In May 2025, something fundamental changed in how Google search works. It wasn’t announced with fanfare, and many store owners didn’t immediately notice. But by mid-summer, the impact was undeniable: organic traffic patterns had shifted dramatically.
Some stores saw 30-50% traffic drops. Others—counterintuitively—saw massive increases. One store we track saw a 700% spike in organic traffic during the exact timeframe these changes rolled out.
What changed? Google launched AI Mode, and it fundamentally altered how the search engine evaluates which sites deserve organic traffic.
Understanding what happened—and why some sites won while others lost—isn’t just interesting history. It’s critical intelligence for your 2026 strategy. Because AI Mode isn’t going away. It’s Google’s future direction for search.
What Is AI Mode? (And How Is It Different From Regular Search?)
Before May 2025, Google search worked in a familiar way:
You typed a query → Google showed a list of blue links → You clicked one → You got your answer.
Sometimes you’d see AI Overviews (formerly called Search Generative Experience) at the top—short AI-written summaries with cited sources. But traditional search results still appeared below.
AI Mode changed everything.
Introduced in May 2025, AI Mode doesn’t show traditional search results at all. Instead, Google’s Gemini AI model runs multiple background searches, synthesizes information from across the web, and generates a comprehensive answer in a conversational format.
No list of links. No traditional “results page.” Just an AI-generated response—with some cited sources embedded.
For certain types of queries, AI Mode became the default experience. And Google clearly signaled this is where search is heading.
Why Google Launched AI Mode
Google didn’t build AI Mode to hurt websites. They built it to solve a real problem: people were getting frustrated with traditional search.
Consider a query like: “What should I know before buying a road bike?”
Traditional search gives you:
- 10 blue links to different articles
- Each article answers part of the question
- You open multiple tabs
- You piece together information yourself
- It takes 15-20 minutes
AI Mode gives you:
- One comprehensive answer synthesizing multiple sources
- Organized by relevant topics
- Covers the most important considerations
- Takes 2-3 minutes to read
For informational queries—where someone just wants to learn something—AI Mode is objectively better. Faster. More efficient. Less cognitive load.
Google knows this. That’s why they’re pushing AI Mode aggressively.
The Zero-Click Problem (And Why Many Sites Lost Traffic)
Here’s where it gets painful for website owners.
When AI Mode answers someone’s question comprehensively, they don’t need to click through to any website. The query ends right there. Zero clicks.
The data is stark:
For news-related queries: Zero-click results increased from 56% to 69% between early 2025 and mid-2025. That means over two-thirds of searches end without anyone clicking a traditional link.
For informational queries: The percentage is even higher. People get their answer and move on.
The result: Many content-driven sites saw traffic drop 30-50% starting in May 2025. Publishers, blogs, how-to sites—anyone whose primary value was “answering questions”—got hit hard.
If your site’s main purpose is providing information that AI Mode can summarize, you’re competing with a summary that appears before users ever see your link. That’s a losing battle.
But Some Sites Saw Traffic Increase. Why?
Here’s what confused everyone: while most sites lost traffic, some saw dramatic gains during the exact same timeframe.
We tracked one e-commerce store that saw 700% organic traffic growth starting in mid-May 2025—precisely when AI Mode launched.
This wasn’t a fluke. Other stores with similar characteristics also reported growth, not decline.
What made the difference?
AI Mode needed to learn something new.
When AI Mode answers a question, it’s synthesizing information it found on various websites. But how does Google’s AI know which websites to trust? Which ones actually deliver value?
Google’s AI can’t just rely on traditional SEO signals (keywords, backlinks, domain authority) anymore. Those tell you if a site should theoretically be good—not if users actually experience it as valuable.
So Google started weighting engagement signals far more heavily. These are behavioral indicators that users find genuine value:
Dwell time: How long users stay on your site before returning to search Pages per session: How many pages users explore Return visits: Do people come back to your site? Interaction patterns: Scrolling, clicking, adding to cart, engaging with content Bounce-back rate: Do users immediately hit “back” and try a different result?
These signals tell Google’s AI: “Users actually liked this site. They explored. They engaged. They came back.”
Sites optimized for engagement—not just information delivery—thrived under AI Mode.
The E-Commerce Advantage No One Expected
Here’s the counterintuitive insight: e-commerce sites with high engagement had an unexpected advantage when AI Mode launched.
Why? Because shopping isn’t just about answering questions. It’s about discovery, exploration, and experience.
Consider these two scenarios:
Store A (Low Engagement): User searches for fabric and lands on the store. They view a couple of products, spend about 90 seconds on the site, then leave. They don’t return.
Store B (High Engagement):
User searches for fabric and lands on the store. They view 8-10 different fabrics, spend 5+ minutes exploring, and bookmark the site. They return the next day and again later that week.
The difference isn’t what the stores say they offer—it’s what users actually do when they get there. Store B creates an experience where shoppers naturally view more products, stay longer, and come back. That behavioral difference is exactly what AI Mode learned to recognize as genuine value.
All the content that used to count is now mostly ignored.
Real Data: What “Winning” Looked Like
Let’s look at the concrete numbers from that store that saw 700% traffic growth.
