Here is the short answer. To track your brand in ChatGPT, build a fixed set of the questions your buyers actually ask, run them on a schedule, and record six things each time: whether you are mentioned, where you rank in the answer, how you are described, your share of voice against competitors, and which sources the model relied on. The metric that matters is not any single answer, it is the trend across runs. You can start by hand and move to an AI visibility tool once you want it consistent, historical, and across every engine.
This is now a real growth metric, not a curiosity. ChatGPT answers questions for roughly 1 billion people every week, and when it names a brand, that recommendation frequently decides the purchase before the buyer ever opens a website. Being absent from the answer is a silent loss: there is no click, no bounce, no trace of the customer you never met.
The short version
Track a fixed prompt set, the real buyer questions, not your brand name.
Record six metrics: mention (presence), position, sentiment (how you are described), share of voice vs competitors, and cited sources.
Watch the trend, not a single snapshot, because AI answers change constantly.
Cover every engine, not just ChatGPT, since each one cites different sources.
Why tracking ChatGPT is now a core marketing metric
Discovery is moving from a page of links to a sentence an AI writes. A Pew Research Center study found that when an AI summary appears on a Google results page, people click a normal result just 8 percent of the time, versus 15 percent without one, and only about 1 percent click a source inside the summary. In a chat tool like ChatGPT there are often no links at all, only the answer and the handful of brands it chose to name. Google's own AI Overviews already reach more than 2 billion people a month, so the "answer layer" is not a fringe channel, it is where a large share of buying decisions now start.
Put plainly: "Does ChatGPT recommend us?" belongs next to "Do we rank on Google?" on your dashboard. And like any metric, you cannot improve what you do not measure.
What "tracking" actually means: the six metrics
Vague monitoring ("do we show up?") is not enough. The AI visibility industry tracks a specific, repeatable set of signals. These are the six you should record for every prompt, every run.
Presence, or mention rate. Across your whole prompt set, in what share of answers does your brand appear at all? This is your baseline visibility number.
Position. When you do appear, are you named first, in the middle, or last? Order in an AI answer carries weight, much like rank on a results page.
Sentiment. How does the model describe you, positive, neutral, or with a caveat? A mention that comes with "but it is expensive" is not the same win as a clean recommendation.
Share of voice. Out of all the brands named across your prompts, what percentage are you, versus each competitor? This is the single clearest competitive scoreboard.
Citations, or sources. Which pages did the answer draw on? These sources are your roadmap: get mentioned there and the answer changes.
Accuracy. Does the model state anything wrong about you, outdated pricing, a discontinued product, a factual error? Misinformation in an answer is its own problem to fix.
To choose the right prompts and read the sources correctly, it helps to understand the mechanism first. See what are AI citations, and why they decide if AI recommends your brand.
Why a single check tells you almost nothing
The most common mistake is asking ChatGPT about your category once and drawing a conclusion. AI answers are probabilistic and dynamic, so one look is noise, not signal:
They change with phrasing. "Best accountant for startups" and "affordable startup accountant" can surface different brands.
They change over time. As models update and the web changes, this month's answer is not next month's.
They change with context. Location, prior conversation, the specific model version, and whether web browsing is on all shift the result.
The sources shift too. The pages a model leans on for your category are a moving target.
That is why tracking must be systematic and repeated. The number that matters is the direction of travel: more mentions, better position, higher share of voice, cleaner sentiment, run after run.
How to track your brand in ChatGPT, step by step
Whether you use a spreadsheet or a platform, the method is the same.
1. Build a prompt set. List 15 to 40 real buyer questions, the ones asked at the moment a recommendation is made: "best [category] for [use case]," "top [category] companies," "[category] alternatives to [competitor]," "is [your brand] good." Fix this list so every run is comparable.
2. Run each prompt and log the six metrics. Mention, position, sentiment, share of voice, sources, accuracy. Keep one row per prompt per run.
3. Record the sources. When the answer cites or clearly draws on pages, note them. Over time, patterns appear: the same reviews, roundups, and threads keep feeding the answer.
4. Repeat on a schedule. Weekly or monthly, same prompts, and chart the movement. The history is the whole point.
5. Cover every engine. ChatGPT is one surface. Buyers also ask Google's AI Overviews and AI Mode, Gemini, Perplexity, Microsoft Copilot, Claude, Grok, and Meta AI, and each pulls from different sources. Tracking ChatGPT alone leaves blind spots.
Doing it by hand vs using an AI visibility tool
Manual tracking is a good way to start and to prove the problem is real. It breaks down once you want scale, history, and multi-engine coverage without spending hours a week.
What you need | Tracking by hand | An AI visibility tool |
|---|---|---|
Run a large prompt set | Slow, one prompt at a time | Automated in bulk |
Trend over time | Your own spreadsheet | Recorded and charted for you |
Cover multiple engines | Repeat everything per engine | Tracked side by side |
Share of voice vs competitors | Easy to miscount | Calculated automatically |
Source and sentiment analysis | Manual reading | Surfaced for you |
This is precisely what an AI visibility platform such as Reddlix does: it runs the questions your buyers ask across ChatGPT, Gemini, Claude, Perplexity and more, then records presence, position, sentiment, share of voice, and the sources behind each answer, so you watch the trend instead of guessing. For a broader comparison of the category, see the best AI visibility tools in 2026.
From tracking to action
Measurement is half the job. The point is to close the gap it reveals:
If you are rarely mentioned, the fix is almost always off-site: earn citations in the sources these models trust, reviews, best-of roundups, authoritative articles, and the community threads in your niche.
If your site blocks AI crawlers (for example OpenAI's GPTBot, PerplexityBot, or Google's Google-Extended) or is unreadable without JavaScript, you are invisible before the race starts. A free Technical Audit shows exactly how AI crawlers see you.
If a competitor keeps winning, your source analysis is the answer key: the pages that name them are your shortlist of where to show up.
The full method of closing that gap is laid out in how to get your business recommended by ChatGPT and other AI models.
The bottom line
For a growing share of buyers, the ChatGPT answer is the shortlist. Tracking your brand in it is now as fundamental as rank tracking was for SEO: build a fixed prompt set, record presence, position, sentiment, share of voice, and sources, repeat on a schedule, and cover every engine. Do that, and "I wonder if AI recommends us" becomes a number you can watch and move.
Start tracking your AI Visibility and boost your presence to get more qualified customers and users.
References
Pew Research Center (July 22, 2025). "Google users are less likely to click on links when an AI summary appears in the results." Users clicked a result 8 percent of the time with an AI summary versus 15 percent without, and only about 1 percent clicked a source inside the summary. https://www.pewresearch.org/short-reads/2025/07/22/google-users-are-less-likely-to-click-on-links-when-an-ai-summary-appears-in-the-results/
OpenAI, reported via DemandSage (2026). ChatGPT reached about 1 billion weekly active users in 2026. https://www.demandsage.com/chatgpt-statistics/
TechCrunch (July 2025). Google's AI Overviews reached more than 2 billion monthly users. https://techcrunch.com/2025/07/23/googles-ai-overviews-have-2b-monthly-users-ai-mode-100m-in-the-us-and-india/
Usage figures are as reported in 2025 and 2026 and change over time. Verify the latest numbers at the linked sources.