AI Search Has Cut UK Website Traffic Growth by 86%, and Most Owners Cannot See It Happening

The uncomfortable argument

The SEO advice most UK small businesses are still paying for is advice for a search engine that no longer exists.

For twenty years the deal was simple. You published a page, Google ranked it, people clicked, and you got a visitor. Rank higher, get more clicks. That was the whole game, and a generation of agencies, courses and freelancers were built on it.

That deal has quietly broken. Google now answers the question itself, at the top of the page, using your content, without sending anyone to your site. People read the answer and move on. The link you worked to rank is still there. Nobody clicks it.

Here is the part nobody is saying out loud to small businesses. This is not a dip you wait out. It is a structural change in how being found online works, and the numbers are already brutal. Worse, the tools most owners use to measure their marketing cannot see most of what is now happening. You are being told traffic is “a bit down” when the foundation has moved.

This report sets out what the data actually shows, why your analytics are lying to you by omission, and what a small business should do about it. None of it requires a bigger budget. It requires fixing structure before chasing scale.

The UK numbers: an 86% collapse

The clearest UK picture comes from a study by the agency Tank, published in October 2025. It tracked organic traffic for 800 companies across 16 sectors, using Ahrefs data spanning three annual periods from August 2022 to August 2025, before and after Google launched AI Overviews in August 2024 and AI Mode in July 2025.

Average monthly organic traffic growth across all sectors fell from 26.3% the year before AI search to 3.7% the year after. That is a decline of 22.6 percentage points, or an 86% collapse in the rate of growth. Sites were not just growing more slowly. The engine that quietly added visitors every year had largely stopped.

Why nobody is clicking any more

The mechanism behind the UK collapse is visible in the click data. The Pew Research Center analysed the real browsing behaviour of 900 US adults in March 2025. Close to 6 in 10 of their searches returned an AI summary at the top of the page.

The effect on clicks was stark. When an AI summary appeared, users clicked through to a traditional website result in only 8% of searches. Without a summary, the figure was 15%. That is nearly half the clicks, gone, simply because Google answered the question itself. Clicks on the links inside the AI summary were rarer still, at 1% of visits. People were also more likely to give up and end their session entirely after seeing an AI answer.

Step back and the trend is industry-wide. Similarweb data shows zero-click searches, where the user gets what they need without leaving Google, rose from 56% to 69% after AI Overviews launched. In Google’s newer AI Mode, analysis from Semrush puts the share of sessions ending with no click at around 93%. The website is becoming the thing the AI reads, not the thing the customer visits.

Click and abandonment figures from Pew Research Center (US browsing data, 2025). Zero-click figures from Similarweb and Semrush. AI Mode figure is an industry estimate and will move as the feature rolls out.

Who is getting hit hardest

The damage is not spread evenly. The pattern is clear and it is bad news for exactly the businesses Whito works with: anyone whose customers start with a question.

In the Tank study, hospitality was the worst affected, with organic traffic actually shrinking by 6.7% in the year after AI search. Content-heavy and advice-led sectors, the how-to guides, the explainers, the “what does this cost” pages, saw some of the steepest drops, because those are precisely the questions an AI answer handles in full. IT and technology firms held up best, with growth slowing only slightly, because their buyers still click through to compare technical detail.

Separate analysis from Chartbeat, reported by Axios in March 2026, found that small publishers were hit hardest of all. Sites in the 1,000 to 10,000 daily page view range, which describes a great many small business websites, saw search referral traffic fall by around 60% over two years. Big brands with direct traffic and strong names absorbed the hit. Small sites that depended on Google to be found took the full force of it.

The blind spot: your analytics are lying to you

Here is the part that should worry owners most, and the reason this change has gone under-reported. The tools small businesses use to judge their marketing cannot see most of what is now happening.

When ChatGPT recommends your business, when Gemini summarises your service, or when Perplexity cites your page, there is often no clickable link and no referral. Someone reads the answer, then types your name into Google or your address straight into the browser. In your analytics that shows up, if it shows up at all, as “direct” or “branded” traffic with no obvious cause. Industry estimates suggest standard GA4 setups capture only 30 to 40% of genuine AI-driven visits. Google only added a dedicated way to spot AI assistant traffic in May 2026, and most small business accounts are not configured to use it.

So the typical small business is flying blind in both directions. The old traffic is falling and they can see that. The new AI-era discovery is happening and they mostly cannot. An owner checks their dashboard, sees traffic “a little soft”, and concludes they need a few more blog posts. The real story, that the channel has fundamentally changed, is invisible on the screen they are looking at. The principle still holds: if you cannot measure it, do not fund it, but first you have to measure the right thing.

The reframe: stop trying to rank, start being the answer

This is where most reactions go wrong. The instinct is to fight harder in the old game: publish more, chase more keywords, buy more SEO. That is spending money to win a race that has moved to a different track.

The principle is the one Whito comes back to on everything. Structure before scale. Before you spend on tactics, fix what the business looks like to the system that now does the answering.

In an AI answer, there is usually one recommendation, not a page of ten blue links. The question is no longer “where do I rank”. It is “am I the business the AI names”. That is won by being clear, consistent and genuinely useful across the web: a sharp description of what you do and who for, real reviews, accurate listings, content that answers questions properly rather than padding for word count. The businesses AI recommends are the ones whose positioning is unambiguous. That is a structure problem, and structure is cheaper to fix than traffic is to buy. Our earlier work on how Google AI Overviews are changing UK search covers the mechanics, and the Start stage is where this work begins.

There is one piece of good news in the data worth holding onto. Visitors who do arrive after an AI answer convert at several times the rate of ordinary search visitors, because the AI has already done the explaining and the trust-building. Fewer visitors, but warmer ones. For a small business, a handful of ready-to-buy enquiries beats a thousand browsers who bounce. The job is to make sure you are the name the AI hands them.

What a small business should do now

Common questions