Gareth Puckey Say hello

How to see if AI answers cite your site

How to check whether AI answers cite your site in 2026, using Search Console's AI report, Bing's AI Performance report and GA4's AI Assistant channel.

In this guide
  1. Check you can be cited at all
  2. Google: the generative AI performance report
  3. Bing: the AI Performance report
  4. GA4: the visits that arrive
  5. Spot checks, done carefully
  6. The three reports side by side
  7. A monthly routine
  8. FAQ

Free tools now show much of whether AI answers cite your site: Search Console reports how often your pages appear as links in Google’s AI Overviews, AI Mode and generative AI features in Discover. Bing Webmaster Tools shows how often Copilot and Bing’s AI summaries cite you, and GA4 has a channel for visits from AI assistants. None of them gives the whole picture, so read them together and judge the result by the enquiries and revenue those visits bring in.

Check you can be cited at all

Before measuring anything, make sure nothing is keeping you out.

For Google, a page has to be indexed and eligible to be shown in Search with a snippet, and Google says “There are no additional technical requirements” (Google Search Central). The same page lists the controls that limit what’s shown: nosnippet, data-nosnippet, max-snippet and noindex. If one of those sits on a page you want cited, that’s the first fix.

Search Console also has a setting under Settings > Search generative AI. Sites are included by default, and excluding one means its content “won’t be eligible to be used as an input to generate an AI response or preview” (Search Console Help). It reached every site on 31 August 2026, so it’s worth a two-minute check that nobody has switched it off.

Other assistants have their own crawlers. OpenAI recommends allowing OAI-SearchBot in robots.txt, and says sites that block it won’t be shown in ChatGPT search answers, while GPTBot is the crawler for model training (OpenAI). Perplexity says PerplexityBot is the crawler that surfaces and links sites in its search results (Perplexity). For Google, training is a separate control, Google-Extended (Search Console Help). Blocking a training crawler and blocking a search crawler are separate decisions.

Google: the generative AI performance report

Google launched generative AI performance reports in Search Console on 3 June 2026, covering AI Overviews, AI Mode and generative AI features in Discover. They reached all sites on 31 August 2026 (Google Search Central Blog).

The report measures impressions: “how many times links to your site were shown to a user in a generative AI feature on Google Search” (Search Console Help). You can break that down by page, country, date and device. There are no clicks, no click-through rate, no position and no queries, so it tells you which pages are being used and where, but not what people did next.

Clicks still exist, just somewhere else. Google says AI feature traffic is reported in the main Performance report under the “Web” search type (Google Search Central). Clicking a link in an AI Overview counts as a click, and the AI Overview takes a single position in the results (Search Console Help). Neither Google page describes a separate figure for AI clicks. The practical answer is to set the pages with AI impressions next to the clicks those same pages earn in the Performance report, month on month.

Bing: the AI Performance report

Bing Webmaster Tools added an AI Performance report as a public preview on 10 February 2026. It shows how your content appears “across Microsoft Copilot, AI-generated summaries in Bing, and select partner integrations”: total citations, cited pages and the grounding queries behind them (Bing Webmaster Blog). Since March you can see which pages were cited for a specific grounding query (Microsoft Advertising Blog), and a Topics Portfolio beta added on 13 October groups those queries into topics (Bing Webmaster Blog).

Microsoft is careful about what the figures mean. The data “represents a sample of overall citation activity” (launch post), and it says to treat the numbers “as directional signals for planning, not as a precise ledger” (Topics Portfolio post). The report lists citations, cited pages and grounding queries, but not clicks. The grounding queries are the useful part: they show the questions where your pages were used, which Search Console’s generative AI report doesn’t give you.

GA4: the visits that arrive

GA4’s default channel group has included an AI Assistant channel since 13 May 2026 (Analytics Help). Google defines it as “the channel by which users arrive at your site from sources like ChatGPT, Gemini, Deepseek, Copilot, or Grok”. It excludes Google’s AI Overviews and AI Mode, which stay in Organic Search (Analytics Help). So a visit from a ChatGPT answer and a visit from an AI Overview land in different channels.

