Measure AI Citation Share in Microsoft Clarity

By UpGeo · 2026-09-04

Short answer: Clarity measures AI referral sessions, not citations

Microsoft Clarity does not show “AI answer engine citation share” as a native metric. You calculate it by building a segment of AI-referred sessions, then dividing that segment by total sessions. The formula is: AI citation share = AI-referred sessions ÷ total sessions × 100. This is really AI referral click share, because Clarity only records visits after a user clicks a link to your site. It cannot count AI answer citations that never produce a click.

What counts as an AI answer engine in Clarity

Start by defining the referrer domains you want to track. Common AI answer engines appear as browser referrers when a user clicks a citation and lands on your site.

AI source	Clarity filter	Notes
ChatGPT	Referrer domain contains chatgpt.com or chat.openai.com	Often appears as chatgpt.com/c/...
Perplexity	Referrer domain contains perplexity.ai	Includes www.perplexity.ai
Copilot	Referrer domain contains copilot.microsoft.com or bing.com/chat	Bing Chat may still show as bing.com
Gemini	Referrer domain contains gemini.google.com	Gemini app clicks can lose the referrer and appear as direct
Google AI Overviews	No reliable Clarity-only filter	Referrer is usually google.com/search, same as regular organic

Do not confuse these browser referrers with AI crawler user agents such as GPTBot, PerplexityBot, or Google-Extended. Clarity records real visitor sessions, not bot fetches. For crawler visibility, work from your server logs or an AI crawler list.

Step 1: Build an AI answer engine segment

1. Open the correct project in Microsoft Clarity.
2. Click Segments in the left navigation.
3. Select New segment.
4. Add the condition Referrer domain contains and enter chatgpt.com.
5. Add additional referrer conditions for Perplexity, Copilot, and Gemini using OR logic if available.
6. Name the segment AI answer engines and save it.

If your Clarity version does not support multi-value OR logic, create one segment per engine, then sum the session counts manually in a spreadsheet. This is also useful for comparing ChatGPT share versus Perplexity share directly.

Step 2: Track UTM-tagged citations you control

Referrer-only tracking has gaps. Some AI apps strip referrers, and Google AI Overviews cannot be separated from normal Google Search by referrer alone. For links you influence, add UTM parameters such as:

?utm_source=chatgpt&utm_medium=ai&utm_campaign=citation

Then create a Clarity filter for UTM Source or UTM Medium instead of relying only on referrer domain. This works best when you publish AI-friendly URLs yourself. Include trackable URLs in your llms.txt file so AI platforms that respect it are more likely to cite a measurable version of your page. You can create one quickly with the llms.txt generator.

Step 3: Calculate AI citation share

Use these formulas in Clarity or after exporting your session data:

• AI citation share = AI segment sessions ÷ total sessions × 100
• ChatGPT share of AI traffic = ChatGPT sessions ÷ AI segment sessions × 100
• Page-level AI share = page sessions in AI segment ÷ page total sessions × 100

Example: if your site records 10,000 sessions in 30 days and 340 come from the AI answer engine segment, your AI citation click share is 3.4%. If 190 of those sessions come from ChatGPT, then ChatGPT represents 55.9% of your AI-referred traffic.

Step 4: Enrich Clarity with custom AI source data

If you want a cleaner dimension than raw referrer strings, set a custom tag when a visitor arrives from an AI engine. Add this script before the Clarity tracking code or immediately after it:

if (document.referrer.includes('chatgpt.com')) {
  clarity('set', 'ai_source', 'chatgpt');
} else if (document.referrer.includes('perplexity.ai')) {
  clarity('set', 'ai_source', 'perplexity');
} else if (document.referrer.includes('copilot.microsoft.com')) {
  clarity('set', 'ai_source', 'copilot');
} else if (document.referrer.includes('gemini.google.com')) {
  clarity('set', 'ai_source', 'gemini');
}

You can then filter heatmaps, recordings, and dashboards by the ai_source custom tag to see how AI-referred users scroll, click, and convert.

What Clarity cannot do

• Cannot count non-click citations. If ChatGPT cites your brand but the user never clicks, Clarity records nothing.
• Cannot isolate Google AI Overviews reliably. Because the referrer is still google.com/search, Clarity alone cannot separate AI Overview clicks from normal organic clicks.
• Cannot fix referrer loss. Users who copy and paste a URL, use privacy browsers, or click from mobile AI apps may appear as direct traffic.

For true mention-level citation share, pair Clarity with answer-engine monitoring that tracks whether your brand appears in responses at all. That is the difference between on-site click attribution and generative engine optimization visibility. Clarity is strongest as the post-click layer: once an AI user arrives, it shows you exactly what they do.