Cate Dombrowski at iPullRank published a study on July 30, 2026, analyzing roughly 13 billion Google searches, and the topline number is the one every recap of it will lead with: 47% of Google searches now end without a click. That's a real, large, well-sourced number, and it's also not the finding worth building a strategy around. The finding worth building a strategy around is buried a few paragraphs into the same study.
The zero-click search number everyone will quote
Zero-click search has been the year's most recycled statistic, and for good reason. It's been reported at 68% by SparkToro's Rand Fishkin using Similarweb clickstream data, and separately by Kevin Indig's Growth Memo in its own H1 2026 halftime numbers. Dombrowski's 47% figure, published a day after Indig's, comes from a genuinely different, larger dataset, 13 billion searches rather than a panel-based clickstream sample, and lands at a meaningfully lower number. That gap between studies is itself worth noting; different methodologies are converging on 'clicks are declining sharply' as a direction while disagreeing by double digits on the exact rate, the same pattern we've flagged before when three trackers disagreed on ChatGPT's market share. Track your own numbers, not any single vendor's topline stat.
The geographic breakdown in Dombrowski's study adds useful texture past the single headline figure too. The zero-click rate ranges from 41% in Japan and Vietnam, the most click-friendly markets in the dataset, up to 54% in the Philippines, the highest zero-click rate measured. The US sits at 49%, close to but slightly above the global 47% average. A team running a single global zero-click assumption across every market it operates in is already working from the wrong number for most of its markets.
The number that actually matters: 86%
Here's the finding that gets far less attention than it deserves. Of the clicks that do happen on Google, 86% go to sites the user did not name in their search query. Only about 14% are navigational: someone typing a brand name, or a specific site they already had in mind, and clicking through to exactly that destination. The overwhelming majority of surviving click volume is Google introducing a searcher to somewhere they didn't ask for by name.
Share of Google clicks, by whether the destination was named in the query (iPullRank, Jul 30, 2026)
Sit with what that means before moving on. If a search engine's surviving click volume were mostly navigational, the strategic implication would be straightforward: rank for your own brand terms, defend the queries where people are already looking for you by name, and treat everything else as a smaller, secondary opportunity. That is not the world this data describes. The actual world is one where the vast majority of the value left in classic organic search is Google acting as a matchmaker between a searcher's problem and a destination they'd never heard of, which is a completely different game to build a content strategy around than defending your own name.
Zero-click search as discovery, not lookup
This reframes the zero-click conversation in a way that's more useful than another round of doom about the shrinking pie. Yes, the pie is shrinking; 47% to 54% of searches end with no click at all, depending on market. But the slice that remains is disproportionately a discovery slice, not a defense-of-existing-relationships slice. That's structurally similar to the argument we've made about AI engines directly: only 15.2% of AI-search categories currently have a stable owner, meaning 89.3% of category-level demand is still unclaimed and winnable. The zero-click data says something adjacent but distinct about classic Google search specifically: even in a shrinking click pool, most of what's left is genuinely up for grabs by whoever answers the underlying problem best, not reserved for whoever already has the brand recognition.
| CONTENT STRATEGY BUILT FOR… | ASSUMES | WHAT THE 86% FIGURE SAYS ABOUT IT |
|---|---|---|
| Defending branded search share | Most surviving clicks are navigational | Wrong — only ~14% are |
| Winning unnamed, problem-first queries | Most surviving clicks are discovery-driven | Right — ~86% are |
| A single global zero-click assumption | One zero-click rate applies everywhere | Wrong — ranges from 41% to 54% by country |
| Segmenting by market and intent before publishing | Discovery share and zero-click rate both vary by segment | Consistent with the data |
There's one more figure in Dombrowski's study worth flagging on its own, because it complicates the simple 'AI is stealing Google's clicks' narrative. ChatGPT ranks as only the sixth most-clicked destination from Google search itself, meaning people are searching Google for ChatGPT and clicking through to it at a lower volume than most would assume given how much of the current discourse frames ChatGPT as Google's replacement. The study also found ChatGPT pulls a notably higher share of paid-ad clicks from Google, roughly 4.75%, than other major destinations; YouTube, by contrast, gets negligible paid-click traffic at around 0.2%. Google is still very much the front door to ChatGPT for a meaningful slice of its user base, which is its own small irony worth sitting with.
What this means for content strategy
The practical implication is that content built to be found by people who already know your brand, glossary pages, branded comparison pages, your own product's feature pages framed for people already evaluating you specifically, is competing for a shrinking 14% slice of an already-shrinking pool of clicks. Content built to answer a problem before the searcher has attached any brand name to it at all is competing for the 86% slice, and that's true whether the searcher ultimately clicks through to a result on classic Google or gets the answer folded into an AI Overview or an AI Mode summary instead.
This lines up with something we've already measured directly in AI answers specifically: comparison and alternatives content earns 32.5% of AI citations, more than any other format, because it answers a buyer's first real question, what are my options, before the buyer has decided on a specific vendor to search for by name. The 86% discovery-click figure is the classic-search version of the exact same behavior. Buyers overwhelmingly start from a problem, not a brand, and the content that wins is the content built to meet them at the problem, not the content built to be found once they already know what to search for.
How to build for discovery instead of just ranking
Building for discovery rather than pure ranking changes a few concrete decisions, not just the framing.
None of this makes zero-click go away, and it isn't a reason to stop tracking it. It's a reason to stop treating the shrinking click pool as uniformly bad news. A pool that's 86% discovery-driven is a pool where being the best answer to an unnamed problem, not the biggest brand, is still the thing that wins the click, whichever surface actually delivers it. That's a strategy question content teams can act on directly, which is more useful than another quarter of watching a topline zero-click percentage creep upward with nothing to do about it. It's the same reorientation behind our own content marketing engagements: build for the problem a buyer arrives with, not the brand name they haven't decided on yet, and both the classic-search and AI-answer surfaces tend to reward the same content.
One caution before you rebuild a whole content calendar around this single number. Dombrowski's 86% figure describes the aggregate across a huge, mixed dataset, and the actual navigational-versus-discovery split will vary by industry and query type the same way the zero-click rate varies by country. A company with unusually strong brand recognition in its category may see a meaningfully higher navigational share than the aggregate suggests, and that's useful information too, not a sign the study is wrong for that business. Pull your own segmented number before treating the global 86% as your specific target.
See where you are cited today
A free snapshot audit of your rankings and AI citations before we ever talk.
Tyler leads work at the intersection of SEO and generative engines at Something Inc., helping B2B brands get ranked and cited across every major AI engine.