Measuring your ChatGPT visibility isn't about typing your brand name once and seeing whether the tool answers correctly. Real measurement must reproduce the questions your prospects ask, compare your presence against your competitors and track how it changes over time.
AI visibility is unstable by nature: the prompt's wording, the context, the language, the country, the available sources and model updates can all change the answer. That's exactly why you need a method.
1. Define what you want to measure
Before running prompts, clarify the expected outcome. Are you trying to find out whether ChatGPT knows your brand? Whether it recommends you? Whether it uses your site as a source? Whether it compares you correctly against your competitors?
You can measure several levels of visibility:
- Brand visibility: your business is named in the answer.
- Source visibility: your site or content is used to generate the answer.
- Competitive visibility: your brand appears alongside the market alternatives.
- Recommendation visibility: AI presents you as a relevant or priority option.
2. Build a panel of realistic prompts
A good measurement panel shouldn't contain only your name. It must reflect your customers' real questions. For example: “which provider to choose for…”, “best solution for…”, “tool comparison…”, “who do you recommend in French-speaking Switzerland for…”.
Sort the prompts by intent: discovery, comparison, purchase, problem, location, alternative to a competitor. This segmentation helps you understand where you're visible and where you're absent.
Avoid overly artificial prompts. If no one would phrase the question that way, the data will be of little use. A good prompt looks like a natural request from a prospect.
3. Add your competitors as benchmarks
An isolated mention doesn't say much. Being cited once can seem positive, but if three competitors are cited ten times more often, your presence remains weak.
So for each prompt, note the brands cited, their order of appearance and how they're described. This lets you calculate a share of voice: across all the answers, what place does your brand hold relative to the market?
4. Capture the right signals
For each answer, collect at least five pieces of information:
- Is your brand cited?
- At what position does it appear in the answer?
- Is the tone positive, neutral or negative?
- Which competitors are mentioned?
- Which sources, arguments or criteria seem to influence the answer?
Position matters, but it isn't enough. A brand cited first with a vague description can be weaker than a brand cited second but backed by precise proof.
5. Repeat the measurement over time
A one-off measurement is a snapshot. A regular measurement becomes a dashboard. Re-run the same prompts at a fixed frequency to observe the trends: progress, decline, disappearance, sentiment change, new competitors.
Ideally, measure each day or each week depending on your market. What matters is keeping a stable cadence to compare consistent periods.
6. Turn the results into actions
Measurement is only valuable if it produces decisions. If ChatGPT cites competitors but not you, look for the sources that make them visible: comparison pages, reviews, directories, expert content, structured data, media mentions.
If your brand is cited but poorly described, work on your proof pages: positioning, use cases, FAQ, figures, “about” pages, external sources. If it's cited but rarely recommended, add content that clearly explains why and for whom your solution is relevant.
Measuring your ChatGPT visibility isn't about monitoring a curiosity. It's about understanding how an AI presents your business at the moment a prospect asks it for advice.