Here’s a number worth sitting with: since Google’s AI Overviews rolled out broadly, click-through rates to third-party sites in certain query categories have dropped 20-40%. That’s not a marginal trend. That’s a structural change.

And yet Google keeps growing. The paradox is the key to understanding what’s actually happening.

A twenty-year model is shifting — quietly

For two decades, Google operated as a traffic router. User asks question, Google shows results, user clicks through. That model — which supported an entire ecosystem of SEO, content marketing, and ad-supported media — is transforming.

With AI Overviews (formerly Search Generative Experience), Google aggregates and summarizes information directly on the results page. For a growing share of queries — definitions, calculations, recipes, factual questions, simple comparisons — users get their answer without clicking anything. Google doesn’t lose revenue short-term: ad placements still appear. What compresses is the value created for the publisher and brand ecosystem that built on organic traffic.

What Gemini changes in this equation

The difference between earlier AI features and what Gemini brings to Search isn’t just degree — it’s kind. Gemini doesn’t only summarize; it reasons, compares, infers, and can guide decision-making paths. For complex queries (choosing a hotel, comparing two products, understanding an insurance contract), Gemini can replace a multi-step search sequence that would previously have generated five to ten clicks.

Two trends are emerging clearly in 2026:

Top-of-funnel informational queries are moving heavily to AI. “How does leasing work?” “What’s the difference between SPF 30 and SPF 50?” “Which Tokyo neighborhoods are best for a week’s stay?” — these questions are now answered directly, bypassing publisher sites entirely.

Transactional queries remain click-driven. “Order Nike Air Max 90 size 9,” “book a Paris-Tokyo flight,” “buy an IKEA EKTORP sofa” — purchase intent requires an action AI alone can’t complete. These queries continue driving traffic and value for commercial platforms.

What brands should do differently

Two postures make sense for brands managing online presence in 2026:

Reduce dependence on informational organic traffic. Concentrate SEO investment on transactional, high-purchase-intent queries. Accept that educational and informational content now serves brand perception more than direct acquisition.

Invest in AI citability. Publish proprietary data, original research, signed perspectives with distinctive value. If AI synthesizes search results, it synthesizes what exists and what’s cited. Brands producing genuinely original content have a structural advantage in this new equilibrium.

Build direct channels. Email, push notifications, owned apps — channels not mediated by a search algorithm. Brands that underinvested in direct audience relationships are most exposed to the traffic compression.

What Google won’t say out loud

There’s a subject Google’s AI Search communications carefully sidestep: compensation. When an AI Overview cites or synthesizes publisher-created content, those publishers receive nothing — no click, no mention, no payment.

Legal challenges are forming, especially in the US and Europe, around IP rights in the content used to train models and feed summaries. The outcome of these disputes could significantly reshape Search economics by 2027-2028.

Short-term, Google is in a dominant position: users are adopting AI Overviews massively, and no competitor offers a comparable alternative at scale. But the tension between value created for users and value destroyed for the content ecosystem is structural — and it will resolve, one way or another.

The search revolution is quiet because it doesn’t have the drama of a platform switch. Google isn’t disappearing. Users aren’t leaving for a competitor. But the rules of the game are changing — and brands that adapt earliest hold competitive advantages five years from now.