Less than a third of Google searches now end in a click: the store owner's response

Less than a third of Google searches now end in a click: 68.01% of them ended without one in the first four months of 2026, up from 60.45% in 2024 (SparkToro's analysis of Similarweb US clickstream data). For a store owner that headline is far less alarming than it reads, because the searches going silent are mostly the ones you never sold anything from — the money queries still click.

The reason we built Contexta on a store's own Search Console data is that a market-wide average like 68% says almost nothing about a single catalog. When we split a real store's queries by intent, the click loss is lopsided: informational "how does X work" searches shed clicks fast, while "buy", model-number and brand searches barely move. One honest caveat before the rest — the zero-click rise is real and still climbing, but pinning it all on AI is wrong. SparkToro found only 0.34% of searches reached Google's AI Mode in that same window, so most of the missing clicks are ordinary answer boxes, not chatbots.

Is the 68% zero-click figure the same for a store as for a publisher?

No — it's an average across every query type, and a store's revenue sits in the query types least affected by it. Publishers live on informational searches, which is exactly the traffic Google now resolves inside the results page, so a news site or recipe blog absorbs the full force of the 68%. A store lives on transactional and brand searches, where the person still has to reach a page to buy, pick a variant or check stock — so the same market average lands far softer on a catalog than on a content site.

The 68.01% figure counts every query identically, from "what time is it in Tokyo" to "buy trail runners size 10". Bundling those into one number is what makes the stat feel apocalyptic. It also over-credits AI: Google's AI Mode absorbed only 0.34% of searches in early 2026, and while AI Overviews are broader, the bulk of clickless searches are still weather boxes, sports scores, definitions and knowledge panels — the same SERP features that have existed for a decade, not a sudden chatbot takeover.

Which searches still send a store owner clicks in 2026?

Transactional and brand searches still click, because a purchase needs a destination. "Buy", "price", a model number, "[product] in stock", "near me", and searches for your brand name all resolve to an action Google can't complete inside the results page, so the click survives. Google can tell a shopper the average price of a running shoe; it can't add your specific variant to a cart, apply their loyalty discount or confirm your warehouse has it — that still lands on your page.

Map your own query mix and the split becomes obvious. Most searches sort into four groups, and only two of them ever reliably fed your store:

↑ Revenue value to the store

AI citations of your store

No click, real value: an assistant names you as the source, paying off later as branded searches and agent-driven purchases

Transactional & brand searches

The money queries: "buy", model numbers, your brand name — a purchase or stock check still needs your page

Informational "how / what / why"

Most of the vanishing 68%: answered in the results, and rarely a query a store sold from anyway

Comparison & research

Softer intent, clicks survive: shoppers still open tabs to compare before they commit

Chance it still ends in a click →

The transactional click hasn't disappeared; it's just surrounded by a growing crowd of informational queries that used to inflate your impression count. A purchase still needs a page to land on — though even that page is quietly turning into a data source AI agents parse rather than a storefront a human reads.

What still drives value when the click never happens?

Three things keep paying a store even when no click arrives: transactional clicks that still complete, brand demand that free answers build, and AI citations that name your store as a source. The first is direct revenue you can measure today; the other two are influence that surfaces later as branded searches, direct visits and agent-mediated purchases you can't trace back to a keyword.

Brand demand is the quiet winner of the zero-click era. Every time an answer box or an AI summary states a fact and attributes it to you, a shopper files your name away without clicking — and comes back later typing your brand directly, which is the search that clicks most reliably of all. AI citations work the same way but through a different door: being the source an assistant quotes puts you in front of buyers who never touch Google, and which engine names you depends heavily on where your facts live, since ChatGPT, Gemini and Perplexity each cite from different places. None of this shows up as a clean line in a clicks report, which is exactly why store owners undercount it.

How does a store owner tell a lost click that matters from noise?

Rank your lost clicks by the revenue behind them, not by their raw count — a page-one product query that slipped from 8% to 3% click-through costs you real orders, while a how-to post shedding impressions was never going to convert. Treating both as equal losses is how owners panic over a scary aggregate number while the actual leak, a handful of high-intent queries losing their click, goes unnoticed. Revenue is the first sort; the second is whether the click was ever available, which you can test by comparing each page against the best CTR your site earns at that position.

The mechanic is a sort in your own Search Console data: pull queries by page, flag the transactional and brand terms losing click-through, and ignore the informational impressions that were never revenue. Doing that by hand across hundreds of URLs is the job nobody finishes, which is what Contexta's Problem Map handles — it imports your GSC data and ranks pages by lost clicks per month, separating the CTR losers on money queries from the informational noise. The value the report can't see, AI-referred visits, has to be counted separately, because analytics systematically undercount the traffic assistants send.

So what's the store owner's actual response to zero-click?

Stop defending every click and defend the three channels that still pay: keep transactional pages winning the click, feed AI clean facts so it cites you, and measure branded and AI-referred demand apart from organic clicks. The panic response — chasing informational rankings that Google now answers for free — spends effort on the exact traffic that's structurally leaving.

The practical sequence for a WooCommerce store:

1. Separate your queries by intent and protect the transactional and brand ones — accurate titles, live stock, clean prices, fast pages. These are the clicks worth every hour.
2. Make your product and answer content extractable, not just rankable: lead each section with the answer, name every entity, and render it without JavaScript so assistants can quote it. This is the sentence-level shift from SEO to GEO, and it's what turns a page into a citable source.
3. Answer the questions that come after the one the page is titled for, on the same page — because a conversation that never returns to the web can only quote what it already pulled in, the reasoning behind the follow-up-question test.
4. Instrument the invisible channels. Count AI referral visits, watch branded-search volume, and treat a flat-clicks-but-steady-revenue reading as a signal of agent-mediated buying, not a failure.
5. Let go of informational vanity traffic. If a query is one Google answers in the box, winning position one buys you an impression and no visit — accept it and reallocate the effort.

The store owners doing well in mid-2026 aren't the ones who reversed the zero-click trend, because nobody can. They're the ones who stopped measuring their business by a metric that now counts mostly other people's queries.