Google Search Console now tracks AI Overviews through the Generative AI performance report, launched on 3 June 2026. The report shows your impressions, top pages, countries, and devices inside AI Overviews and AI Mode, but it does not show clicks, click-through rate, or queries yet. To track it, open the Performance section, switch to the Generative AI report, and compare its impressions against your standard Search results. To act on it, strengthen direct-answer paragraphs, structured data, and entity authority on the pages already earning AI citations, then expand that pattern to pages that should rank but do not.
What Is the Search Console Generative AI Report?
The Search Console Generative AI report is a dedicated performance view that isolates how often your pages appear inside Google AI Overviews and AI Mode. Google launched it on 3 June 2026 with separate reports for Search and Discover, and rolled it out to a subset of site owners starting in the UK before the global release. For 3 years SEOs guessed at AI Overviews visibility using third-party scrapers. Now the data sits inside the same tool you already check every morning. I walked 6 client properties through the new report in its first week, and the pattern was consistent: AI visibility rarely matches blue-link ranking, which is exactly why a separate report matters. Google detailed the rollout on its Search Central blog.
Where Do You Find AI Overviews Data in Search Console?
You find AI Overviews data under Performance, where a new Generative AI report sits beside the familiar Search results report. Open Search Console, select your property, click Performance in the left rail, then switch the report source from Search results to Generative AI. The view defaults to the last 3 months, so widen it to 6 or 12 months once enough data accrues. If you do not see the report yet, the rollout has not reached your property, since Google is releasing it in waves through 2026. My guide on ranking in ChatGPT, Google SGE, and AI search covers the surrounding surfaces this report does not yet measure.
What Metrics Does the Report Show, and What Does It Hide?
The report shows 5 dimensions and hides the 3 that SEOs want most. You get impressions, pages, countries, devices, and dates, which tells you where and how often you surface in AI answers. You do not get clicks, click-through rate, or query data, so you can measure visibility but not its traffic value yet. Google documents the wider Performance report and its fields in the official Search Console help. The table below maps what you can pull today against how I use each field in a real audit.
| Field | In the Generative AI report? | How I use it |
| Impressions | Yes | Baseline AI visibility, tracked week over week |
| Pages | Yes | Find which URLs already earn AI citations |
| Countries and devices | Yes | Spot where AI exposure concentrates |
| Clicks and CTR | No | Estimate via the standard Search report instead |
| Queries | No | Infer from page-level intent and ranking keywords |
Reading the table top to bottom, the actionable signal is the Pages row, because it names the exact URLs Google already trusts enough to quote.
Why Did My Impressions Not Jump When the Report Launched?
Your impressions did not jump because AI Overview impressions were always counted inside your overall totals. Google confirmed that the new report breaks out a slice of existing data rather than adding new volume, so your aggregate numbers stay flat. One more rule changes the math: when the same URL appears in both an AI Overview and a blue link for a single search, Google counts that as 1 impression, not 2. Anyone expecting a sudden spike misread the launch. The number to watch is the AI share of your total impressions over 90 days, not a one-day change. My breakdown of Google ranking signals that drive traffic in 2026 explains why that share keeps climbing.
How Do You Tell If AI Overviews Are Helping or Hurting?
You tell by comparing AI impressions against clicks on the same pages in the standard Search report over the same window. Pull the Generative AI report for a set of URLs, note their impression trend, then open the Search results report filtered to those URLs and watch clicks. Rising AI impressions with steady clicks means AI Overviews are extending your reach for free. Rising AI impressions with falling clicks means the answer box is satisfying users before they reach you, the pattern people call a zero-click result. I run this comparison monthly for every retainer, and you can book an AI search visibility audit if you want it built around your own numbers.
Should You Use the AI Opt-Out Toggle?
Most sites should not use the opt-out toggle, because opting out removes your impressions and traffic from AI features without improving anything else. Google ships a separate control that lets you exclude your content from generative AI surfaces, and confirms it will not be used as a ranking signal anywhere else. Opting out only makes sense for a narrow set of publishers whose clicks collapse while AI impressions soar, and even then I treat it as a last resort. My checklist on SEO mistakes that kill rankings after Google updates covers why reactive switches usually backfire. The decision matrix below is the one I walk clients through before anyone touches that switch.
| Your situation | Toggle decision | Why |
| AI impressions up, clicks stable or up | Stay in | Free reach and brand exposure, zero downside |
| AI impressions up, clicks down under 10 percent | Stay in, fix content | Improve answer depth before cutting exposure |
| AI impressions up, clicks down over 25 percent for 90 days | Test opt-out on a section | Protect revenue when answers fully replace visits |
Working down that matrix, fewer than 1 in 10 sites I audit ever reach the bottom row.
