Prompt
How do I audit which pages are showing up in AI answers?
Latest observation
To audit which of your pages are showing up in AI answers, you need to track AI citations, referrals, and search visibility in AI surfaces. There isn’t one perfect tool yet, so the best approach is to combine several methods.
1) Check analytics for AI referral traffic
Start in GA4, Adobe, or your analytics platform and look for traffic from:
chatgpt.comperplexity.aicopilot.microsoft.comgemini.google.comclaude.aiopenai.com/chat.openai.comwhere applicable- AI search/referral sources in Bing or Google properties
What to look for
- Landing pages receiving sessions from AI referrers
- Engagement and conversion rates for those sessions
- New or unusual pages getting traffic after AI model changes
Tip
Create a segment or custom exploration for:
- Session source/medium containing AI domains
- Landing page dimension
- Conversions / revenue / key events
2) Use AI visibility tools
Several SEO platforms now track how often brands and URLs appear in AI-generated answers. Depending on the vendor, this may be called:
- AI visibility
- Answer engine optimization (AEO)
- Generative engine optimization (GEO)
- AI citations
- LLM visibility
Common tool categories include:
- Brand monitoring platforms
- SEO suites adding AI tracking
- Dedicated AI answer tracking tools
These tools typically show:
- Which prompts triggered your brand/page
- Whether you were cited
- Share of voice vs competitors
- Citation frequency over time
3) Manually test high-value prompts
Build a list of questions your audience would ask, then test them in:
- ChatGPT
- Perplexity
- Gemini
- Copilot
- Google AI Overviews / AI Mode where available
For each prompt, record:
- Query
- Tool used
- Date
- Whether your page/brand is mentioned
- Whether there’s a citation/link
- Which URL is cited
- Competitors cited instead
Best practice
Use the same prompts repeatedly over time, because AI answers change.
4) Audit your content pages most likely to be cited
Look at pages that are:
- Clear, factual, and specific
- Product pages with specs
- FAQ pages
- Comparison pages
- Glossary/how-to pages
- Research/statistics pages
- Pages with unique data or authoritative claims
These are often the pages AI systems prefer to cite.
5) Cross-check with search console data
AI answers often pull from pages that already rank well in search. In:
- Google Search Console
- Bing Webmaster Tools
Check:
- Queries with rising impressions but modest clicks
- Pages with strong impressions for informational queries
- Pages that rank for question-based searches
If a page performs well in classic search, it has a better chance of being surfaced in AI answers.
6) Build a simple audit spreadsheet
Track rows like this:
| Prompt | AI Tool | Date | Brand Mentioned | URL Cited | Result Type | Competitor Cited | Notes |
|---|
Add another sheet for URL-level tracking:
| URL | Topic | AI Citations | Mentions | Referrals | Conversions | Priority |
|---|
7) Watch for “hidden” exposure
Sometimes your page influences the answer even without a visible citation. To capture that:
- Track brand/entity mentions
- Track assisted conversions from AI referral traffic
- Compare prompts before and after content updates
- Use logs or prompt testing if you have a proprietary AI monitoring setup
8) Prioritize by business value
Not every page matters equally. Focus on pages tied to:
- Revenue
- Lead generation
- High-value product categories
- Brand reputation
- FAQs that drive support deflection
Recommended workflow
- Pull AI referral traffic from analytics.
- Identify top landing pages from AI sources.
- Run a prompt set in major AI tools.
- Record which URLs are cited.
- Compare with Search Console/Bing data.
- Update pages that are missing or underrepresented.
If you want a more advanced setup
I can help you build:
- A GA4 audit report
- A prompt testing spreadsheet
- A dashboard for AI citations
- A SQL query to find AI referral landing pages
If you want, I can also give you a step-by-step audit template you can use today.