Users Behave Differently in AI Overviews vs. AI Mode And Most SEO Strategies

A new clickstream study looking at 846,000 Google search sessions found something most marketers are still missing. Users do not behave consistently across Google’s AI search experiences. In AI Overviews, people compare. In AI Mode, people comply.

That difference sounds small, but it is not. And you can feel it in the way the sessions move, it is not just a “reporting” thing.

Many SEO teams are treating AI Overviews and AI Mode like the same channel. They measure visibility the same way, optimize content the same way, and they report success the same way. Even the dashboards end up wearing the same blindfold.

But the data hints that this is a mistake.

If older search was about earning the click, then AI search is splitting into two different games.

AI Overviews reward comparison and validation.

AI Mode rewards authority and selection.

Getting your head around that split could become one of the most important search marketing skills for the next two years.

Introduction

Users actively weigh options inside AI Overviews, and they often scroll backward, re-read the results, then click.

Users in AI Mode tend to accept recommendations with very little extra checking, making the whole thing feel winner takes most, even when context is missing.

Small businesses can still compete pretty well in AI Overviews even if they have a harder time showing up inside AI Mode recommendations.

The future of SEO isn’t just ranking once. It’s learning which AI interface you need to beat.

Why Are Users Acting Differently in AI Overviews vs. AI Mode?

Quick answer: AI Overviews drive more wandering, while AI Mode drives quicker acceptance.

The clickstream research showed users treated AI Overviews like a little research room, not a shortcut. Cursor movement got wider, scrolling became less linear, and people often went back to material they already had seen, sometimes twice.

AI Mode, on the other hand, did the reverse.

In a previous AI Mode study, users accepted AI generated shortlists in 88% of tasks, picked the first suggestion 74% of the time, and a lot of times they never clicked through to websites at all.

Think about how you watch Netflix.

Sometimes you already know what you want, and you hit play right away.

Other times you spend fifteen minutes scrolling, comparing, reading the descriptions, and second guessing yourself.

AI Mode is the first behavior.

AI Overviews is the second.

Thats why treating them as a single traffic source creates misleading conclusions.

AI search isnt one user experience anymore. It is two different decision making spaces.

AI Overviews vs. AI Mode What Changes for Marketers?

Short answer: Visibility matters in both environments, but the reason people engage is completely different.

Factor	AI Overviews	AI Mode	Marketing Priority
User behavior	Compare and validate	Accept recommendations	Different optimization approach
Click likelihood	Higher comparison driven clicks	Lower click-through rates	Optimize for different outcomes
Brand influence	Helpful but questioned	Strongly rewarded	Build authority signals
User attention	Distributed across SERP, while AI response stays focused, with different content formatting showing up	strongly rewarded	built for user module
Decision speed	Slow(er), but faster when it matters, and conversion paths can follow different routes	higher auth	upgarded
Winning factor	Trust and differentiation, inclusion in the AI shortlist, and then separate measurement models	yes	yes

Takeaway: Winning AI Overviews and winning AI Mode are getting separated into two distinct marketing disciplines, now.

One of the most interesting things from the research is that AI Overviews generated behavior that looked realy similar to online shopping.

People paused.

They compared.

They revisited.

They checked alternatives.

That is not passive consumption, it is active evaluation.

For marketers, this means metadata titles reviews and brand positioning suddenly matter right inside the search result, not only after the click.

What Does This Mean for Small Businesses?

Short answer: AI Overviews may actually open more doors for smaller brands than AI Mode.

A lot of conversations about AI search tend to drift toward the big brands, you know, taking over the results.

And yeah that worry is legit.

BrightEdge research shows brands show up far more often in AI Mode than they do in AI Overviews. AI Mode kind of repeats well known entities again and again, while AI Overviews bring more swingy behavior, with more varied sources.

For local businesses and smaller niche companies, this could be pretty positive.

1. Picture a small accounting firm going up against Deloitte.
2. In classic search, getting placed above Deloitte is already hard.
3. In AI Mode, getting suggested instead of Deloitte is also a rough climb.
4. But AI Overviews are a different lane.
5. People are actively comparing sources there.
6. They do not just accept the largest name right away.
7. They are checking what fits, for their situation.
8. They are confirming what was claimed.
9. They are hunting for cues that actually address the exact question.
10. That is the area where specialized know-how, becomes a real advantage.

Real Example: The Rise of Niche Authority

A 2025 SparkToro study found that niche publishers with strong topical authority, often generated higher engagement rates than large media sites even while getting less traffic overall. That was surprising, a bit. We’re seeing a similar pattern show up with ai search, and it is not just a random thing.

