Google searches that end without anyone clicking anything hit 68% in the first four months of 2026. In 2024 the figure was 60.45%. That's a 9.5-point swing in the share of searches producing a click — a 22.9% decline — inside two years.
Put differently: fewer than one in three Google searches now sends a visitor anywhere.
Where the clicks went
AI answers absorbed them. AI Overviews now appear on more than 20% of all queries, and AI Mode has passed a billion monthly users with query volume more than doubling each quarter.
The referral maths is where it gets uncomfortable. Traditional Google search refers traffic on roughly 17–19% of queries. AI Mode refers on 1.6–2.5%. That's not a modest decline in efficiency; it's close to an order of magnitude, and it applies to a surface growing faster than anything else in search.
The downstream damage is already measurable. Around 73% of B2B websites recorded significant traffic loss between 2024 and 2025, with year-on-year declines averaging 34%.
The number that complicates the panic
One finding runs against the grain: AI-referred visitors convert at about 4.4 times the rate of traditional organic visitors.
It makes sense on reflection. Someone arriving from an AI answer has usually had their question already handled. They're not clicking to learn what a thing is — they're clicking because they've decided to act. The volume is a fraction of what it was, but the intent is far more concentrated.
Whether that trade works depends entirely on your model. If your revenue depends on ad impressions or affiliate volume, a 34% traffic decline is not offset by better conversion. If you sell something with a meaningful order value, fewer and better-qualified visits may leave you roughly where you started.
What marketers are actually doing about it
The strategic advice converging across the analyses is that visibility inside the answer is now worth pursuing on its own terms.
Being cited in an AI-generated answer, a featured snippet or a knowledge panel delivers value even when no click follows. The user still sees your brand, your claim, your framing. That's closer to how advertising and PR have always been measured than to how SEO has been measured, and it requires a different scorecard than sessions and bounce rate.
The complementary recommendation is blunter: invest in brand presence on the platforms where your audience already is, whether or not those efforts produce site visits. In an environment where most queries resolve without leaving the results page, being known matters more than being findable.
The measurement problem nobody has solved
The awkward part is that citation-based value is hard to count. You can measure clicks precisely. Measuring how often you were mentioned inside an AI answer, to whom, and with what effect on a purchase weeks later is considerably messier, and the tooling is immature.
That gap is going to cause a lot of internal argument. Traffic dashboards will keep showing decline, because traffic is declining. Whether the business is actually worse off is a separate question that current analytics can't answer cleanly.
The useful move for now is to stop treating organic sessions as the headline number. Track them, but track branded search volume, direct traffic, and conversion rate by source alongside them. If AI-referred visitors really do convert at four times the rate, that shows up in the revenue line long before it shows up in the traffic chart.
Image: TREEDEO.ST, via Pexels
