How to Track Brand Mentions in ChatGPT and Google AI Mode (2026 Guide)
How to Track Brand Mentions in ChatGPT and Google AI Mode (2026 Guide)
You can track brand mentions in ChatGPT and Google AI Mode with three data sources: a fixed set of buyer prompts run on a schedule, referral data in Google Analytics, and the Generative AI performance report in Google Search Console. The prompts show whether you are mentioned, the analytics show whether those mentions send visitors, and Search Console shows how often Google’s AI features display your pages.
The two platforms need different methods. ChatGPT is a separate product with no publisher console, so most of its tracking depends on sampling answers yourself. Google AI Mode sits inside Google Search, which means some of its data already appears in the Search Console account you use for SEO.
This guide covers both, step by step, and includes the figures we found when we tracked our own site at Creative For More.
Why Do ChatGPT and Google AI Mode Need Different Tracking Methods?

ChatGPT and Google AI Mode need different methods because Google reports AI Mode data to site owners and OpenAI does not report ChatGPT data in the same way. For ChatGPT, you measure mentions by sampling answers and measure clicks through referral tags. For AI Mode, you can combine sampling with Google’s own reporting.
| What you want to know | ChatGPT | Google AI Mode |
|---|---|---|
| Is my brand mentioned in answers? | Prompt sampling, manual or with a tool | Prompt sampling, manual or with a tool |
| How often are my pages shown? | No official publisher report | Search Console Generative AI performance report |
| Do mentions send visitors? | Google Analytics, source chatgpt.com | Included in Google organic traffic, not separated by default |
| Can the engine reach my site? | Check robots.txt allows OAI-SearchBot | Page must be indexed and eligible for a snippet |
How Do You Track Brand Mentions in ChatGPT?
You track brand mentions in ChatGPT by running the same list of buyer prompts at regular intervals and recording whether your brand appears, whether your site is cited, and which competitors are named. Referral data in Google Analytics then shows whether those mentions turn into visits.
- Write 20 to 50 buyer prompts. Use the questions a customer asks before choosing a provider, such as “best B2B marketing agency in Singapore” or “how to enter the Japanese market as a US brand.” Include category questions, comparison questions and problem questions, not only your brand name.
- Fix the test conditions. Use a logged-out session or a clean account with memory turned off, and keep the location and language the same between runs. Personal chat history changes the answers, so results from your own account are not representative.
- Run each prompt more than once. ChatGPT gives different answers to the same prompt, so run each one at least three times per check and record how many runs mention you.
- Record five fields per answer. Log whether the brand was mentioned, whether your site was linked as a source, where the mention appeared in the answer, which competitors were named, and whether the description was accurate.
- Check referral traffic. According to the OpenAI Publishers and Developers FAQ, ChatGPT automatically adds the parameter
utm_source=chatgpt.comto referral URLs, so these visits appear as a named source in Google Analytics. - Confirm that ChatGPT can reach your site. The same OpenAI FAQ states that content will appear in ChatGPT summaries and snippets only if the site isn’t blocking OAI-SearchBot. Check your robots.txt file and any firewall or bot-protection rules.
In Google Analytics, open the Traffic acquisition report and filter the session source for “chatgpt.” A custom channel group that collects ChatGPT, Perplexity, Gemini and Copilot sources into one “AI referrals” channel makes the trend easier to follow month to month.
How Do You Track Brand Visibility in Google AI Mode?
You track brand visibility in Google AI Mode through the Generative AI performance report in Google Search Console, combined with prompt sampling in AI Mode itself. The report shows how often your pages appear in AI features, and sampling shows the context in which your brand is described.

