AI Overviews now appear on roughly 48-50% of U.S. Google searches, up from just 6.49% in January 2025, Google itself has confirmed the figure sits around 50% of U.S. queries. That’s not a gradual shift, it’s a near eight-fold increase in about a year. But the popular framing, that AI has “killed” SEO, doesn’t match what the data actually shows. What’s changed is what earns a ranking, not whether ranking still matters.
What’s actually happening to clicks
The headline number is real: Ahrefs found AI Overviews reduce the organic click-through rate for the position-one result by 58%, and roughly 58-60% of all Google searches now end without a single click. That’s a genuine structural shift, not a temporary dip.
But the picture is more nuanced than a straight decline. Seer Interactive’s own tracking shows organic CTR on AI Overview queries, which had collapsed to a low of 1.3% in December 2025, recovering to 2.4% by February 2026, a real rebound even if still well below pre-AI-Overview levels. And critically, 99.5% of AI Overview citations still pull from pages already ranking in the top 10 of standard organic results, with close to 80% coming from the top 3 specifically. Traditional ranking signals haven’t become irrelevant, they’ve become the prerequisite for AI visibility rather than the finish line.
Why E-E-A-T matters more than ever
Google’s own helpful content guidance puts real experience, not just expertise, at the center of what it rewards. In practice, that means:
| What actually works now | Why it matters |
|---|---|
| Real case studies with specific numbers | Demonstrates genuine outcomes an AI system can’t fabricate on your behalf |
| Named author bios with real credentials | A direct, verifiable signal of expertise, not just claimed authority |
| Citations to credible external sources | Shows your content is grounded in something beyond opinion |
| Original data or a genuine point of view | The one thing generic AI-generated content can’t easily replicate |
This is the practical reason AI hasn’t displaced human-created content, an AI model can process information faster than any person, but it can’t have actually solved a client’s specific problem or watched a campaign succeed in real time. That lived experience is exactly what current ranking systems are built to detect and reward.
This shows up in a concrete way when you compare two similar pages. A generic explainer on “how SEO works” competes against thousands of nearly identical pages saying the same thing in slightly different words. A page built around a real client result, with a specific number, a named business, and an honest account of what worked and what didn’t, gives both search engines and AI systems something genuinely distinct to point to. The former is replaceable by the next AI-generated draft; the latter isn’t, because the underlying experience it’s describing actually happened.
How this connects to the bigger picture
This shift is part of the same discipline we cover in more depth in our Generative Engine Optimization and Answer Engine Optimization services, and in our breakdown of the top AI search optimization companies in Canada, including an honest look at where even our own approach has room to grow. If you want the concept explained visually, we also built a simple comparison of SEO, AEO, and GEO that shows how the three fit together rather than compete.
What’s actually working for businesses adapting well
Showing up across more than one platform. People increasingly discover content on ChatGPT, research further on YouTube, and fact-check on Reddit before ever visiting a website directly. Being visible only in traditional Google results now means being invisible for a meaningful share of the research journey.
Treating content quality as a real arms race. AI made content production faster for everyone, which paradoxically made deeply researched, expert content more valuable, not less, since it’s now competing against a much larger volume of generic material saying the same thing.
Not neglecting technical SEO fundamentals. With mobile now accounting for roughly 71% of all Google search traffic, page speed, proper schema markup, and Core Web Vitals remain non-negotiable regardless of how AI reshapes the front end of search.
A realistic starting point
Rather than a rigid 90-day plan, the practical priorities are straightforward: audit existing content against Google’s own helpful-content guidance, identify which of your keywords are already triggering AI Overviews using a tool like Ahrefs or Semrush, and make sure older content is actually updated, not just re-dated with no real changes. From there, building real topic depth and consistent branding across platforms matters more than chasing any single tactic.
Measuring what actually matters now
Traditional metrics still count, but they no longer tell the whole story on their own. Ranking position and click-through rate remain relevant, but tracking whether your brand actually gets cited or named inside an AI-generated answer is becoming just as important, and neither Google Analytics nor Search Console fully captures that on its own yet. Tools like Ahrefs’ Brand Radar are starting to fill that gap, showing which specific queries surface your business across ChatGPT, Gemini, and AI Overviews, and which competitors are being cited instead.
This is worth checking on a real schedule, not just once out of curiosity. A quarterly review of your most important service pages, run against the actual AI platforms your customers use, gives a clearer read on whether your content strategy is working than watching click-through rate alone, especially now that a meaningful share of valuable visibility happens without a click ever being recorded.
Leave a Reply