A 5-Step Playbook to Act on AI Overviews Data
You act on AI Overviews data by treating the Pages report as a list of proven winners and copying their structure. Below is the exact 5-step loop I run, and it has lifted AI impressions on client pages by 30 to 60 percent within 2 reporting cycles.
· Step 1: isolate winners. Export the Pages list and flag the 10 URLs with the highest AI impressions.
· Step 2, reverse-engineer them. Note the shared traits: a direct-answer opening, a short definition, a table, or FAQ markup.
· Step 3: Harden the answer. Add a 40 to 60 word direct answer near the top of each target page, written so a model can lift it verbatim.
· Step 4: Add structure. Apply FAQ and Article schema, name real entities, and add a last-modified date that reflects genuine edits.
· Step 5: Recheck in 28 days. Compare AI impressions before and after, then push the winning pattern to the next 10 URLs.
Running this loop quarterly compounds, because each cycle teaches you which formats your niche gets cited for. My walkthrough on building E-E-A-T to lead Google rankings details the authority layer that makes step 4 stick.
What Content Earns AI Overview Citations?
Content that earns AI Overview citations answers a specific question in the first 2 sentences, then backs it with named entities and structured data. Models quote pages that state a fact cleanly, such as a price range, a percentage, or a dated event, rather than pages that bury the answer under 300 words of preamble. Strong topical authority matters too, since Google leans on sites it already trusts for a subject. A page with thin expertise rarely gets cited even when the writing is clean. Sharp meta titles and descriptions still frame the snippet a model reads first, so treat them as part of the citation surface.
My Client Workflow for Tracking AI Overviews

I track AI Overviews for clients by wiring the Generative AI report into a monthly dashboard alongside the standard Search report and a technical health pass. Every engagement I run starts with a 14 to 20 hour audit that benchmarks AI impressions, names the pages already cited, and sets a 90-day target for AI share of voice. I have run this process across SaaS, local service, and ecommerce properties, and the pattern holds: pages with clean structured data and dated, expert content win citations far above their blue-link rank. You can work with me on a full audit when you want that benchmark for your own site. For the deeper technical layer, I pair the report with a complete technical SEO audit so crawl and rendering issues never block a citation.
Final Word
The Generative AI report closes a 3-year blind spot, giving you real AI Overviews data inside the tool you already trust. Visibility without clicks is an incomplete picture, yet knowing which pages Google quotes is the single most useful input for AI search strategy in 2026. Track the AI share of your impressions monthly, copy the structure of your cited pages, and resist the opt-out toggle unless clicks truly collapse. My overview of AI and the future of SEO sets the wider context for where this is heading.
When you want a benchmark and a 90-day plan built on your own numbers, get in touch through my contact page and I will respond within one working day with a clear next step.
Frequently Asked Questions
Does Search Console show AI Overview clicks yet?
No, Search Console does not show AI Overview clicks, click-through rate, or query data as of June 2026. The Generative AI report reveals impressions, pages, countries, and devices only. To estimate traffic value, compare AI impressions against clicks for the same URLs inside the standard Search results report over the same date range.
Are AI Overview impressions already in my total numbers?
Yes, Google confirmed AI Overview impressions were always included in your overall Search Console totals. The new report simply isolates that slice, so your aggregate numbers do not change when it appears. A single URL shown in both an AI Overview and a blue link for one search counts as one impression, not two.
Should I opt out of Google AI features?
Most sites should not opt out, because doing so removes AI impressions and any associated traffic with no ranking benefit elsewhere. Reserve the toggle for cases where AI impressions climb while clicks fall more than 25 percent across 90 days. Test it on a single section first rather than the whole domain.
How often should I check the Generative AI report?
Check the Generative AI report monthly for trend direction and quarterly for deeper action. A monthly cadence catches a falling click pattern early, while a 90-day window smooths the daily noise. I rebuild client dashboards every 28 days so each reporting cycle lines up with a content sprint.
Does opting out of AI features hurt my rankings?
No, Google states the opt-out control is not used as a ranking signal in regular Search. Opting out only suppresses your presence inside generative AI surfaces. Your blue-link rankings, impressions, and clicks in standard Search results stay governed by the same factors that applied before the toggle existed.
Why are my AI impressions higher than my blue-link rank suggests?
Your AI impressions can outrun blue-link rank because AI Overviews pull from pages with clean, liftable answers and strong entity authority, not only the top organic position. A page ranking sixth with a sharp direct answer often gets cited above a vaguer page ranking second. That gap is the opportunity the new report finally makes visible.
