Publishers now need to know how often AI answers cite their pages, and the first-party reports that show this count different things in different ways. AI search citation reporting means the reports a search provider gives site owners about how their pages appear as links or sources in AI-generated answers, such as Google Search Console's Generative AI performance reports and the AI Performance report in Bing Webmaster Tools. This tracker sets out, from each provider's own help pages and announcements, what each report covers, which metrics it offers, how those metrics are counted, which surfaces are included and what the provider says the numbers do not mean. It is written for SEO and content teams, analysts, publishers and marketing leaders who need to report AI visibility accurately, and for security and privacy teams who manage the webmaster accounts that hold this data. Providers are listed alphabetically; order is never a ranking.
Last checked: 2026-11-15.
This is an independent reference page from NHI Mgmt Group based on public provider documentation; it is not sponsored or endorsed by any provider.
Key takeaways
- Two providers publish dedicated AI citation reports for site owners. Google offers Generative AI performance reports for Search and for Discover in Search Console, and Microsoft offers AI Performance in Bing Webmaster Tools.
- The headline metric differs. Google's generative AI reports describe impressions of links to your site in AI features; Bing's AI Performance counts citations, cited pages and the grounding queries behind them.
- AI traffic is also inside the classic reports. Google states that AI features are included in overall search traffic in the Performance report under the "Web" search type, with defined rules for clicks and position in AI Overviews and AI Mode.
- Both providers warn against over-reading the numbers. Microsoft says grounding query data is a sample and Citation Share is "not a ranking system"; Google notes chart and table totals can differ because of aggregation rules.
- Coverage is limited to each provider's own surfaces. Google covers AI Overviews and AI Mode (and generative AI features in Discover); Microsoft covers Copilot, AI-generated summaries in Bing and select partner integrations.
Which first-party AI citation reports exist?
The table gives one row per report with identical fields. Status and dates are as printed on the provider's page.
Table as of 15 November 2026.
| Report | Provider and tool | Status | AI surfaces covered | Headline metric |
|---|---|---|---|---|
| AI Performance | Microsoft, Bing Webmaster Tools | Public preview announced 10 February 2026; further insights in preview from 16 June 2026; Topics Portfolio beta from 13 October 2026 | Microsoft Copilot, AI-generated summaries in Bing, select partner integrations | Total citations |
| Generative AI performance report (Discover) | Google, Search Console | Rolled out to all websites worldwide as of 31 August 2026 | Generative AI features in Google Discover | Impressions |
| Generative AI performance report (Search) | Google, Search Console | Introduced June 2026; rolled out to all websites worldwide as of 31 August 2026 | AI Overviews and AI Mode | Impressions |
| Search performance report, "Web" search type | Google, Search Console | Generally available | All Search traffic, including AI features | Clicks, impressions, CTR, position |
OpenAI's crawler documentation explains how OAI-SearchBot controls whether a site appears in ChatGPT search answers, and does not describe a publisher citation report. For the controls that decide whether you appear at all, see the AI Crawler and Grounding Controls Tracker.
Google Search Console: Generative AI performance reports
Google publishes two generative AI reports, one for Search and one for Discover, plus counting rules for AI features inside the main Search performance report.
Search report
- Surfaces: AI Overviews and AI Mode. Google says it expects to update this list over time. Search Labs experiments are excluded.
- Metric: impressions, defined as how many times links to your site were shown to a user in a generative AI feature on Google Search.
- Aggregation: the chart aggregates by property, so two results from one site in the same feature count as one impression. In the table, country, device and date rows aggregate by property while page rows aggregate by page, which is why totals can differ.
- Dimensions: Pages (the canonical URL), Countries, Dates (Pacific Time) and Devices (desktop, tablet, mobile).
- Search type filter: "Web: text-based" for standard text queries and "Web: multimodal" for image-based searches, including Google Lens and Circle to Search.
- Freshness: the newest data may be preliminary, shown as a dotted line, and may change in the next few hours.
- Eligibility: the report appears only when a site has enough impressions in generative AI features, and may be missing if the site is excluded from those features.
- Limits and export: the Search performance report limits apply, including the 1,000-row table limit. Chart and table data can be exported; cells shown as "~" or "-" export as zeros.
Discover report
- Surface: generative AI features in Google Discover.
- Metric: impressions, counted when links are scrolled into view; only one impression is counted per result per session, and two generative AI results from the same property in one Discover list count separately.
