Do AI Overviews Actually Kill Your Organic Clicks? What the 2026 Data Shows

A client asks almost every week now: is Google's AI Overview quietly killing our organic traffic? The honest answer, backed by the best data available so far, is more specific than a yes or no. AI Overviews measurably reduce the odds that a given search ends in a click, but they don't appear on every search, and Google disputes how much they actually hurt sites in aggregate. Here's what the 2026 data actually shows.

The click-through rate drop, in real numbers

Pew Research Center analyzed 68,879 real Google searches from 900 U.S. adults, collected in March 2025 and published that July. When an AI Overview appeared, only 8% of visits ended in a click to a search result link. When no AI Overview appeared, that number was 15%. That's roughly a 47% relative drop in click-through rate on searches where Google decided to show a summary. Clicking a link inside the AI Overview itself happened even less often, in just 1% of all visits to pages carrying one.

What happens after the click doesn't happen

The same study found users ended their entire browsing session 26% of the time after a search page showed an AI Overview, compared to 16% of the time when it didn't. That gap matters more than the raw click number. It suggests AI Overviews aren't just changing which result gets clicked, they're often closing the loop on the search entirely, with the user getting a usable answer and moving on without visiting any site at all.

Not every search triggers one, and length matters

AI Overviews showed up on about 18% of all searches in the Pew dataset, and 58% of participants encountered at least one during the study period. Query phrasing predicts this closely. Searches of ten words or more triggered an AI Overview 53% of the time, against just 8% for one- or two-word searches. Question-phrased searches triggered one 60% of the time, versus 36% for full sentences that weren't phrased as questions. A typical AI Overview ran about 67 words and cited three or more sources in 88% of cases, with Wikipedia, YouTube, and Reddit together accounting for roughly 15% of the links cited.

Google pushes back on the numbers

Google called Pew's methodology "flawed" with a "skewed queryset," and the company maintains that AI Overviews drive over 10% more queries globally for the search types where they appear, a figure CEO Sundar Pichai has repeated publicly. Both things can be partly true at once: an individual search that shows an AI Overview may be less likely to produce a click, while the total volume of searches Google receives keeps growing. What the two sides don't agree on is whether that growth actually offsets the per-search click loss for the sites whose content gets summarized instead of visited. Independent data on that broader tradeoff is still mixed by industry and query type.

What this actually changes about a content strategy

Ranking first in classic organic results no longer guarantees a click, and it doesn't guarantee AI Overview citation either, since Google draws overview citations from a broader pool of sources than just the top result. What consistently shows up in the data is that AI Overviews favor specific, well-sourced answers to precisely phrased questions over broad or promotional pages, and they trigger more often on the longer, question-style searches that also tend to signal real buying or research intent. That means the practical response isn't to abandon organic content work, it's to apply the same discipline good answer engine optimization already calls for: write directly to the specific question a buyer types, back claims with named data the way a generative engine optimization program should, and track visibility inside AI answers as a second, additional metric next to standard rankings rather than a replacement for them. Tools built for this, like the ones covered on our AI search visibility page, exist because ranking position alone no longer tells the whole story of whether a brand is actually being seen.

This is precisely the tracking gap a one-time report can't close, and it's the work Seal Global's AI search visibility team does month over month for its own clients: running weekly citation checks across ChatGPT, Gemini, and Google's own AI Overviews so a client knows whether a content change actually moved their visibility, not just their rank.

Related Reading

• How ChatGPT, Gemini, and Google AI Overviews Actually Decide What to Cite
• Does Your Brand Actually Need AI Search Optimization, or Is This Hype?
• How to Check Whether Your Brand Shows Up in AI Search
• What Agentic AI and Human-in-the-Loop Workflows Mean for Brand Trust

Sources: Pew Research Center, "Do people click on links in Google AI summaries?"; Search Engine Land, "Google's AI Overviews are hurting clicks: Pew study".