Google Says Billions of AI Search Clicks. The Data Says a 58% Drop.

Google’s search leader Nick Fox posted on July 17, 2026 that AI features in Google Search are “now sending billions of clicks to websites every week.” Independent measurement tells a different story about ai search clicks: top-ranking pages have lost 58% of their click-through rate when an AI Overview appears, according to Ahrefs’ study of 300,000 keywords using aggregated Search Console data (Ahrefs, February 2026). Both claims can be true at once. What matters for you as a marketer is which one shows up in your own account.

This post breaks down what Google said, what the data shows, and the exact checks you should run this week to see whether your traffic is holding up or bleeding out under the new AI Search behavior.

What Google Actually Said About AI Search Clicks

On July 17 at 9:46 PM, Nick Fox, Senior Vice President of Knowledge and Information at Google, posted on X that Google is “now sending billions of clicks to websites every week through AI features in Search alone” (PPC Land, July 18). He said the team is “not done with this work” and that Google keeps improving how it displays links inside AI features to make it easier for people to click through.

The post landed as publishers, SEO teams, and lead-gen operators reported ranking-stable, click-declining pages inside Search Console. Google Search Advocate John Mueller has argued that clicks from AI Overview pages are “of higher quality, where users are more likely to spend more time on the site.” Liz Reid, Google’s Vice President of Product for Search, has said AI Overview clicks show greater purchase intent and longer downstream engagement.

Two things are missing from that framing. First, the ai search clicks headline has no denominator. Fox did not say what share of impressions those clicks represent, whether the base is growing or shrinking, or how the number compares to the pre-AI baseline. Second, quality claims about intent are not falsifiable without account-level data. You can measure this. Google has not published it.

What Independent Data Shows About AI Search Clicks

Ahrefs re-ran its 300,000-keyword CTR study using December 2025 Search Console data and compared it to December 2023, before AI Overviews launched. The result: the top-ranking page for a keyword that triggers an AI Overview now gets a 58% lower average click-through rate than the same slot without an AI Overview (Ahrefs, February 2026). That is up from the 34.5% drop Ahrefs measured in April 2025. In eight months the effect nearly doubled.

Position one on an AI Overview keyword dropped from a 7.3% CTR in December 2023 to a 1.6% CTR in December 2025. Position two lost 50.8%. Position three lost 46.4%. Even position ten lost 19.4%. A randomized field experiment with 1,065 desktop Chrome users, published on the Social Science Research Network in April 2026, found that when an AI Overview appeared, outbound organic clicks fell 39.8%, zero-click searches rose 34.5%, and clicks on sponsored results stayed flat.

Both numbers can hold at the same time. A rising absolute count of ai search clicks can coexist with a falling per-page rate if Google is running many more searches and covering more topics. The question you have to answer for your accounts is which direction your own numbers are moving, and whether the click quality claim shows up in your revenue reports.

Why This Data Matters for Ai Search Clicks in Your Account

If you run lead generation or e-commerce sites, the risk is not that traffic goes to zero. The risk is a slow bleed on informational and mid-funnel pages that used to fill the top of your funnel. Those pages still show impressions, still rank the same, and still look healthy in a dashboard that only tracks average position. The click-through column is where the loss lives.

This is what Ahrefs calls “the great decoupling.” Rankings stay flat, impressions rise, clicks fall. If your team only reports rank and impressions, you will miss it for months. Our team has written before about the citation and ranking shift the industry has been tracking in the top 6 AI search ranking signals for 2026 and how the shift toward on-SERP answers is now producing a 68% zero-click rate for informational queries.

The other risk is misreading Fox’s claim as good news. A page that used to get 100 organic clicks from position one and now gets 40 has lost 60% of its clicks. If those 40 remaining clicks convert at a slightly higher rate, revenue can still drop. Higher intent per click does not offset a large drop in click volume.

The Five Checks to Run This Week on Ai Search Clicks

You do not need a new tool stack to see what is happening. You need five checks against your existing Search Console and analytics data. Run these on your top 50 pages by impressions.

Check 1: Split queries into AI Overview shape and non-AI Overview shape. Filter Search Console for informational, question-based, and how-to queries. Compare CTR for that segment to your commercial and branded queries. If the informational segment CTR is more than 30% below your commercial segment CTR, AI Overviews are absorbing your top-funnel traffic.

Check 2: Compare CTR over 12 months at flat rank. Pull your top 20 URLs where average position stayed within one slot over the last year. If CTR fell by more than 25% at flat rank, that is decoupling. It is not a ranking problem, and refreshing the post to chase a rank you already have will not fix it.

Check 3: Segment by device. The Ahrefs study measured desktop CTR. Mobile behavior differs. Pull the same 20 URLs and split desktop against mobile. In our client accounts, mobile drops are usually larger because the AI Overview takes up more of the fold.

Check 4: Tag GA4 for AI-referrer patterns. Set up a custom dimension for referrers with no source, no medium, or clearly AI-related paths. Compare bounce rate, session length, and conversion rate against your normal organic segment. This is where you test whether Google’s “higher quality clicks” claim holds in your own numbers. If it does, you keep publishing. If it does not, you reallocate.

Check 5: Track AI Overview citation rate. A page can lose organic clicks and still gain brand visibility if it is being cited inside the AI Overview itself. Use Ahrefs Brand Radar, Semrush AI Toolkit, or a manual sample of 40 queries to check where your domain appears inside overviews. Cited pages recover roughly 38% of the click gap. Uncited pages sit at the full 58%.

What to Change Once You Have the Data

The measurement work above tells you which of your pages are healthy and which are bleeding out ai search clicks. The response depends on which bucket a page falls into. This is where a real GEO content audit beats a repackaged SEO crawl, because the fix for a decoupled page is not the same as the fix for a ranking problem.

For pages losing clicks but keeping rank, the play is citation capture. Rewrite the top of the page so it answers the primary question in the first 60 words with a clean, extractable structure. Add specific numbers, dates, and named sources. Google’s AI features cite pages that read like reference material more often than pages that read like sales copy. Our team has documented the four-pass workflow that separates the two in the diagnostic playbook on why AI Overviews quote Reddit instead of your page.

For pages that never had citation potential, such as sales pages, service pages, and localized landers, the answer is different. You defend those with intent-heavy queries, longer commercial keywords, and paid coverage where the organic decline is steepest. The Highlighted Answers Ads playbook covers how ad units are now sitting inside AI Mode answers and what that means for spend allocation. For lead-gen teams weighing whether to keep certain informational content live at all, our Google AI Search opt-out guide walks through the trade-offs.

Where Ai Search Clicks Reporting Goes From Here

Google is unlikely to publish the denominator Fox left out. The company has strong reasons to frame AI Search as additive rather than substitutive, and independent studies keep finding evidence for both effects at once. That means you have to build the measurement into your own reporting stack, not wait for a Google announcement.

Update your reporting on GA4 configuration so that consent, referrer, and query source are captured cleanly. Segment by AI Overview shape. Report CTR at flat rank against a prior baseline every month. Track citation share alongside ranking share. That combination will tell you within a quarter whether Fox’s “billions of clicks” or Ahrefs’ 58% CTR drop is closer to your reality.

The operators who win the next 18 months are the ones who stop debating whether AI Search is good or bad and start measuring the two effects separately in their own accounts. If you want a second set of eyes on your Search Console and GA4 to run the five checks above, our team does account audits on the same framework. Book a free consultation and we will show you exactly where your top-funnel pages sit on the decoupling curve.

Let’s Grow!