Generative AI Performance Reports: How to Track Your Impressions Inside AI Overviews and AI Mode

For the past two years, business owners and SEO professionals have been flying blind. Google's AI Overviews and AI Mode were reshaping how people find information online — pulling content from websites, summarising it, and serving it directly in search results — but nobody could actually measure how often their site was featured in those AI-generated answers. That changed on June 3, 2026, when Google launched dedicated Generative AI Performance Reports inside Google Search Console.

This is arguably the most significant Search Console update since the product launched. Here is what it is, what it shows you, and — more importantly — what you should actually do with the data.

Why This Report Exists (and Why It Took So Long)

When Google rolled out AI Overviews in the US in May 2024, it confirmed that clicks and impressions from AI features would count toward standard Search Console performance report totals. That sounds helpful, but the problem was there was no way to separate those AI-driven numbers from your regular organic search data. Everything was blended together, which made it impossible to know whether your impressions were coming from traditional blue links or from AI-generated responses.

The situation got more confusing in September 2025 when false information circulated online suggesting Google had already added an AI Overviews filter to the Performance report. It had not. Publishers were seeing impression counts climb while click-through rates dropped, but had no clean way to attribute that divergence to AI features specifically.

The data points were alarming enough on their own. Research found that organic click-through rates dropped by as much as 61% on queries where AI Overviews appeared. Separately, zero-click searches on Google climbed from 56% to nearly 69% after AI Overviews launched. BrightEdge documented a 49% year-over-year jump in impressions alongside a 30% decline in click-through rates — more visibility, less traffic. And a major Search Console impressions bug that ran from May 2025 onward was inflating the numbers even further, making trend analysis unreliable.

The new Generative AI Performance Reports are Google's answer to years of requests from the publisher and SEO communities. You can now isolate what is happening inside AI features, separately from your standard organic results.

What the New Reports Actually Show

The reports live inside Google Search Console under the Performance section. You can access them by appending /ai to your existing performance report URL. They track five data dimensions:

Impressions — How many times a URL from your site appeared inside a generative AI feature, including AI Overviews, AI Mode in Search, and generative AI features in Discover. If two pages from the same site appeared in a single AI result, that counts as one impression in the chart total. If you add a URL filter, the chart then aggregates by URL instead.

Pages — Which specific URLs on your site are surfacing inside AI features. This is the dimension most actionable for content strategy, because it tells you exactly which pieces of content Google considers citation-worthy.

Countries — A geographic breakdown of where your AI impressions are originating, useful for businesses with regional content or those considering investment in localised pages.

Devices — Desktop versus mobile breakdown. Available for Search results only; the Discover report does not include device data.

Dates — Granularity down to the hourly level, with daily, weekly, and monthly views also available. All dates are reported in Pacific Time.

One important clarification: this data is not siloed away from your existing reports. Generative AI impression data continues to flow into the overall performance report. What you are getting with the new reports is a dedicated, isolated view of that same data — the ability to look at AI visibility on its own terms, without it being diluted by your broader organic numbers.

What the Reports Do Not Show

There is a significant limitation worth calling out directly: there is no click data. The reports track impressions only. You can see how often your URLs appeared inside AI features, but Google is not currently telling you how many users clicked through from those AI responses to your actual website. Query-level data is also absent.

For an industry that has spent two years trying to quantify the traffic impact of AI features, this is a real constraint. Microsoft, by comparison, launched AI Performance reporting inside Bing Webmaster Tools in February 2026 and included query-level citation data. Google's current implementation omits that layer entirely, though Google has said they are continuing to work with website owners and plan to introduce additional metrics over time.

Who Has Access Right Now

The rollout is phased and initially limited to a subset of websites in the United Kingdom, connected to regulatory pressure from the UK's Competition and Markets Authority, which is legally requiring Google to give publishers more control over how their content is used across AI features.

Your site will not see the report if it has not received enough impressions in generative AI features to meet Google's minimum threshold, or if your property is excluded from AI features entirely. The broader global rollout has no confirmed date yet, though Google has indicated it intends to expand access after completing testing with the initial UK cohort.

If your property is eligible, navigate to the Performance section in Search Console. The new AI-specific tab should appear alongside your standard Search and Discover reports.

The Opt-Out Control

Alongside the new reports, Google is testing a toggle that lets website owners choose whether their site appears in AI Overviews, AI Mode, and generative AI features in Discover at all.

Choosing to exclude your site means your URLs will not appear in those features and will not receive any clicks or impressions from them. Google has confirmed this opt-out choice will not be used as a ranking signal for traditional search results, meaning it should not negatively affect your position in standard organic results.

This builds on existing tools like snippet controls (which manage how content displays in traditional results) and Google-Extended (which lets you block content from being used to train Google's AI models). The new toggle is specifically about live generative AI search experiences.

Most site owners should not use the opt-out. Being cited inside AI features is increasingly how brands build authority at the top of the funnel. The exception might be publishers with subscription-based paywalled content, where surfacing answers for free inside AI Overviews directly cannibalises paid readership.

Why This Matters for Your Business

The Scale of AI Search Is No Longer Marginal

AI Overviews now trigger on approximately 48% of all tracked search queries — a 58% year-over-year increase from early 2025 — and that coverage spans over 200 countries across more than 40 languages. For high-volume informational queries, the figure is even higher, with some research placing AI Overview appearance rates at over 80% of that query category.

If your business produces any content that answers questions — about your services, your industry, how things work, what things cost — there is a high probability your content is already being surfaced inside AI responses. The new reports let you confirm that for the first time.

The Impression/Click Disconnect Is Now Visible

The divergence between impressions and clicks is the defining characteristic of the AI search era. When an AI Overview answers a query fully, many users never visit a source website. The new reports let you quantify exactly how much of your impression count is coming from AI features, which in turn helps you understand what your true organic click data actually looks like underneath.

