Search stopped sending the traffic. Commerce content still has to pay.

The range in the chart above is wide because the studies measure different things: different query sets, different position bands, desktop versus mobile. Treat any single figure as directional. What is not ambiguous is the slope: the April 2025 measurement of −34.5% had become −58% by February 2026.

Pew Research approached it from the reader side and found that when an AI Overview is present, 8% of users click a traditional search result, against 15% when no overview appears. Roughly half the onward clicks, gone.

It shows up in publisher numbers, not just SERP studies

Google search traffic to publishers fell 33% globally in the year to November 2025, with US publishers down 38%. Digital Content Next reported most of its members losing between 1% and 25% of traffic, and some losing more than 75%.

Dotdash Meredith named AI Overviews explicitly as a traffic factor in its Q1 earnings, while in the same quarter reporting affiliate commerce revenue up 26% and performance marketing up 11%.

The conclusion is about revenue per session, not traffic

Those two Dotdash facts sitting side by side are the whole strategic point. Sessions fell and commerce revenue rose, because commerce revenue is a function of intent and monetisation quality, not raw pageviews. A reader who arrives already deciding what to buy is worth many multiples of one who bounces off a news page.

If the top of the funnel is structurally shrinking, the only durable responses are to raise revenue per remaining session and to acquire traffic that does not depend on ranking. Which is a plain argument for better program economics on every click, and for distribution that isn't Google.