The Maze: Google search is not becoming one thing. It is splitting into two behaviors. Desktop looks more like research again: in 2026, 60.6% of US desktop searches still produced at least one click. Mobile became the answer machine. Only 17.7% of mobile searches produced a click, while zero-click jumped to 82.3%. That is the part that matters for ecommerce and content teams. Search visibility is still valuable, but mobile visibility increasingly means influence without a visit.
The all-device number hides the operating system underneath. SparkToro found that 68.01% of US Google searches in January-April 2026 ended without any click. Giacomo Iotti's device split explains why that number can rise even while desktop improves. Desktop zero-click fell from 43.2% in 2024 to 39.4% in 2026. But mobile zero-click rose from 69.1% to 82.3%. The weighted average moved toward mobile behavior. This is not a neat decline of search. It is a redistribution of where clicks survive.
Mobile search is now mostly a visibility channel, not a traffic channel. In 2024, the mobile click share was already weak at 30.9%. By 2026, it had fallen to 17.7%. That means four out of five mobile searches end inside the results experience. For marketers, the old funnel metric gets slippery. A product can be discovered, compared, and remembered without the browser ever visiting the site. The LinkedIn comment thread made the right point: visibility and traffic are now separate concepts. That is uncomfortable, because traffic is easy to measure and influence is not.
Desktop is where deeper intent still leaks out. The desktop pattern is the surprise. At least one click rose from 56.8% to 60.6%, while zero-click declined. That suggests desktop still carries more complex behavior: comparisons, tabs, deeper research, and tasks that need a destination. The strategic implication is not "stop doing SEO." It is split the job. Mobile content needs to be machine-readable, quotable, and entity-clear. Desktop content can still earn the click when the query is complex enough to need a page, a product detail, a calculator, a review, or a checkout path.
The methodology is good enough to act on, but not clean enough to worship. SparkToro's historical series combines different clickstream panels: Jumpshot for older studies, Datos for 2024, and Similarweb for 2026. SparkToro also notes that the 2026 study covers mobile browser searches, not the Google mobile search app. The source visual's mobile split is derived from a two-thirds mobile, one-third desktop assumption. So the exact mobile number should not become a tattoo. The direction is still commercially useful: mobile is where click supply is being removed fastest.
The replacement scoreboard is not raw traffic. It is named demand. SparkToro's companion analysis argues that SEO still matters, but its job changes. Search presence now does two jobs: it helps machines describe and cite you, and it captures branded or high-intent demand when users still click. That is why the practical KPI stack needs more than sessions: branded search, direct traffic, assisted conversions, share of voice, retail search visibility, review presence, and revenue movement. In ecommerce, this is especially important because AI-referred or zero-click-influenced shoppers may arrive later, fewer, and more qualified.
Why it matters: The mobile zero-click jump is a margin problem disguised as a media problem. If discovery happens inside Google, marketplaces, AI answers, and social feeds, retailers lose cheap attribution and gain a harder task: being the brand the machine remembers. The winners will not optimize every page for a visit. They will make product facts, category expertise, reviews, and proof easy for both humans and machines to reuse. Traffic may fall. The cost of being unknown will rise.
