AI Visibility: How to Check Whether ChatGPT and Gemini Mention Your Brand

Short answer: AI visibility is how often and how accurately AI assistants mention your brand. Track it with three layers: a prompt set of 30–150 real customer questions tested regularly in ChatGPT, Gemini, Perplexity, Claude, Copilot and Google; the official platform reports (Google Search Console's generative AI report and Bing Webmaster Tools' AI Performance report); and AI referral traffic and leads in your analytics. For each prompt, record whether you're mentioned, cited, accurately described, and who else appears.

This guide gives you a practical system and a template. It's the measurement side of our generative engine optimization guide. Want us to run it for you? See our AI search optimization service.

What is AI visibility?

AI visibility is how often, and how accurately, AI assistants mention your brand when people ask questions you should be the answer to. It covers four things:

• Mentions: does ChatGPT, Gemini, Perplexity, Claude, Copilot or Google AI name your business?
• Citations: does the answer link to your website as a source?
• Accuracy: are your services, prices, locations and details described correctly?
• Share of voice: how often you appear compared with competitors. See AI share of voice.

How to check your AI visibility in 10 minutes

For a quick first look before setting up full tracking:

1. Write down five questions a customer might ask, such as "best [service] in [town]" or "[your brand] reviews".
2. Ask each one in ChatGPT (with search on), Gemini, Perplexity, Copilot and Google AI Mode.
3. Note whether you're mentioned, whether your site is cited and whether the details are right.
4. Note which competitors and sources appear instead.

If you're missing, or the details are wrong, the rest of this guide explains how to track it properly and what to fix.

Why AI visibility is hard to measure

• Answers vary between sessions, users, locations and wording.
• Assistants don't share query data the way search engines do (with some new exceptions below).
• Many answers don't include links, so you can be mentioned without getting a click.
• Model updates can shift answers suddenly.

That's why you need repeatable, structured tracking rather than occasional spot checks.

Layer 1: Build a prompt set

Step 1: Collect real questions

Sources:

• Sales and support conversations ("customers always ask…")
• Google Search Console queries, especially long questions
• People Also Ask boxes and related searches
• Reviews (what people care about)
• Forums and Reddit threads in your niche

Step 2: Organise by intent

Aim for 30 prompts for a small local business, 100+ for B2B or e-commerce.

Step 3: Choose assistants

At minimum: ChatGPT, Google (AI Overviews and AI Mode), Gemini, Perplexity and Copilot. Add Claude if your audience uses it.

Step 4: Run tests consistently

• Use logged-out sessions or temporary chats where possible, to reduce personalization.
• Use the same wording each time.
• Note the date, assistant and (for local prompts) location.
• Run important prompts two or three times per session.

Step 5: Record results

Track these fields for each prompt and assistant:

A simple spreadsheet works. Calculate monthly mention rate (% of prompts where you appear) and share of voice versus each competitor.

Layer 2: Use the official reports

Google Search Console: Generative AI performance report. Shows impressions when links to your site appear in AI Overviews and AI Mode, by page, country, device and date. Google says it rolled out to all sites in 2026; sites with few AI impressions may not see it yet. There's also a Discover version.

Bing Webmaster Tools: AI Performance. Shows when your pages are cited in Microsoft Copilot and Bing's AI-generated answers, which URLs are cited, and the "grounding queries" used to retrieve your content.

These are the only first-party data sources for AI visibility right now, so set them up first.

Layer 3: Measure AI referral traffic and leads

In Google Analytics 4 (or your analytics tool):

1. Create a custom channel group or segment for AI assistants, matching referrers such as chatgpt.com, perplexity.ai, gemini.google.com, copilot.microsoft.com and claude.ai.
2. Track conversions from that segment: leads, sales, sign-ups.
3. Add "How did you hear about us?" to forms with an option for "ChatGPT or another AI assistant".

Some AI-driven visits arrive without a referrer and show as direct traffic, so treat referral numbers as a minimum.

Turning tracking into action

Use your data to decide what to fix:

See how to write content AI search engines cite and how to get recommended by ChatGPT.

Should you use an AI visibility tool?

Third-party tools can automate prompt runs at scale and chart share of voice. They're useful for larger prompt sets. Google's guidance cautions that no third-party tool has access to its internal ranking or AI systems, so judge tools on transparency: can you see the prompts, raw answers and methodology?

A monthly AI visibility report template

1. Summary: mention rate and share of voice vs last month
2. By assistant: ChatGPT, Gemini, AI Overviews/AI Mode, Perplexity, Copilot, Claude
3. By intent: discovery, local, comparison, pricing, brand
4. Accuracy issues found and fixes planned
5. Official reports: Search Console AI impressions, Bing AI citations
6. Business impact: AI referral sessions, leads, sales
7. Actions for next month

For context on how fast AI search is growing, see our AI search statistics.

A sample tracking sheet

Summarise monthly: mention rate by assistant, share of voice vs each competitor, accuracy issues found and fixed.

A worked example

A UK conveyancing firm tracks 60 prompts monthly. In month one it's mentioned in 12% of answers, and three assistants list an old office address. It fixes listings, adds fee information to its conveyancing pages and earns reviews on two legal review sites. By month four its mention rate across the same prompts is noticeably higher, the address errors are gone, and AI referral enquiries appear in its CRM for the first time. (Illustrative.)

Sampling tips for reliable results

• Run high-value prompts two or three times per session and record the majority result.
• Use the same location settings for local prompts.
• Keep prompt wording fixed; add new prompts separately rather than editing old ones.
• Record the model or mode where visible (for example, search on vs off).
• Note major model updates in your sheet; they can explain sudden shifts.

Who should own AI visibility tracking?

Usually whoever owns SEO or content, with input from PR and customer teams. The data is most useful when it drives action: content updates, listing fixes, PR priorities and review campaigns.

The bottom line

AI visibility can be measured well enough to manage. Build a prompt set, test it consistently across assistants, use Search Console and Bing's new AI reports, and connect it all to referral traffic and leads. Review monthly and act on what you find.

Want a baseline AI visibility report for your business? Request a free growth audit.