How AI Search Is Eating Organic Traffic: Benchmarks by Industry for 2026

How much organic traffic is AI search actually taking?

A lot, and the number is still climbing. Google searches in the United States ended without a click 68.01% of the time between January and April 2026, according to a SparkToro study built on Similarweb clickstream data, up from roughly 60% just two years before. That means fewer than one in three searches now sends a visitor anywhere at all.

The mechanism is AI Overviews and chat-style answers absorbing the question before a user ever scrolls to a blue link. Ahrefs tracked 300,000 keywords, half of them triggering an AI Overview, and found the average clickthrough rate for the number one ranking page down 58% when an AI Overview is present, up sharply from a 34.5% reduction the firm measured just eight months earlier. Pew Research Center reached a similar conclusion from actual browsing panels: users clicked a traditional result only 8% of the time when a Google AI summary appeared, versus 15% when it did not, and only 1% of visits involved clicking a link inside the summary itself.

None of this means organic traffic is dead. It means the traffic that remains is more concentrated, more competitive, and increasingly gated by whether a model chooses to cite you at all rather than whether you rank.

1. 68.01% of U.S. Google searches ended without a click, January to April 2026 (SparkToro / Similarweb)
2. 58% lower clickthrough rate for the top-ranking page when an AI Overview appears, December 2025 data (Ahrefs)
3. 8% click rate with an AI summary present versus 15% without one (Pew Research Center, March 2025 panel data)
4. 1% of visits with an AI summary present included a click on a link inside that summary (Pew Research Center)

Which industries are losing the most traffic to AI answers?

Information-heavy, advice-driven categories are losing the most ground. BrightEdge tracked AI Overview coverage across industries through December 2025 and found healthcare queries triggering an AI Overview 88% of the time, the highest of any category measured, followed by education at 83% and B2B technology at 82%. Restaurant and local queries sat close behind at 78%.

A separate study from NP Digital, built on Adthena search-index data through May 2026, measured trigger rates by query length across five verticals and found retail and finance climbing to 84% and roughly 79% respectively once queries stretched past nine words, the range where people ask fuller questions instead of typing bare keywords. Automotive and travel queries lagged well behind, peaking at 48% and 28%.

The table below lines up both studies. They use different methodologies, category coverage in one case, peak rate by query length in the other, so read it as a directional index of where AI answers are displacing clicks fastest, not a single unified benchmark.

Why are some industries barely touched by AI Overviews?

Commercial intent is the dividing line. Google still routes high-purchase-intent queries, the kind that precede an actual transaction, toward shopping ads, maps, and organic product listings rather than a generated summary. That is why travel and automotive, both categories where people compare specific bookable or buyable options, show the lowest AI Overview trigger rates in the NP Digital data, and why e-commerce and transactional queries generally sit in the low single digits.

Advice and research queries behave the opposite way. 'What causes X symptom' or 'best way to structure Y' has no single correct product to sell, so an AI summary can answer it fully without sending a click anywhere, which is exactly the pattern behind healthcare's 88% and education's 83% trigger rates.

The practical read: if your category lives on comparison and price, AI search is currently taking a smaller bite out of your organic traffic. If it lives on explanation and advice, plan for the bite to keep growing.

What do AI search traffic trends look like heading through 2026?

Two things are happening at once: AI Overviews keep expanding inside Google, and traffic is now also arriving from AI chat platforms directly, not just being absorbed by them. Similarweb puts average monthly visits across generative AI platforms worldwide at 9.5 billion between June 2025 and May 2026, a 70% year-over-year increase, with Google's Gemini rising from under 9% to roughly 27% to 28% share of that traffic and Claude moving from about 2% to close to 9%, the largest proportional gain of any platform tracked.

ChatGPT itself changed shape mid-year. After a May 7, 2026 product update that made cited brand names clickable and more prominent inside answers, total ChatGPT referral traffic rose 157.7% week over week, and the share of those referrals landing on a brand's homepage jumped from roughly 27% to about 62%, a level that has held rather than faded. That is a structural shift in how AI-referred visitors arrive, not a one-week spike.

Inside Google itself, AI Mode is still small in raw share, accounting for only about 0.34% of all searches from January through April 2026, but Google reported it passed one billion monthly users at I/O 2026 with query volume doubling quarterly. The traffic pattern most sites are adjusting to today (AI Overviews eating clicks from classic search) is likely the first phase of a longer shift, not the whole story.

What should marketing and content teams do about the traffic AI search is taking?

Stop measuring success by rankings alone. A page can rank first and still lose 58% of its historical clicks to an AI Overview sitting above it, so the more useful question becomes whether the model is citing you inside that overview at all. That is the idea behind Citation Share: the percentage of relevant AI answers in a category that cite a given brand, tracked across ChatGPT, Perplexity, Gemini and Google AI Overviews the same way rank tracking once covered page one of Google.

The categories with the highest AI Overview exposure (healthcare, education, B2B technology) are also the categories where being the cited source matters most, because there is no page two in an AI answer. A brand that never gets quoted inside the summary gets none of the 88% of healthcare queries or 82% of B2B technology queries that now resolve inside one.

This is not about gaming a model. It is about publishing enough authoritative, current material that a model has something worth citing in the first place, at a scale and freshness most in-house teams cannot sustain alone. LeadHaste, the operation behind this publication, took its own site from 0 to 1 million impressions and over 200 AI citations, with Domain Rating moving from 1 to 24, in four months of exactly this kind of publishing. The traffic environment described above is the reason that work was worth doing before the shift finished, not after.

1. Track Citation Share by engine (ChatGPT, Perplexity, Gemini, AI Overviews), not just organic rank
2. Prioritize content depth and freshness in the highest-exposure categories: healthcare, education, B2B technology, local and restaurant queries
3. Treat comparison and 'best X' pages as citation targets first, ranking targets second
4. Watch homepage-referral share from AI platforms as a leading indicator, it moved structurally in 2026
5. Re-audit quarterly. Trigger rates and referral patterns moved multiple times within 2026 alone

Sources

U.S. Google searches ended without a click 68.01% of the time, January to April 2026 Search Engine Land (SparkToro / Similarweb study), 2026-06-09

AI Overview presence correlates with a 58% lower average clickthrough rate for the top-ranking page, December 2025 data Ahrefs, 2026-02-04

Users clicked a traditional result 8% of the time with an AI summary present versus 15% without one Pew Research Center, 2025-07-22

AI Overview coverage by industry: Healthcare 88%, Education 83%, B2B technology 82%, Restaurants 78%, December 2025 Search Engine Journal (BrightEdge data), 2026-03-01

AI Overview trigger rates by query length across Retail, Finance, Automotive and Travel, May 2026 NP Digital (Adthena data), 2026-05-20

Average monthly visits across generative AI platforms reached 9.5 billion, June 2025 to May 2026, a 70% year-over-year increase; Gemini and Claude gained share as ChatGPT's share fell Similarweb, 2026-07-29

ChatGPT total referral traffic rose 157.7% week over week after a May 7, 2026 update, and homepage referral share rose from roughly 27% to about 62% of all ChatGPT referrals Similarweb, 2026-06-10