Google still dominates search, but a 2025 dataset across 41 major sites shows a meaningful share of queries happening on commerce, social, and AI platforms—changing how demand gen leaders should allocate effort in 2026.

Google’s grip on search is usually described as absolute: “over 90% market share.” The number is repeated so often it’s treated like physics. But when search is defined as the thing people do—typing a query into whatever box is in front of them—the picture gets messier fast.

A year-long analysis published in March 2026 by Rand Fishkin, based on SparkToro and Datos (a Semrush company) research, looked at search behavior across 41 large websites using US and EU/UK desktop panels covering all of 2025. Under that broader definition, Google is still the biggest player. It’s just not the whole story.

That distinction matters in 2026 because demand gen teams are being asked to do two contradictory things at once: protect the inbound engine that still runs through Google, and show a credible plan for how the brand will show up in the places buyers increasingly treat as search engines—social feeds, marketplaces, maps, and AI tools.

Here’s the first uncomfortable truth: search isn’t a channel. It’s a behavior.

The “90% market share” claim breaks the moment search is defined like a buyer defines it

Conventional market share charts generally count activity across recognized search engines. That’s useful, but incomplete. Fishkin’s dataset widens the lens to include search behaviors occurring inside major digital properties, not only Google/Bing-style engines.

In that analysis, Google’s Q4 2025 US market share is reported at 73.7%—still dominant, still the default, but far from 90%. In the EU/UK desktop panel, Google is higher (roughly 78–80%), yet still below the mid-90s numbers that show up in common market share tools. The gap isn’t academic. A CMO building a plan around “Google is basically everything” will allocate budget, content formats, and measurement differently than a CMO who accepts that a meaningful share of search behavior is happening elsewhere.

But the context, however, is more complex. This research is desktop-only, and mobile represents a massive share of search behavior for many categories. Separate industry figures underscore that reality: reports cited in the research brief put phone-based search at 54%–81% of searches overall, with Gen Z at 80%–81% and Millennials at 62% (Sources: [1][2][3]). So treat the 41-site dataset as directional evidence, not a complete census.

Even so, it’s enough to force a reframing: the strategy can’t be “SEO plus maybe some AI.” It has to be intent-first distribution across the places buyers actually query.

Where the “other searches” go—and why that changes demand gen operations

Across the 41 sites in the study, the non-Google share doesn’t collapse into one new winner. It splinters. That fragmentation is the point: when buyers leave Google, they don’t all move to the same place.

And the stakes are high because click distribution inside Google is already brutally concentrated. In 2023, the #1 organic result captured 27.6% CTR, the top three results captured 54.4% of clicks, and only 0.63% of users clicked to page two (Source: [2]). That means a small ranking drop can create a pipeline problem that looks, from the board’s perspective, like marketing “suddenly stopped working.” It didn’t. Gravity just got stronger.

At the same time, buyers are not only searching for answers. They’re searching for reassurance. In 2023, 43.4% of internet users researched products and brands via search (Source: [2]). That’s not trivia. It’s a reminder that “search” often sits in the middle of a commercial decision, not at the beginning of a content journey.

Seen from the other side, this is also why “website diversity” keeps showing up in expert guidance: multiple intent-specific pages and formats tend to perform better because they match what the searcher is trying to do in that moment (Sources: [2][5][6]). One page rarely satisfies informational, comparison, navigational, and transactional intent at once. Trying anyway is how teams end up with content that ranks for nothing and convinces nobody.

For demand gen leaders, the operational implication is straightforward: the content plan can’t be a single blog cadence and a single set of landing pages. It needs assets that map to intent—guides for learning, comparison pages for evaluation, proof for validation, and crisp paths to conversion when the buyer is ready.

AI, social, maps, mobile: the new search surface area (and the measurement headache)

The 2025–2026 reports cited in the research brief describe search behavior fragmenting across AI tools, social platforms, and mobile, with Gen Z and Millennials leading shifts toward social discovery (Sources: [1][2][3]). The numbers attached to that shift are no longer small: 22% of consumers already use generative AI tools for product research and 72% plan to use them (Sources: [1][2][3][4]). Social platforms are also used as search by 31% of consumers, with Gen Z and Millennials in the cited 29%–49% range (Sources: [1][2][3]).

None of this means Google stopped mattering. In fact, the research brief is blunt on that point: in 2023, 68% of online experiences began with a search engine, and Google’s scale remains enormous (Sources: [2][3][7]). The better read is that “search everywhere” expands the playing field rather than replacing the core.

But the data also introduces a second uncomfortable truth: visibility and clicks are drifting apart. Multiple reports cited in the brief suggest AI answers can reduce clicks to traditional sites, with a wide reported range of up to 8.9%–50% (Sources: [3][5]). The exact impact varies by query type and SERP features, and cross-website measurement is limited in the provided results, so precision claims about channel splits should be treated carefully. Still, the directional message is hard to ignore: buyers can get what they need without leaving the platform.

That is why the practical demand gen question in 2026 isn’t “How do we win SEO?” It’s “Where do buyers seek information, comparison, and validation—and what format do they trust there?”

Three examples make the point:

And then there’s the part teams don’t like: attribution. When discovery happens in social search, when evaluation happens in AI tools, and when validation happens in maps and directories, the conversion path stops looking like a clean funnel. It looks like a buyer doing their own homework.

What to do with this: an intent-first “search everywhere” playbook for 2026

Start with a constraint, not an ambition: the top three organic results take 54.4% of clicks, and page two might as well not exist (Source: [2]). So the goal can’t be “rank for everything.” The goal is to win the handful of queries that reliably signal buying intent, while building enough presence elsewhere that the brand shows up during validation.

From there, the best approach is mapping buyer intent to the search environment that matches it:

To decide which formats belong where, experts recommend SERP analysis—look at what Google actually rewards for the query, including “People Also Ask” and the dominant result types—then build assets that match that revealed intent (Sources: [2][3]). Not guesses. Evidence.

The final loop closes back where it opened. Google is still the main highway, and in 2023 organic results captured 73% of clicks (up from 65% in 2022), which helps explain why 47% of marketers planned to increase SEO investment while only 10% planned to reduce it (Source: [2]). The money is following the gravity.

But gravity doesn’t mean monopoly. The 41-site research exists because buyers already treat many destinations as search engines in their own right. In 2026, the teams that keep calling this “an SEO problem” will keep optimizing one box while buyers ask their questions in dozens of others. The teams that accept the broader definition—search as behavior—will build presence where intent actually shows up, and stop being surprised when demand appears somewhere they weren’t measuring.