AI Growth Hub
← AI SEO

Search Console's New Generative AI Reports: Reading the Data and the Opt-Out Question

google-search-console ai-search seo-strategy Google Search Console Google Analytics

What Google Shipped on June 3

On June 3, Google announced generative AI performance reports in Search Console via the Search Central Blog. For the first time, site owners get first-party data on how often their pages surface inside AI Overviews, AI Mode, and other AI-generated search experiences. Until now, DTC teams had to rely on third-party visibility trackers and educated guesswork.

Temper expectations: the report shows impressions only. No clicks, no CTR, and no query-level data. You can see that a product page earned 3,000 AI impressions last month, but not what shoppers asked or whether anyone clicked through.

The dimensional breakdowns are more generous than the metric itself.

DimensionAvailable
ImpressionsYes
Clicks / CTRNo
Query-level dataNo
Page breakdownYes
Country and device breakdownYes
Date granularityHourly to monthly

Treat this as a visibility dashboard, not a performance dashboard. Any attempt to compute ROI from it alone starts from a broken premise.

Reading Impressions Without Clicks

The trend line is the highest-value read. Pull weekly impressions and watch whether your AI-surface visibility is growing or shrinking, especially around content pushes and algorithm updates. That directional signal used to cost a third-party subscription; now it is free and first-party.

Segment by page template next. Product pages, category pages, and editorial comparison pages behave differently in AI answers, and blending them hides everything. A template that consistently earns outsized impressions has found a format the AI systems favor; one that consistently underperforms needs restructured content, not more volume of the same.

Country and device splits matter for brands selling across the US and EU. AI Overviews rollout pace varies by market, so a flat aggregate line can hide a US surge offset by a quiet European market. Read each market on its own curve.

Be equally clear about what the report cannot tell you: without queries, you cannot map impressions to shopper intent, and without clicks, you cannot say whether that visibility converts into sessions.

Pairing It with GA4 LLM Referral Data

The missing half of the funnel lives in GA4. Search Console shows exposure inside AI answers; GA4 referral tracking shows sessions arriving from AI surfaces. We have covered building a custom AI referral channel group in GA4 before, and this report is exactly why that setup pays off.

The workflow: take your top 10 pages by generative AI impressions in Search Console, then check those same pages as landing pages in GA4 filtered to your AI referral channel. Align the date windows and compare week over week. Three patterns typically emerge:

  • Impressions up, AI referrals up: the page is cited and drives real visits. This is a GEO asset; protect it during redesigns.
  • Impressions up, referrals flat: your content is being absorbed into answers with no click-through. Brand exposure still counts, but consider adding assets that require the visit, such as sizing tools or downloadable templates.
  • Impressions flat, referrals up: traffic is likely arriving from surfaces outside this report, such as the Gemini app or ChatGPT. Do not credit Google Search for it.

Hourly granularity has one practical use for commerce teams: during launches and promotions, you can watch how quickly new pages start earning AI impressions instead of waiting on weekly rollups.

The June 17 Opt-Out Toggle

Two weeks after the reports, on June 17, Google shipped a companion toggle that lets site owners block their content from appearing in AI Overviews, AI Mode, and Discover AI responses. Google states that using it carries no organic ranking penalty.

Rollout is UK-first, with global expansion pending. If you do not see the toggle in your property yet, that is expected. Use the waiting period to settle your decision framework rather than improvising when it arrives.

One trap deserves top billing: opting out does not remove your content from the Gemini app. The toggle governs AI surfaces inside Google Search, not the standalone Gemini product. Anyone flipping it to “get out of Google’s AI” ends up only partially out, which undercuts the most common argument for blocking.

A Decision Framework for E-Commerce Sellers

For DTC brands and online stores, the default answer is: leave it alone. Visibility in AI answers is upstream demand. A shopper who sees your brand in an AI Overview may not click immediately, but branded searches, direct visits, and marketplace lookups typically follow that exposure.

Blocking removes you from a fast-growing surface and gives you nothing in return. Google promises no penalty for opting out, but it also promises no reward. And with the Gemini app exemption, the “content protection” argument is incomplete at best.

The rare cases where blocking makes sense:

  • Pure-content publishers whose traffic is the revenue through ads or subscriptions. When AI answers absorb the content, the clicks won back by opting out can plausibly outweigh lost exposure.
  • Companies using content access as licensing leverage. Withholding content from AI surfaces strengthens a negotiating position, and the toggle is the enforcement mechanism.

E-commerce operations match neither profile. Your guides, comparisons, and tutorials exist to acquire customers, not to monetize pageviews, so more AI exposure is strictly better. Spend two to three months building a baseline from the new reports, stitch it to your GA4 referral data, and quantify what AI surfaces are worth. By the time the toggle reaches your market, you will be deciding with numbers.

阅读本文中文版: Search Console 上线生成式 AI 报告：数据怎么读，屏蔽开关要不要开

Related Articles