According to Google’s Search Console documentation on the Generative AI performance report (2026), the report covers AI Overviews and AI Mode and counts impressions as the number of times links to your site were shown in a generative AI feature. It can be broken down by page, country, device and date.
The report has limits worth knowing before you rely on it. Google states that it is rolling out over time and not every property has access yet, data from Search Labs experiments is excluded, and multiple links from the same site in one AI answer count as a single impression at property level.
What if your property doesn’t have the report yet?
AI Mode data is still counted in your standard Search Console figures. According to Google Search Central’s guide to AI features, sites appearing in AI Overviews and AI Mode are included in the Performance report under the “Web” search type. You can’t isolate AI Mode there, but the traffic isn’t lost.
Google’s guidance also states that there are no extra technical requirements for AI Mode beyond being indexed and eligible to show with a snippet. A page that can’t appear in standard Google results can’t appear in AI Mode either, so indexing problems are the first thing to rule out.
Which Metrics Should You Record for ChatGPT and AI Mode?
The most useful metrics are mention rate, citation rate, share of voice, accuracy and referral sessions. Together they show whether AI engines know your brand, trust your pages as a source, rank you against competitors and send visitors.
| Metric | How to calculate it | Where the data comes from |
|---|---|---|
| Mention rate | Answers that name your brand ÷ total answers sampled | Prompt sampling |
| Citation rate | Answers that link to your site ÷ total answers sampled | Prompt sampling |
| Share of voice | Your mentions ÷ mentions of you and named competitors | Prompt sampling |
| Accuracy | Answers that describe your offer correctly ÷ answers that mention you | Prompt sampling, reviewed by a person |
| AI impressions | Count reported by Google | Search Console Generative AI report (AI Mode and AI Overviews) |
| Referral sessions | Sessions with source chatgpt.com | Google Analytics |
A monthly check is enough for most brands. AI answers change often, so a single month’s figure means little, and the trend across three to six months is what shows whether your content and coverage are working.
What Did We Find When We Tracked Our Own Site?

When we checked creativeformore.com, Google AI Mode cited the site 28 times across 16 pages, while ChatGPT cited it 0 times. The figures come from Ahrefs’ AI citation data, pulled on 24 September 2026.
| Engine | Citations to creativeformore.com | Distinct pages cited |
|---|---|---|
| Google AI Mode | 28 | 16 |
| Perplexity | 27 | 15 |
| Microsoft Copilot | 23 | 12 |
| Google AI Overviews | 5 | 5 |
| Gemini | 2 | 2 |
| ChatGPT | 0 | 0 |
Source: Ahrefs Site Explorer AI citations report for creativeformore.com, 24 September 2026.
The gap shows why ChatGPT and AI Mode have to be tracked separately. A brand that checked only ChatGPT would conclude AI engines ignore it, and a brand that checked only AI Mode would assume its visibility carries over to ChatGPT. Neither conclusion would be correct.
For a site in this position, the first checks are the ones in the steps above: confirming OAI-SearchBot can reach the site, and testing which buyer prompts produce answers that cite competitors instead. The pages that already earn AI Mode citations also show which topics and formats the site is trusted for.
Can You Track Brand Mentions in ChatGPT for Free?
You can track brand mentions in ChatGPT for free by running prompts manually, logging the results in a spreadsheet and reading ChatGPT referrals in Google Analytics. For AI Mode, the Search Console report is also free once it reaches your property.
The free method works for 20 to 30 prompts in one market. It becomes hard to maintain once you track several markets, several competitors or more than one engine, because each prompt has to be run repeatedly and logged by hand.
Paid trackers automate the sampling and add history, competitor benchmarks and cited-source lists. Ahrefs Brand Radar, which we use, bases its prompts on real search demand. Our comparison of AI visibility tools covers the others, with pricing checked on each vendor’s own page.
How Do You Track ChatGPT and AI Mode Across Different Markets?

You track visibility across markets by running each prompt set separately for each country and language, because ChatGPT and AI Mode give different answers to the same question in Singapore, Japan and the United States. One global prompt list hides those differences.
For a brand selling in several countries, set up one prompt list per market in the local language, record results by country, and filter the Search Console Generative AI report by country to match. A brand can be the default recommendation in one market and absent in the next, and only market-level tracking shows it.
This is the part of AI visibility work our Generative Engine Optimisation team spends the most time on for cross-border clients. Tracking shows where a brand is missing, and the fix usually combines content built on a sound SEO content strategy with coverage on the third-party sites AI engines cite in that market.
We run this process across our own markets using the tools described in the AI stack we use, which keeps prompt tracking, keyword data and citation data in one place.
If you want to know how your brand appears in ChatGPT and Google AI Mode across Singapore, Japan and the US, a market-by-market baseline is the place to start. If your company is looking for a strategic marketing partner, the Creative For More team can help. Book a discovery call to explore how we can support your brand’s growth.