Google just told the market how many clicks its AI features send to websites. It will not tell you how many of those clicks are yours. On July 17, Nick Fox, Google's SVP of Knowledge and Information, wrote on LinkedIn that "we're now sending billions of clicks to websites every week through AI features in Search alone," on top of the long-standing claim that Search overall sends billions of clicks daily (Search Engine Land). It is the first time Google has attached a number to AI-feature-specific traffic. It is also a number no publisher on earth can audit.
The incumbent way
Most marketing teams will read the headline, not the fine print, and slot "AI search sends billions of clicks" straight into next quarter's channel narrative. That is the incumbent way: treat a platform's aggregate PR claim as a planning input, the same way teams once treated "mobile-first" or "voice search is the future" as settled fact rather than a framing exercise.
The fine print changes the read. Fox's post gives no baseline, no denominator, and no methodology. It does not say whether the weekly AI-feature figure is rising, flat, or falling relative to last quarter. It cannot be compared to the daily Search-wide figure, because the two numbers use different scopes and different time windows (Search Engine Journal). "Billions" sounds precise. It is closer to a marketing tagline than a metric.
The cost of standing still
Adopt the number this quarter without checking it, and the first cost is a false sense of coverage. A leadership team that hears "billions of clicks from AI search" assumes the open web is thriving under AI Overviews and AI Mode. The budget conversation shifts toward defending AI visibility spend rather than questioning whether it is working for this specific business.
Carry that assumption two quarters, and the second cost is a forecasting error nobody can trace. If your own AI-feature-driven traffic is flat or declining while Google's aggregate claim keeps growing, you have no way to know it, because the report built for exactly this question does not carry a click column. You will discover the gap in a quarterly business review, not in a dashboard, and by then it is a credibility problem, not a data problem.
Carry it a full year, and the third cost is a vendor-risk exposure. Marketing directors are already being asked to shift budget toward AI-visibility work, GEO audits, citation tracking, structured data, on the premise that AI search sends meaningful referral traffic. If that premise rests entirely on an unverifiable aggregate from the platform selling the outcome, the budget line has no independent audit trail behind it. That is the kind of gap procurement and finance eventually ask about, and "Google said so on LinkedIn" is not an answer that survives that conversation.
What the evidence shows
Google's own tooling confirms the gap. Search Console's "generative AI features performance" report, the sanctioned surface for tracking AI Overviews and AI Mode visibility, exposes five dimensions: impressions, pages, countries, devices, and dates. Clicks are not among them (Search Engine Journal). That structure has held since the report's original rollout on June 3, 2026, and remained unchanged through its most recent update (RanketAI). Google has said it will "gradually include more metrics," but has not committed to adding clicks or to a date.
Independent measurement points the other way. A randomized field experiment covering 1,065 U.S. participants, run in early 2026, found organic clicks to external sites fell roughly 38 percent on queries where an AI Overview appeared, with zero-click searches rising about 34.5 percent. Separately, SparkToro's modeling of Similarweb panel data found the share of Google searches ending with no click at all climbed from 60.45 percent in 2024 to 68.01 percent in the first four months of 2026, while the share of clicks reaching the open web fell from roughly 66 percent to roughly 27.6 percent over the same stretch (Digital Applied). Neither study proves Google's aggregate false. Both prove it is not the only credible read of the same underlying behavior, and both come with a stated sample, date range, and method that Google's own post does not.
The binary marketing directors now face
That is not a data quality nitpick. That is a governance decision about which numbers are allowed to move a budget.
Path one: keep citing the aggregate. Let "Google says billions of clicks" stand in for a measurement strategy, and find out during the next budget cycle that nobody can reconcile it against a single session in your own analytics.
Path two: build your own AI-visibility scorecard now, using what Search Console will actually give you, impressions, pages, and query-level detail, paired with server log analysis for AI-assistant referrer traffic and a standing skepticism toward any platform's aggregate click claim that arrives without a denominator. Treat Google's number as a market signal worth watching, not a figure worth forecasting against.
Verdict
Google has the number. Your Search Console account has impressions. Until that changes, the gap between what Google publishes and what any single business can verify is not a rounding error, it is the entire measurement problem. Marketing directors who build their own evidence now will still be standing on solid ground when this quarter's board deck asks where the AI traffic numbers came from. Everyone else will be citing a LinkedIn post.
Magnet builds AI-visibility measurement that starts with what you can actually verify, not what a platform claims in a press post. If your reporting on AI search traffic currently traces back to a headline instead of a log file, talk to Magnet about a measurement audit.
