best invoicing tool for freelancers?"
• "Alternatives to [the market leader]"
• "Which [category] tool is cheapest for a small team?"
• "[Category] tool that integrates with Slack"

Write down 5 to 10 questions like these. Use the words your customers use on sales calls, not your marketing words. Our guide on choosing buyer questions goes into detail.

Step 2: Ask in a clean chat

ChatGPT remembers things about you. If you've talked about your product before, it may mention you because of that memory, not because it would recommend you to a stranger. To get a fair answer:

1. Use a temporary chat, or log out, or use a browser profile with no history.
2. Paste the question exactly as a buyer would type it, with no extra context.
3. Don't follow up with "what about [your product]?" A buyer wouldn't.

Step 3: Ask more than once

ChatGPT doesn't give the same answer every time. Ask the same question three times and you might be named twice, or once, or in a different position. That's normal: why ChatGPT gives different answers explains why.

So ask each question at least three times in fresh chats and count how often you're named. "Named in 2 of 3 answers, usually third" is a result. "Named once" is a guess.

Step 4: Record what matters

For each answer, write down four things in a simple table:

Also note any links or sources ChatGPT shows. Those pages shaped the answer, and they're your best clue for what to fix. See which websites AI assistants cite.

Step 5: Check the other assistants too

ChatGPT is the biggest, but Gemini and Claude are growing fast and often give different answers. A product can be named by Claude and missing from ChatGPT, or the other way round. Ask the same questions in each one. Results that differ between assistants usually point to different sources being cited.

How to read the results

• Named in most answers, near the top: you're in good shape. Keep tracking so you notice if a rival moves past you.
• Named sometimes, low down: you're on the edge. Getting onto the pages that are cited is usually what moves you up.
• Never named, same rivals every time: those rivals are on pages you're not. Start with the sources cited in their answers.
• Nobody is named: the question is wide open. A clear page that answers it directly can win it.

Why one check isn't enough

A single check tells you where you are today. It doesn't tell you whether you're going up or down, or whether the blog post you published last month made any difference. Answers move week to week as engines update and new pages appear.

That's what tracking is for. Citemark puts your buyer questions to ChatGPT and Gemini (and Claude on Pro) on a schedule, with your most important questions checked daily. You get a score and its trend, the rivals named instead of you, the cited pages, and alerts when you drop out of an answer.

Turn the result into action

Once you know where you stand, the playbook for getting recommended by ChatGPT covers what to do next, in order of impact. The short version: make sure AI crawlers can read your site, publish clear comparison and pricing pages, and get onto the pages the answers already cite.

Common mistakes to avoid

A few habits make manual checks look better or worse than reality:

• Asking about yourself by name. "Is [product] good?" tests whether ChatGPT knows you, not whether it recommends you.
• Leading the question. "Best tools like [your product]" puts your name in the question. Buyers rarely do that.
• Checking once and stopping. A single good answer can be luck, and a single bad one can be too.
• Only checking ChatGPT. Gemini and Claude can give completely different lists.
• Ignoring position. Being named sixth in a list of six is very different from being named first.
• Changing the questions every time. If the wording changes, you can't compare this month with last month.

Keep the same list of questions, ask them the same way, and write the results down. That's what turns a curious look into a measurement you can act on.

Try it now

Run the free check to see today's answer in about a minute. To follow it every day, start a 7-day Pro trial with no card.