- Dimensions: Pages (canonical URL), Countries and Dates (day, week or month, Pacific Time).
- Limits: the Discover performance report limits apply, including 1,000 table rows; daily values below a threshold may be hidden on the chart but still count toward totals.
AI features in the main Search performance report
Google's Search Central page on AI features states that AI features are included in overall search traffic in Search Console, under the "Web" search type. The Search Console help centre defines how they are counted:
| Surface | Click | Impression | Position |
|---|---|---|---|
| AI Mode | Clicking a link to an external page in AI Mode | Standard impression rules | Same methodology as a Google Search results page; carousels and image blocks use their standard rules. Each follow-up question is treated as a new query |
| AI Overviews | Clicking a link to an external page in the AI Overview | Standard impression rules | The AI Overview occupies a single position and all its links share that position |
For a wider explanation of how Google's AI surfaces select and show links, see Google AI Mode Explained.
Bing Webmaster Tools: AI Performance
Microsoft's report counts citations rather than impressions, and adds query-level and topic-level views.
Core metrics (public preview, 10 February 2026)
- Surfaces: Microsoft Copilot, AI-generated summaries in Bing and select partner integrations.
- Total Citations: how often your content is displayed as a source in AI answers over the selected period; it does not show placement within an answer.
- Average Cited Pages: the average number of unique pages from your site displayed as sources per day, aggregated across surfaces; Microsoft says it does not indicate ranking or authority.
- Grounding queries: the key phrases the AI used to retrieve your referenced content. Microsoft states the data shown is a sample of overall citation activity.
- Page-level citation activity: citation counts for individual URLs, reflecting how often pages are cited, not their importance, ranking or placement.
- Visibility trends: a timeline of citation activity across supported AI experiences.
A March 2026 Microsoft Advertising post adds Grounding Query to Page Mapping, which links each grounding query to the pages cited for it and each page to its queries.
Insights added in preview (16 June 2026)
| Insight | What it measures | What Microsoft says it is not |
|---|---|---|
| Citation Share | The percentage of citations attributed to your site out of all citations shown across all sites for a grounding query | "not a ranking system or a competitive scoreboard"; it does not reveal competitor domains, traffic share or quality scores |
| Compare | Overlays a previous period on the current view, such as the last 30 days against the prior 30 days, or custom ranges | It shows what changed, not why |
| Intents | Grounding queries classified into categories such as Informational, Commercial, Navigational, Learn and Solve, Research, Creation and Local | Labels may be broad in specialised or niche domains during preview |
| Topics | Related grounding queries grouped into thematic clusters | Classification precision is expected to improve as models mature |
Topics Portfolio (beta, 13 October 2026)
- Topic table: columns for Topic, Topic Size, Sites Cited and Queries, over a one-, two- or three-month window.
- Opportunity queries: grounding queries where other sites are cited and yours is not, with citation volumes shown as ranges.
- Topic detail pages: three KPI cards plus Your Grounding Queries and Opportunity Queries tables.
- Caveats: Microsoft asks users to treat the numbers "as directional signals for planning, not as a precise ledger", and notes that AI classification can be imperfect, topics with too little data are hidden, and branded queries can appear as opportunities.
How do the reports compare field by field?
This table places the same fields side by side so analysts can label dashboards correctly. It describes what each report counts; it does not judge them.
Table as of 15 November 2026.
| Field | Bing Webmaster Tools AI Performance | Google Search Console Generative AI report (Search) |
|---|---|---|
| Unit counted | Citations (content displayed as a source) | Impressions (links to your site shown in a generative AI feature) |
| Query data | Grounding queries (sampled), with intents and topics | Not listed among the report's dimensions |
| Page data | Page-level citation activity; query to page mapping | Pages dimension (canonical URL) |
| Other dimensions | Topic, intent, period comparison | Countries, dates, devices, text-based or multimodal search type |
| Share metric | Citation Share per grounding query | Not stated by Google for this report |
| Date window | One, two or three months in Topics Portfolio; custom ranges in Compare | Changeable date ranges and granularity |
| Export | Not stated by Microsoft in the posts used here | Chart and table export |
| Provider caveat | Sampled data; directional, not a ledger | Preliminary recent data; chart and table totals can differ |
How to build an AI visibility report from these sources
The steps below turn the first-party data into a monthly report that states clearly what each number is.
- Verify ownership: confirm the site is verified in both Google Search Console and Bing Webmaster Tools, using the domain property where possible so all subdomains and protocols are included.