A site receiving ten million total impressions per month might find that two million of those come from AI features where click-through rates are near zero. That changes how you read your traffic trends, how you set benchmarks, and how you evaluate content performance.

Position Data Works Differently in AI Features

There is an important nuance to understand about how position is recorded. AI Overviews have historically assigned position one to all URLs contained within them, regardless of where they actually appeared inside the response. AI Mode uses a different method, assigning positions based on the URL's actual location within the response structure.

Google's John Mueller confirmed that when the same URL appears in both an AI Overview and a standard blue link on the same search results page, Search Console records it as a single impression. Understanding this helps you interpret your position data more accurately and avoid drawing false conclusions about ranking performance.

What to Do With This Data: A Practical Action Plan

1. Establish a Baseline Immediately

Once you have access, capture your current AI impression numbers before making any content changes. Note which pages are generating AI impressions, what countries they are originating from, and what device split looks like. This baseline becomes the benchmark against which you measure every content and technical change going forward.

2. Audit Your Citation-Winning Pages

Look at the Pages dimension and identify which URLs are consistently surfacing in AI results. Study what those pages have in common — content structure, question-and-answer formatting, schema markup, topic specificity, depth of coverage. These characteristics are your template for producing more AI-citation-worthy content across the rest of your site.

If you find that blog content surfaces consistently but your service pages and category pages do not, that gap shows you where topical authority needs to be built out. Service pages that answer the questions buyers actually type into search — costs, timelines, comparisons, alternatives — are often significantly underoptimised for AI citation.

3. Use Geographic Data to Validate Regional Investment

If your AI impressions are skewing heavily toward one geography, verify whether your business is actually structured to serve that market before committing budget to localised content. Conversely, if you are trying to grow in a specific region and seeing low AI impressions there, that is a signal your topical authority in that geography needs work.

4. Compare AI Impression Trends Against Your Overall Traffic

Because AI impression data now sits alongside your standard performance data, you can start building a picture of the relationship between AI visibility and actual site traffic. If AI impressions are climbing while organic clicks remain flat or decline, you are gaining awareness inside AI responses but the answers those responses provide are satisfying user intent without a click. The solution is not to produce less useful content — it is to produce content that establishes your brand as the authority and brings users back for more specific, higher-intent queries.

5. Structure Your Content for Extractability

The pages that earn AI citations share common structural characteristics. Search engines interpreting content for AI responses reward content that is easy for AI to pull clear, quotable answers from. Practically, this means:

Writing direct answers to specific questions within the first few sentences under each heading, not buried three paragraphs deep. Using descriptive subheadings that mirror the natural language of search queries. Adding FAQ sections that explicitly state the question before the answer. Implementing schema markup — particularly FAQPage, HowTo, Article, and Organisation schemas — that signals to Google exactly what your content is and who it is from. Keeping definitions concise and verifiable, with facts that can be attributed and checked.

6. Build Topical Authority, Not Just Individual Pages

AI systems do not just evaluate individual pages in isolation. They evaluate whether a website represents a genuinely authoritative source on a topic. A single well-structured blog post is far less likely to earn consistent AI citations than a cluster of deeply interconnected content that covers a topic from multiple angles — the overview, the specifics, the comparisons, the how-to guides, the FAQs.

Pillar pages supported by cluster content, internally linked and covering the full breadth of a topic, signal the kind of topical authority that AI systems recognise as citation-worthy. This is a content architecture shift, not just a writing improvement.

7. Monitor Competitor AI Visibility

The Generative AI Performance Reports tell you about your own site. Tools like Semrush, Ahrefs, and dedicated AI visibility platforms like Profound and SE Ranking can show you which competing sites are earning AI citations for the queries that matter to your business. Understanding what your competitors' content looks like inside AI responses tells you what you need to match or surpass.

The Bigger Picture: Redefining What Visibility Means

The launch of these reports forces a shift in how businesses think about search performance. For years, the core metric was page ranking. Then it became organic clicks. The AI search era is introducing a third metric: citation presence inside AI-generated responses.

A brand that ranks third organically but is consistently cited inside AI Overviews may be building more authority and trust than the brand sitting at position one. Users who encounter your brand through an AI response arrive with more context, having already received an answer that referenced you as a source. Research suggests those visitors convert at higher rates precisely because they arrive pre-qualified.

This is not about abandoning traditional SEO. The technical foundations — site architecture, crawlability, Core Web Vitals, mobile optimisation, quality backlinks — remain the prerequisite for any AI visibility. Without solid SEO foundations, AI systems have nothing reliable to ingest, understand, or cite. What is changing is the layer above those foundations: the content strategy, the structural choices, and the metrics used to measure success.

Being cited by AI is now a performance metric in its own right. Google has just given you the first tool to measure it.

The Bottom Line

The Generative AI Performance Reports are a genuine step forward, and a long-overdue one. The absence of click data and query-level breakdowns is a real limitation that Google will need to address for the reports to become fully actionable. But page-level AI visibility tracked by country, device, and date is a foundation that did not exist before June 3, 2026.

If you have access to the report now, start building your baseline data this week. If you are waiting for the global rollout, start preparing: audit your content structure, identify your highest-priority pages, implement schema markup if you have not already, and start building out the topical content clusters that AI systems recognise as authoritative.

The businesses that treat AI Overviews and AI Mode as a channel to be optimised — not a threat to be feared — are the ones that will be cited most consistently as search continues to evolve. The data to measure that optimisation is finally available. Now it is time to use it.

TechBuild.me helps small and growing businesses build digital infrastructure that performs. If you want to audit how your current website is positioned for AI search visibility, get in touch with the team.