Search hasn't been "disrupted" in some abstract sense. Specific, dated things happened this year that changed how visibility works, and most SEO content published even a few months ago is already missing them.
If you type "best CRM for small business" into Google today, you're less likely to get a list of blue links and more likely to get a direct AI-generated answer with a shortlist and reasoning attached. That much hasn't changed since earlier this year. What has changed is more specific, and more actionable, than most people realize.
What Just Happened: Google I/O 2026 and What Changed After
A handful of concrete developments landed this year that most "SEO in 2026" content, including our own earlier coverage, was written before.
Gemini 3.5 Flash became the default AI Mode model.
At Google I/O on May 19, Google announced Gemini 3.5 Flash as the new default model powering AI Mode globally, with faster, more capable query handling across desktop and mobile.
Google launched "information agents." These run in the background, continuously monitoring the web, news, and real-time data (finance, shopping, sports) on a topic you care about, and surface updates without you having to re-search. This is Google search shifting from a query-response tool toward something closer to an ongoing subscription to a topic.
"Preferred Sources" now directly shapes AI Overview visibility.
In a May 26 interview with The Verge's Decoder podcast, Google CEO Sundar Pichai confirmed that Google uses long-term user satisfaction signals, engagement, return visits, and bounce-back behavior, to calibrate what AI systems show. Google has since let users mark specific publishers as "Preferred Sources," and that selection now factors directly into AI Mode and AI Overview visibility, not just traditional rankings.
Search Console now reports AI visibility, and lets you opt out.
Google has rolled out new Search Console reporting that shows impressions specifically inside AI Overviews and AI Mode, separate from standard organic data. More recently, it added a setting letting site owners exclude their content from AI Overviews, AI Mode, and Discover's AI features specifically, without affecting normal indexing or rankings.
Google's spam policy now explicitly covers AI manipulation.
Following a May clarification, Google's spam policies extend to attempts to manipulate generative AI responses, not just traditional rankings. An August 2026 Spam Update began rolling out on August 18, confirmed via Google's Search Status Dashboard.
Old SEO vs. AI-Era SEO: A Straight Comparison
Neither column replaces the other. Google still handles roughly 5 billion users and 16.4 billion searches daily, dwarfing every AI assistant combined. Navigating both channels simultaneously is the defining challenge for the future of link building strategies. But the two columns are now measured separately, and optimizing only for the left one means missing a growing share of how people actually find you.
The Real Click Data (Not the Scary Headline Version)
A lot of AI Overview click-through-rate stats floating around right now are wildly inconsistent, some articles cite a 34.5% drop, others cite 58%, others 61%. That range itself is worth understanding rather than picking whichever number sounds most alarming.
Pew Research, measuring actual user behavior, found a click rate of 8% on searches with an AI Overview present, versus 15% without one, a relative decline of roughly 47%. Ahrefs' analysis of ranking data found the top-ranking page for AI Overview-triggering keywords sees about 58% lower click-through than it would without the Overview present. Seer Interactive, which has tracked this longest, documents organic click-through dropping from 1.76% to 0.61% year-over-year, with a tentative stabilization showing in early 2026 data.
The honest takeaway: the exact percentage depends on the study and the query type, but every credible source agrees on the direction and rough magnitude, a meaningful, double-digit decline in clicks on queries where an AI Overview appears.
There's a flip side worth knowing: pages that do get cited inside an AI Overview can see click-through increases of up to 35% compared to non-cited competitors on the same query. The click isn't gone, it's redistributed toward whoever gets named.
Who's Actually Winning Citations Right Now
Citation sources vary a lot by category, and knowing where matters more than a general "be citable" strategy.
That last point matters specifically for SaaS: if you're optimizing content for AI visibility in a software category, a G2 profile and active presence there may do more for your citation odds on comparison queries than another blog post will.