A regional cybersecurity consultancy that publishes very specific ransomware recovery content can become citation-worthy because its expertise stays focused and easy for AI systems to pull from. It reads clean and has clear signals, you know.

So the takeaway isn’t, “be bigger.”

It’s “be more useful”, practical, and precise.

AI Overviews also create openings for specialists because users are still comparing options, not simply accepting whatever shows up first.

Why Is Reverse Scrolling Suddenly Important?

Short answer: Because users are validating information before they make decisions.

One of the most overlooked findings in the study was reverse direction scrolling.

When AI Overviews showed up, nearly half of users ended up doing that thing where they scroll back up, like, checking again. With no AI Overviews, the upward scrolling was way lower.

And that pattern tells us something pretty useful.

It suggests people aren’t just passively reading.

They’re reviewing, assessing, checking.

That gives marketers a very different lane to work with.

How to Optimize for “Validation” Behavior

1. Create stronger title tags that clearly show expertise, not just generic promises.
2. Use specific proof points instead of broad claims that sound shiny but thin.
3. Put credentials up front, early on, not hidden later.
4. Include original data whenever it’s feasible.
5. Make your snippets easy to judge fast, at a quick glance.

Classic SEO was mostly about grabbing the first click. AI Overviews may mean you need to win the second look.

If someone returns to your result while evaluating options, you better have enough credibility signals to hold up, even against better-ranked neighbors.

• Imagine it like a product page.
• People rarely purchase on the first glance.
• They look at features, compare pricing, scan reviews, and weigh alternatives.
• AI Overviews are turning a lot more searches into that same kind of decision process.
• The second impression might even matter more than the first one.

Are Branded Searches Losing Their Edge?

Short answer: yes, surprisingly.

In the study, people who searched for a specific brand still spent more time weighing the AI Overview results than anyone would guess. And even when users already knew exactly which company they wanted, they still often stopped to check what was around it first.

For about twenty years, branded search was pretty direct.

User search for your company.

1. User clicks your website.
2. Journey finished.
3. That shortcut seems to be loosening, a bit.

AI Overviews add this extra evaluation layer between intent and action.

Before someone lands on your site, they run into summaries, alternative sources, comparisons, and citations.

So even loyal prospects may pause to look at what Google puts in front of them.

That creates some real risk, but also room for opportunity. Risk because rivals get more visibility when branded searches happen. Opportunity because weak players become easier to unseat with stronger content, plus better authority signals, a bit sooner than expected.

The new question is not:

“Did they search for us?”

It is: “What did they notice before they clicked us?”

Brand awareness still matters. Brand validation matters even more than it used to.

The Counter-Argument: Are We Overreacting to AI Search?

Short answer: Probably, but pretending these behavior changes are minor is riskier.

Some marketers say AI search is getting more attention than it earns.

They are not entirely wrong.

• Organic search still produces huge traffic.
• Several sectors haven’t felt dramatic AI disruption just yet.
• Even Google has suggested that AI Overviews appear selectively and get dialed back when people do not find them useful.
• Yet that argument misses the point.
• Even if total traffic volumes do not drop, how people decide is shifting.
• The biggest impact of AI search might not be fewer clicks.
• It may be different clicks, ah.
• People arrive later in the assessment process.
• They’re more informed, definitely.
• They already compared options.
• They already consumed summaries, in advance.

Several marketers report lower traffic but higher engagement from AI-generated visits, suggesting the folks who do click are often further down the decision funnel.

That doesn’t look like a traffic issue.

It looks like a measurement problem.

A lot of organizations are still judging success using metrics made for a search ecosystem that no longer exists.

The real disruption may be user psychology, not just traffic volume.

The Strategic Shift Many SEO Teams Need to Make

The biggest mistake marketers can make right now is treating AI Overviews and AI Mode as variations of the same search feature.

They’re not.

They’re different environments that trigger different behaviors.

AI Mode is becoming recommendation-driven and presented like guidance.

AI Overviews stay comparison-driven.

One pays back authority.

The other pays back differentiation.

One behaves like a digital assistant.

The other behaves like a research assistant, I mean.

So, future SEO strategies need two separate questions, unfortunately:

1. How do we become a source AI systems trust enough to recommend.
2. How do we become the option users choose after looking at alternatives.

The companies that answer both of those questions will get a real advantage.

Everyone else will keep optimizing for a version of search people are slowly moving away from.

That search results page is no longer just a hallway to websites.

It is turning into the place where decisions happen.