Google's AI click figures are big, unauditable, and self-reported

Google's Nick Fox told the world recently that AI features in Search send billions of clicks to websites every week. Search Engine Journal reported the claim. Nobody can confirm it.

That is the problem in its entirety. Fox made the statement, Google attached a large-sounding figure to it, and third-party verification does not exist. The company controls the data, publishes what it chooses, and invites the market to take the rest on faith.

The asymmetry that matters

Google's position is structurally comfortable. It can cite aggregate traffic numbers when critics argue that AI Overviews cannibalise organic clicks, then withhold the methodology that would let publishers, advertisers, or regulators test the claim. A number without a denominator is not evidence; it is a press statement. The distinction is not pedantic. For a CMO at a major industrial group or a multilateral institution deciding how much budget to allocate to SEO versus paid channels, the absence of auditable data means every strategic inference rests on Google's word alone.

The parallel to Google's daily search count is instructive. That figure has circulated for years, cited in boardrooms and analyst decks, and it too has never been independently verified. Adding a second unauditable metric alongside the first does not validate either number. It compounds the opacity.

What brands are actually flying blind on

The specific claim matters for AI visibility in a way that goes beyond generic scepticism about vendor statistics. If Google's AI Overviews and AI Mode are genuinely routing billions of clicks outward each week, then being cited within those AI answers is a meaningful traffic channel, and brands should invest accordingly in the content signals, structured data, and authoritative sourcing that earn those citations. If the figure is inflated, or if it aggregates clicks that would have occurred through conventional results anyway, then the strategic case for that investment changes materially.

Financial services firms and policy institutions face this uncertainty most acutely. Their audiences make high-stakes decisions and query AI search for synthesis and recommendation, not casual browsing. The click-through rate from an AI answer to a named authoritative source is likely higher in those contexts than in consumer categories. But "likely" is doing heavy lifting in that sentence, because the segmented data does not exist in the public domain.

Google's incentive structure runs directly against disclosure. Releasing query-level or category-level data on AI-driven clicks would expose the degree to which AI Mode redistributes traffic across domains, and specifically whether established publishers and institutional sources gain or lose relative to AI-generated summaries. That is not information Google has any commercial reason to publish voluntarily.

The signal for anyone building an AI visibility strategy

The practical consequence is this: brands cannot wait for Google to supply the evidence before acting. The absence of auditable click data from Google does not mean AI search has no effect on brand visibility; it means the effect is unquantifiable from the outside. The rational response is to monitor what can be measured, which is share of citation in AI answers across the major models, not to assume Google's aggregate figure tells you anything about your own position.

Perplexity, ChatGPT, and Microsoft Copilot all surface attributable sources in ways that allow at least partial citation tracking. Google remains the largest search engine by volume, and it is precisely the one offering the least transparency. For senior communications and content leaders at organisations that depend on being cited authoritatively in AI answers, that asymmetry is now a structural feature of the information environment, not a temporary gap waiting to be filled.

Google will keep publishing large numbers. The numbers will keep being unverifiable. The brands that treat unverifiable claims as a reason to diversify their AI visibility measurement beyond Google's own reporting will be better positioned than those waiting for data that will not arrive.