- Check eligibility first: confirm key pages are indexed and snippet-eligible, and that nosnippet, data-nosnippet, max-snippet or noindex are only where intended. Google names these as the controls that limit what is shown from your pages.
- Pull Google AI impressions: export the Generative AI performance report (Search) by page and by country, keeping text-based and multimodal search types separate.
- Pull Google clicks: from the main Search performance report ("Web"), remembering that AI Overview and AI Mode clicks are mixed with other Search clicks there.
- Pull Bing citations: export total citations, cited pages and grounding queries, then note Citation Share for the grounding queries that matter most to the business.
- Label units: keep "Google AI impressions" and "Bing citations" as separate series. Adding them together mixes two different units.
- Add outcomes: Google suggests measuring conversions and time on site in analytics tools; join those to landing pages from the reports.
- Annotate changes: record dates of content releases, robots.txt edits and provider announcements, so a change in the trend line can be read against known events.
Metrics worth reporting
- Google AI impressions by page: month-on-month change for your priority pages.
- Bing total citations and cited pages: trend over the selected period.
- Citation Share on priority grounding queries: as an observational measure, as Microsoft describes it.
- Opportunity queries in Topics Portfolio: count of queries where other sites are cited and yours is not.
- Conversions from AI-influenced landing pages: from your analytics tool, joined by URL.
Common failure patterns
- Adding impressions to citations: Google counts impressions of links and Bing counts citations; a single combined "AI visibility" number has no defined meaning.
- Treating Citation Share as rank: Microsoft states it is not a ranking system or competitive scoreboard.
- Reading sampled data as exact: Bing grounding query data is a sample, and Topics Portfolio numbers are directional.
- Expecting position inside an AI Overview: all links in an AI Overview share one position in the Search performance report.
- Blocking the wrong control: a training opt-out and a search or grounding opt-out are separate switches; blocking the wrong one can remove pages from answers. Microsoft's new ai-training and ai-grounding preferences take effect from 1 February 2027, and NoArchive and NoCache stop being read for training or grounding after 1 March 2027.
- Weak access control on webmaster accounts: these consoles also hold removal and crawl tools, so a shared or unmanaged login is a real risk to search presence.
Where identity and access fit
Search Console and Bing Webmaster Tools accounts carry powers beyond reporting, including URL removals, sitemap submission and, in Bing's case from 15 December 2026, temporary grounding preferences for pages that used NoArchive or NoCache. Give each analyst a named account, use the lowest permission level that lets them read reports, remove users when they leave, and protect owner accounts with phishing-resistant MFA. Service accounts or API keys used to pull reports into dashboards are non-human identities: store them in a secrets manager and review who can use them. See the AI Crawler and Scraper Controls guide for the crawl side of the same decision.
Practitioner checklist
- Verify the site as a domain property in Google Search Console and in Bing Webmaster Tools.
- Check that the Generative AI performance reports for Search and Discover appear; if not, confirm the site has enough AI impressions and is not excluded from AI features.
- Export Google AI impressions and Bing citations monthly and keep them as separately labelled series.
- Track Citation Share and opportunity queries for a fixed list of priority grounding queries.
- Record robots.txt, meta robots and X-Robots-Tag changes with dates beside the trend lines.
- Review Bing ai-training and ai-grounding preferences before 1 March 2027 if pages use NoArchive or NoCache.
- Audit users and permission levels on both webmaster consoles each quarter.
- Recheck each report's documented definitions before quarterly reporting, since preview features change.
Standards and references
- Google Search Console Help: Generative AI performance report (Search)
- Google Search Console Help: Generative AI performance report (Discover)
- Google Search Console Help: Performance report (Search results)
- Google Search Central: AI features and your website
- Microsoft Bing: Introducing AI Performance in Bing Webmaster Tools Public Preview
- Microsoft Bing: New AI Visibility Insights in Bing Webmaster Tools
- Microsoft Bing: Introducing Topics Portfolio beta in AI Performance
- Microsoft Bing: Clearer preferences for training and grounding
Also checked on 2026-11-15: Microsoft Advertising: The AI Performance dashboard, OpenAI: Overview of OpenAI crawlers.
Related NHI Mgmt Group resources: AI Answer Engines Hub · What Is Answer Engine Optimisation (AEO)? Definition, Signals and Measurement · What Is Generative Engine Optimisation (GEO)? Definition, Research and Limits · Ads in AI Answers Tracker · Microsoft Copilot Web Answers Explained