How to measure your brand's visibility in AI answers

1. Choose the questions your buyers really ask

Start from the buyer, not from your keywords. Write the questions someone would type into an assistant when looking for what you sell, in their language and with their constraints.

• Discovery: "What are the best payroll tools for a small company in France?"
• Comparison: "Is X or Y better for a 20 person team?"
• Use case: "Which CRM works well for a real estate agency?"
• Alternatives: "What are good alternatives to X?"
• Keep most questions unbranded. Branded questions tell you how you are described, not whether you are found.

2. Group them by theme

Group questions by line of business or use case. A single average across everything you sell hides the theme where you are absent. Twenty to fifty questions spread across a few themes is a solid start for most companies.

3. Run each question several times, on each assistant

Because answers vary, ask each question several times and record every answer. Run them on each assistant separately: ChatGPT, Google AI Overviews and AI Mode, Gemini, Perplexity. Results on one say little about the others.

Keep the conditions consistent between measurements: same country, same language, same questions. Otherwise a change in the numbers may come from the setup, not from your work.

4. Read the answers for a few simple figures

For each answer, note who was named and which sources were shown. Then compute:

• Mention rate: the share of answers that name your brand at least once.
• Share of voice: your mentions compared with every brand named in the same answers.
• Position: whether you appear first, in the middle or at the end of the list.
• Sources: the domains the assistant cites most for your questions, and whether you appear on them.

5. Repeat on a schedule

Measure weekly or monthly with the same set, and log what you changed in between. The trend on each theme and each assistant is what tells you whether your content and technical work is paying off.