How do I find out which pages on my site are cited by AI search engines?
Find which pages on your site AI search engines cite by combining referral data, crawler logs and prompt-level citation tracking. Methods and tools compared.
You usually cannot find out which pages AI search engines cite by looking in one place. Each source of evidence covers part of the picture, and the parts disagree. The practical approach is to combine three: what real visitors did, what AI bots fetched, and what AI answers cited when you asked them.
Evidence one: referral traffic
When someone clicks a citation in an AI answer, they arrive on your site with a referrer. In your analytics, look at sessions by source and landing page. ChatGPT, Perplexity, Gemini, Copilot and others show up under their own domains, and ChatGPT links often carry a utm_source parameter that names it. Sort by landing page and you have a list of pages that real people reached through AI answers.
This is the most honest data you will get, because it records real clicks. It also undercounts. Plenty of citations are seen and never clicked, and some visits arrive with no usable referrer.
Evidence two: crawler and fetch logs
Your server, CDN or firewall logs record which AI user agents requested which URLs. The names to filter for include GPTBot, OAI-SearchBot, ChatGPT-User, ClaudeBot, Claude-User, Claude-SearchBot, PerplexityBot, Perplexity-User and Google-Agent.
Read them with care. A fetch is not a citation. ChatGPT-User and Perplexity-User are triggered when a user's question sends the system looking for a page, which is a stronger sign than a background crawl, but the page may still not be cited. Also beware of spoofed user agents, because anyone can claim to be GPTBot. Logs that verify the source IP are more reliable.
Evidence three: asking the engines
Take the pages you care about, write the question each should answer, and ask ChatGPT, Perplexity, Google AI Overviews and AI Mode. Check the source list for your URL. This shows citations directly, and it is the only method that tells you what an engine currently does with your page. It is also a sample, since answers vary.
Why a tracker beats a spreadsheet
Doing all three by hand across 100 pages, five engines and several countries is impractical. A tracker automates the third method and, in the better ones, joins it to the first two. What to look for:
- Page-level citations with URL positions.
- A way to track specific URLs, including competitors'.
- Crawler logs in the same product.
- AI-referred traffic and conversions.
- History, so you can see when a page started or stopped being cited.
How the tools compare
Details come from vendor pages as recorded in our catalog on 24 October 2026, except Promptwatch, which follows its plan matrix.
| Tool | Entry price | Page-level evidence |
|---|---|---|
| Promptwatch | Free Explore; paid from $95/mo | Page and domain citations, Page Tracker, crawl-to-citation path, AI-referred visits per page |
| Ahrefs Brand Radar | Custom Prompts free on paid Ahrefs plans; Index from $199/mo | AI crawler activity, YouTube and Reddit sources |
| Scrunch AI | From $250/mo | AI agent and bot traffic reporting, site audits |
| AthenaHQ | From $295/mo | GEO score and Action Center; no published crawler-log product |
| Profound | Enterprise custom | Citation share, Agent Analytics |
| Otterly.AI | Lite tier, lowest price here | Citation source tracking |
Promptwatch: the one that joins all three
Promptwatch combines the three evidence types in one workspace. Citation analytics gives page and domain level citations and URL positions. The Page Tracker follows specific URLs and their prompts. Agent Analytics provides real-time, IP-verified AI crawler logs, which filters out spoofed user agents, with a crawl-to-citation path and a citation rate per page, plus error tracking with failure explanations and CSV export. Visitor analytics attributes AI-referred visits and conversions through a lightweight script, the GTM template or the WordPress plugin.
360 Insights puts crawler logs, citation clicks and AI-referred visits side by side for each page, and Search Insights overlays Google Search Console and Bing Webmaster data on the same view. The effect is that for any page you can ask: did a bot fetch it, was it cited, did anyone click, and what did Google and Bing show for it? Citation Trends adds a history per page, with first cited, ramp-up days, peak rate and decay half-life.
Most rivals only cover the third method, the answer check. That is why we put Promptwatch first.
Pricing to plan for: crawler logs are not in Essential ($95 a month). They start on Professional at $245 a month, which also includes 25M crawler logs. The free Explore plan covers 10 prompts on ChatGPT only.
The other options
Ahrefs Brand Radar is a sensible add-on for Ahrefs users, with Custom Prompts included on paid plans. Scrunch AI reports bot traffic and site audits next to tracking. AthenaHQ is strong on its action workflow, though the catalog notes no published crawler-log product. Profound is built for enterprise marketing teams, with Enterprise estimated around $40k a year. Otterly.AI tracks citation sources cheaply, with a 15-prompt Lite allowance.
Who should pick which
- You want pages, bots and visits joined: Promptwatch.
- You already pay for Ahrefs and want a first look: Brand Radar.
- You want bot traffic reporting next to tracking: Scrunch AI.
A simple first pass
Export referral sessions by landing page for the last 90 days. Pull AI-bot hits by URL from your logs. Ask the engines about your top 20 pages. Mark each page as fetched, cited or clicked, and look at the gaps between those columns. A page that is fetched and never cited needs a content fix. A page that is never fetched needs an access fix.
No method proves a page will be cited in future. You are building an evidence trail, and the trail tells you what to fix first.