This store installed Stylaquin in February 2024. Over the following 15 months, shoppers who used Stylaquin’s visual browsing features engaged very differently than those who shopped traditionally. By May 2025, when AI Mode launched, the engagement difference was stark.
Comparing shoppers who used Stylaquin versus those who didn’t:
Session duration: 4:06 → 5:24 (70% longer) Products viewed per session: 4.9 → 10.0 (104% more) Events per session: 5.3 → 11.2 (111% more) Returning visitor rate: 14.5% → 26.2% (80% higher) Returning visitor conversion: 3.76% → 8.13% (116% higher)
These weren’t site-wide improvements over time. These were measurable differences in behavior: shoppers who engaged with Stylaquin’s Idea Boards viewed twice as many products, stayed 70% longer, and came back 80% more often than shoppers who used traditional product grids.
Their site wasn’t just getting traffic—it was keeping users genuinely engaged.
When AI Mode launched and started evaluating sites based on engagement signals, Google’s AI saw exactly what it was looking for: a site where users explore, discover, interact, and return.
The result? 700% organic traffic growth during the period when most sites were declining.
Why This Matters for 2026
AI Mode isn’t a temporary experiment. It’s Google’s long-term direction.
Google announced in late 2025 that they’re expanding AI Mode to more query types. The zero-click trend will accelerate. More searches will end without anyone clicking a traditional result.
If your 2026 strategy assumes traditional search results will still dominate, you’re planning for a past that’s disappearing.
Here’s what merchants need to understand:
1. Information Alone Won’t Drive Traffic Anymore
If your site’s primary value is answering questions—”What’s the best [product]?” or “How do I [solve problem]?”—AI Mode will answer those questions before users reach you.
You need to offer something AI can’t replicate: experiential value.
Visual discovery. Interactive exploration. Curated collections. Social proof. Personal recommendations. The tactile experience of browsing.
These create engagement that signals value to Google’s AI.
2. Engagement Metrics Are Now Primary Ranking Factors
Keywords still matter. Backlinks still matter. Technical SEO still matters.
But if users bounce immediately, view one page, and never return, the AI learns your site doesn’t deliver genuine value—regardless of your keywords and backlinks.
Track these metrics as closely as you track traffic:
- Average session duration
- Pages per session
- Bounce rate (or engagement rate in GA4)
- Returning visitor percentage
- Interaction depth (clicks, scrolls, product views)
If these metrics are weak, you’re fighting the algorithm.
3. The Shopping Experience Is Your Competitive Advantage
E-commerce has an inherent advantage in the AI Mode era: shopping is experiential by nature.
But only if you design for it.
Traditional product grids where users search, filter, click isolated product pages, and leave? That’s low engagement. That looks like “didn’t find value” to the AI.
Visual browsing where users explore collections, discover related items, save favorites, and explore multiple products in a session? That’s high engagement. That signals “genuine value delivered.”
The stores that thrive in 2026 won’t be the ones with the best product descriptions (AI can summarize those). They’ll be the ones with the most engaging shopping experiences.
4. Mobile Experience Isn’t Optional
Google uses mobile-first indexing, and mobile users have even less patience for slow, clunky experiences.
If your mobile site loads slowly, has tiny buttons, requires pinch-to-zoom, or makes discovery difficult, you’re penalized heavily in AI Mode’s evaluation.
Test your mobile experience weekly. Fix issues immediately.
5. Authority and Trust Signal Real Value
AI Mode needs to know which sites to cite and recommend. Authority signals help:
- Customer reviews and ratings
- Clear about/contact information
- Author credentials on content
- External mentions and backlinks from reputable sources
- Transparent policies
- Secure checkout (HTTPS)
These aren’t just nice-to-haves. They’re how the AI determines if your site is trustworthy enough to recommend.
The Shift From Keywords to Experiences
The fundamental shift AI Mode represents is this: Google is moving from evaluating content to evaluating experiences.
For 20+ years, SEO was primarily about content optimization. Get the right keywords, structure your pages correctly, build quality backlinks, load fast. If you did those things, you ranked.
AI Mode changes the equation. Now Google asks: “After users reach this site, do they have a genuinely satisfying experience? Do they engage? Do they find value? Do they return?”
Content optimization still matters, but experience optimization matters more.
The merchants who recognize this shift early will dominate organic search in 2026. The ones who keep optimizing for 2015’s algorithm will wonder why their traffic keeps declining.
Start 2026 With the Right Strategy
As we head into 2026, ask yourself these questions:
If AI Mode answers my customers’ questions before they reach my site, why would they click through?
If you don’t have a compelling answer, neither will Google’s AI.
When users do reach my site, do they engage deeply or bounce quickly?
If they’re bouncing, the AI is learning your site doesn’t deliver value.
Am I designing for discovery and exploration, or just displaying products?
Discovery creates engagement. Engagement signals value. Value drives rankings.
The stores that thrive in 2026 won’t be the ones with the best keywords. They’ll be the ones with the most engaging experiences—because that’s what AI Mode rewards.
Want to see what high-engagement e-commerce looks like? Visit the Stylaquin demo store and experience how visual, magazine-style browsing creates the kind of engagement AI Mode recognizes as valuable.
Google’s AI is watching how users behave on your site. Make sure what it sees signals genuine value.