The pattern across all of it: AI Overviews increasingly cite pages outside the traditional top 10, and increasingly synthesize from three to five sources per answer rather than leaning on one dominant page. Understanding how AI search chooses which sources to trust explains why ranking #1 no longer guarantees inclusion in the answer built from that ranking.
What We've Seen Play Out in Our Own Client Work
To be upfront: the campaigns below were measured on rankings, keywords, and traffic, not AI Overview citation specifically, since that's a newer tracking discipline most tools (including ours) are still catching up to. But the underlying principle connects directly to what the current AI-citation research shows, and it's worth laying out honestly.
Ahrefs' own analysis of 75,000 brands found that branded web mentions, being talked about across relevant sites, correlate with AI Overview visibility roughly three times more strongly than raw backlink counts (0.664 vs. 0.218). That's a relevance-and-mentions story, not a link-volume story.
That's exactly the pattern we've seen drive traditional ranking results too. One SaaS client started at a Domain Rating of 11 with 10 ranking keywords. By prioritizing niche-relevant placements over the biggest names available, guest posts, broken link building, and the skyscraper technique aimed at sites already covering their space, they reached a Domain Rating of 61, ranked for 208 keywords, and grew from roughly 1,000 to 19,000 monthly visitors over nine months.
A communication SaaS client saw a similar pattern: shifting from general, high-authority-but-irrelevant links toward niche-specific placements drove a 199% traffic increase and nearly doubled their keyword count in eight months.
Neither campaign was built with "get cited in AI Overviews" as the explicit goal, that framework didn't fully exist yet when the work started. But the mechanism, being genuinely, repeatedly associated with the right topic in the right places, is the same mechanism the current AI-citation data points to. We think that's not a coincidence: relevance-first work was already the right call before AI search made it measurable in a new way.
What to Actually Do About It
Frequently Asked Questions
Does SEO still matter in 2026, or should I focus entirely on AI search optimization?
Both matter, and they're not the same discipline anymore. Google still processes an estimated 16.4 billion searches a day, far more than all AI assistants combined, so traditional ranking fundamentals remain the foundation. But AI Overview citation is now a separate, trackable layer on top of that foundation, and ignoring it means missing a real and growing share of visibility, especially on informational queries.
How do I know if I'm getting cited in AI Overviews?
Google's Search Console now includes reporting specifically for AI Overview and AI Mode impressions, separate from standard organic data. Check there first before assuming you have no visibility, many sites are cited without realizing it since traditional rank trackers don't capture this layer.
What is a Preferred Source, and does it actually affect rankings?
A Preferred Source is a publisher a user has explicitly chosen to see more often in their results. Google has confirmed this selection influences AI Mode and AI Overview visibility directly. It doesn't affect traditional organic rankings, but it's a genuine, newly available lever for AI-specific visibility.
Why do some pages get cited in AI Overviews even when they don't rank in the top 10?
AI systems select sources based on content structure, claim clarity, and entity authority, not purely on organic rank. Citation overlap with the traditional top 10 has weakened over the past year as AI systems increasingly pull from a wider set of sources per answer.
Do backlinks still matter if brand mentions are what drives AI citation now?
Yes, but the mechanism has shifted. Ahrefs' research on 75,000 brands found branded mentions correlate with AI Overview visibility roughly three times more strongly than raw backlink counts. Relevant, contextual links still build the kind of topical authority that drives both rankings and mentions, it's volume-focused link building specifically that's lost ground—ultimately rewriting modern standards for AI vs. human content evaluation.
What should I do first if I haven't adjusted my strategy for AI search yet?
Start with your Search Console AI reports to see where you already have (or lack) visibility, then prioritize category-specific presence (G2, LinkedIn, Reddit, depending on your industry) alongside your own content, since a growing share of citations now come from outside a brand's own domain entirely.
Takeaway
Search didn't get simpler this year. It got more measurable, in ways that reward the same relevance-first thinking that's always driven good SEO, now with new tools to actually track it. If you want help figuring out where your brand currently stands across both layers, that's exactly what we help clients with at LinkyJuice.