If you want your own definition, custom channel groups live under Admin > Data display > Channel groups. Traffic goes into “the first channel whose definition it matches”, so an AI channel has to sit above Referral, and custom groups can be applied to past data (Analytics Help). Test any pattern against your own list of sources before you trust it.

Some AI visits can show up as Direct. Microsoft’s Clarity documentation notes that “If the source is hidden, the traffic might appear as Direct” (Microsoft Learn), so the AI Assistant figure may undercount.

Then look at what these visits do: the AI Assistant channel’s key events and revenue by landing page, judged the same way as any other channel. The one-page GA4 report is a reasonable place for that line to live.

Spot checks, done carefully

It’s tempting to type your customers’ questions into an AI tool and see who gets named. That’s useful, as long as you know its limits. Google says AI Mode and AI Overviews “may use different models and techniques, so the set of responses and links they show will vary” (Google Search Central). Its optimisation guide also describes query fan-out, “a set of concurrent, related queries generated by the model” (Google Search Central). My view is that one check on one day proves very little.

If you do spot checks, I’d suggest a fixed list of 10 to 20 questions your customers actually ask. Check them on the same day each month, and write down the date, the tool and whether you were cited. Over a few months that shows a direction, even if any single answer varies.

The three reports side by side

Report What it counts Clicks Queries Covers
Search Console generative AI report Impressions of links to your pages No No AI Overviews, AI Mode, generative AI in Discover
Bing AI Performance Citations and cited pages, from a sample Not listed Grounding queries Copilot, Bing AI summaries, some partners
GA4 AI Assistant channel Visits, key events and revenue No (sessions instead) Not shown Sources such as ChatGPT, Gemini, Deepseek, Copilot and Grok
Summarised from Search Console Help, the Bing Webmaster Blog and Analytics Help, as cited above.

A monthly routine

Pulled together, that makes a short monthly routine. SEO vs GEO vs AEO covers how it fits with the rest of your search work.

  1. Search Console: open the generative AI report, note AI impressions by page, then check clicks for the same pages in the Performance report.
  2. Bing Webmaster Tools: note total citations, the pages cited and any new grounding queries.
  3. GA4: check the AI Assistant channel’s sessions, key events and revenue by landing page.
  4. Run your fixed list of spot-check questions and record the results.
  5. Compare with last month and write down what changed. Treat every figure as directional.

FAQ

Can I see clicks from AI Overviews separately in Search Console?

Not according to Google’s documentation. Google reports AI feature traffic inside the main Performance report under the “Web” search type (Google Search Central), and the generative AI report shows impressions only (Search Console Help).

Does GA4 count AI Overview visits as AI Assistant traffic?

No. The AI Assistant channel excludes Google’s AI Overviews and AI Mode, which stay in Organic Search (Analytics Help).

Do I need an llms.txt file to be cited by Google?

No. Google says you don’t need new machine-readable files, AI text files, markup or Markdown to appear in Google Search (Google Search Central).

15 sources, all opened and checked
  1. AI features and your website, Google Search Central
  2. Google’s guide to optimizing for generative AI features on Google Search, Google Search Central
  3. Introducing Search generative AI performance reports in Search Console, Google Search Central Blog
  4. Generative AI performance report (Search), Search Console Help
  5. What are impressions, position, and clicks?, Search Console Help
  6. Search generative AI control, Search Console Help
  7. Introducing AI Performance in Bing Webmaster Tools public preview, Bing Webmaster Blog
  8. The AI Performance dashboard, Microsoft Advertising Blog
  9. Introducing Topics Portfolio (beta) in AI Performance, Bing Webmaster Blog
  10. Default channel group, Analytics Help
  11. What’s new in Google Analytics, Analytics Help
  12. Custom channel groups, Analytics Help
  13. AIPlatform and PaidAIPlatform, Microsoft Learn
  14. Overview of OpenAI crawlers, OpenAI
  15. Perplexity crawlers, Perplexity