846,000 Google Searches Reveal How AI Overviews Are Changing User Behavior: the Practical Angle

For a long time, we’ve operated under a comfortable assumption in digital marketing: if someone types your brand name into a search bar, they are essentially already on their way to your website. Google was the hallway, and your homepage was the destination. You didn't worry about the hallway; you just wanted the user to get through it as quickly as possible.

But the hallway is changing. With the introduction of AI Overviews (AIO), Google is no longer just a pass-through. It is becoming a destination in its own right—a place where users pause, read, and reconsider before they ever decide to click. When the interface changes, human behavior follows. To understand exactly how this is happening, we have to look past the standard click-through rates and look at the actual movement of the user.

Recent analysis of 846,000 U.S.-based Google Search sessions from early 2026 provides a window into this shift. By examining anonymized clickstream data, we can see the nuances of how people actually interact with the Search Engine Results Page (SERP)—where their cursors linger, how far they scroll, and when they decide to turn back. The findings suggest that the "quick click" era is evolving into an "evaluation era."

The Engagement Equalizer: Why Intent No Longer Dictates Dwell Time

Historically, the way a person interacted with a search page depended entirely on why they were there. If you were looking for a local plumber, you behaved differently than if you were researching a complex medical condition or trying to find a specific login page. The data shows that without AI Overviews, searchers fall into five distinct behavioral buckets based on intent: informational, local, navigational, transactional, and video.

However, when an AI Overview is present, these distinctions largely vanish. The AI acts as an "engagement equalizer." Regardless of whether the user is looking for a product to buy or a specific brand, they all begin to behave like a single, unified audience. They stay on the page longer, and the time between seeing a listing and taking action expands significantly.

At the three-second mark, almost all users (between 91.9% and 93.1%) remain active on the page across all intent types when an AIO is present. This suggests that the AI summary captures attention immediately and holds it, regardless of the user's original goal.

Expert Interpretation:
This matters because it breaks the traditional "intent-based" SEO playbook. We used to optimize for "fast" intents (like navigational) and "slow" intents (like informational). Now, the AI is forcing a "slow" experience on everyone. The tradeoff here is a loss of immediate traffic in exchange for a more "considered" visitor. If a user spends 20 seconds evaluating an AIO before clicking your link, they are arriving at your site with more context and a higher level of intent than someone who clicked in two seconds. The decision you need to inspect is whether your landing pages are prepared to handle a user who has already consumed a summary of your value proposition on the SERP.

The Erosion of the Guaranteed Brand Visit

The most striking shift is happening with navigational searches—those queries where a user types a specific brand name or URL directly into the search bar. These have always been the "safest" bets for organic traffic because the user already knows exactly where they want to go.

Without an AI Overview, these users are the fastest to leave the SERP. Only 12% of navigational searchers are still active on the results page after 21 seconds. They see their brand, they click, and they are gone. But when an AI Overview appears, that number jumps to 46%. Nearly half of the people searching specifically for a brand are now lingering on Google instead of heading straight to the site.

This isn't just about time; it's about attention. Cursor movement data shows that without AIO, navigational searchers have a very tight "cursor spread" of about 8%, meaning they move their mouse in a straight line toward their known destination. With AIO, that spread increases to 27.5%. They are exploring the page, reading the AI's summary of the brand, and scanning the surrounding area before committing to the click.

Expert Interpretation:
This is a critical vulnerability for brand equity. When Google introduces friction into a branded search, it creates a window for "distraction." The tradeoff is that while the user is still searching for you, they are now exposed to the AI's interpretation of your brand and potentially other competing suggestions within the AIO. You can no longer assume a brand search is a guaranteed visit. The decision to inspect here is your brand's "digital footprint" in the data sets that feed these AI Overviews. If the AI is summarizing your brand, is it doing so accurately, or is it providing a version of your identity that might make a user hesitate?

From Scanning to Evaluating: The Shift in Cursor Behavior

There is a fundamental difference between "scanning" a page and "reading" a page. Scanning is a fast, erratic movement used to find a specific keyword or link. Reading is characterized by pauses and focused attention.

The data reveals that AI Overviews push users into a reading-and-evaluating mode. Users encountering an AIO keep their cursors still significantly more often—about 44% of the time, compared to 29% when no AI is present. While a still cursor might look like inactivity, the context tells a different story: these same users are covering more of the screen. They sweep across 83% of the viewport, compared to 66% in standard searches.

In short, users are pausing more, but they are pausing over a wider area of the page. They aren't just skimming for a link; they are consuming the content provided by the AI and comparing it against the organic listings below.

Expert Interpretation:
This suggests that the SERP is evolving from a directory into a research hub. The tradeoff is that the "first-click" advantage is diminishing. If users are spending more time evaluating the page as a whole, the quality of your meta-description and the clarity of your snippet become even more vital. You are no longer just competing for a click; you are competing for a place in a comparative analysis happening in real-time on the screen. The decision to inspect is your snippet's "complementary value." Does your listing provide a unique piece of information that the AI Overview missed, giving the user a reason to click through for the "full story"?

The Rise of the Reverse Scroll

We typically think of scrolling as a linear journey: the user starts at the top and moves down until they find what they need. However, the data shows that "back-scrolling" is a common behavior that intensifies with AI Overviews.

While the percentage of users who reverse direction at any point increases modestly (from 51% to 59% with AIO), the actual volume of upward scrolling is where the real story lies. Users are not just clicking and leaving; they are scrolling down to the organic results, then scrolling back up to the AI Overview to verify information, and then potentially scrolling back down again.

This means your listing is likely being seen multiple times in a single session. The user is treating the page as a living document, bouncing between the AI's synthesis and the source links to validate the claims being made.

Expert Interpretation:
The "reverse scroll" indicates a lack of total trust in the AI. Users are using the organic results as a fact-checking mechanism for the AIO. The tradeoff is that your site may get fewer total clicks, but the clicks you do get are coming from users who have already "vetted" the AI and decided that your specific source is the one they need to verify the truth. The decision to inspect is the "trust signal" of your listing. If a user scrolls back to you after reading an AI summary, your listing needs to look authoritative and definitive to capture that second, more intentional look.

The New Reality of the Search Experience

The overarching takeaway from these 846,000 sessions is that AI Overviews have transformed the SERP from a click surface into a decision-making environment. The presence of AI is now a stronger predictor of user behavior than the actual intent of the search. Whether someone is looking for a local business or a global brand, the AI slows them down and forces them to compare.

We are moving toward a world where the "zero-click search" isn't just about Google answering a question—it's about Google creating a space where the user feels they have enough information to avoid clicking entirely, or at least to be much more selective about where they go.

Data and Methodology

These findings were produced by ClickStream Solutions using anonymized data provided by Surfer SEO. The study analyzed approximately 846,000 Google Search sessions from users in the USA during February and March 2026. It is worth noting that because the data comes from a marketing-centric audience, the users may be more search-savvy than the average consumer. However, the vast majority of queries (98.9%) were non-marketing related, providing a broad look at general user behavior across various categories.

Practical next steps

The useful part is not only the idea itself, but the operating habit behind it. Use it as a checklist for decisions: what deserves attention now, what should be monitored, what needs a stronger evidence base, and what can wait until the system has more scale.