The click collapse.
Here's a number that should terrify every marketing team in America: for every 100 clicks your #1 ranked page could earn two years ago, Google now keeps 58 of them.
That's not my number. It's Ahrefs', published in February after they analyzed 300,000 keywords across Google Search Console data. Seer Interactive ran their own study across 3,119 queries and 25.1 million organic impressions and got even uglier numbers — organic CTR dropped from 1.76% to 0.61% on queries where AI Overviews appear. That's a 65% decline.
Paid is worse. CTR on AI Overview queries dropped from 19.7% to 6.34%. You're still paying Google the same CPC. You're just getting a third of the clicks.
This isn't a blip. This isn't something Google will roll back. It's a permanent structural change to how search works. And most teams are still fighting it with the 2022 playbook.
What actually happened
Google rolled out AI Overviews in May 2024. By early 2026, AI Overviews appear on ~48% of all tracked queries and about 80–88% of informational queries. When they appear, the user's question gets answered right there on the results page — synthesized from 3-5 cited sources, presented as a confident paragraph, no click required.
Pew Research did the most rigorous controlled study: 68,000 queries, clicks dropped 46.7% when AI Overviews appeared. SparkToro puts 58% of all Google searches at zero clicks now. Gartner projects 25% of all organic search traffic will shift to AI chatbots entirely by end of 2026.
For every 100 clicks you could historically earn for a top-ranking page, Google now "keeps" 58. — Ahrefs, February 2026
And here's the real kicker — even queries without AI Overviews are seeing 41% CTR declines. Users are clicking less across the board. They've been trained by ChatGPT and Perplexity to expect answers, not links. Google is just catching up.
Why your dashboard looks fine (until it doesn't)
The reason this is catching teams off guard: your rankings are probably holding. Your impressions might even be up (Google now double-counts — you're "impressed" both in the AI Overview citation and in the regular results).
Clicks drop. CPL rises. Pipeline softens. But no single metric screams "we have a problem." Marketing looks the same on paper. Budget gets trimmed from channels that are actually working. Competitors who adapt gain an invisible advantage.
One case documented by Ahrefs: a client's impressions went up 27.56% year-over-year. Rankings improved 14%. Clicks dropped 36.18%. CTR fell from 5.98% to 3.35%. On a surface read, everything looks better. In reality, the top of funnel just got hollowed out.
Three things to do this quarter
1. Change what you measure
Traffic is no longer the right KPI. Neither are rankings in isolation. Start tracking:
Share of voice in AI citations. When a user asks ChatGPT, Perplexity, Gemini, or Claude about your category, do you appear? How often? Among which competitors? Tools like ZipTie and Seer's Generative AI Tracker do this automatically. You can also just run the queries manually in a spreadsheet once a week — takes 20 minutes and tells you more than your rank-tracking dashboard.
Branded search lift. AI Overviews increasingly cite brands by name. If your brand shows up in an AI answer, users often search your name directly afterward. That direct/branded traffic is the new "organic click" — you just have to instrument for it.
Assisted conversions from pre-qualified visitors. People who click through to your site after reading an AI Overview citation are already ~35% more qualified than generic organic traffic (they read the summary, made a decision, chose your link specifically). Track their conversion rate separately from regular organic.
2. Get cited in the AI answers your buyers see
This is the work. It's different from SEO, even though it looks similar from across the room. A few patterns that actually move the needle:
Original data beats rewritten insight. AirOps' research found that ~60% of AI Overview citations come from URLs that don't rank in the top 20 organic results. The models reward originality more than rankings. Publish a study. Run a survey. Analyze your own client data and share the findings. Become the source, not a source.
Freshness is weighted heavily. Pages not updated in 90+ days are 3× more likely to lose AI citations. Seer documented one client who refreshed existing content (no new pages, just updating what was already there) and saw a 300% increase in AI-driven traffic. Build a quarterly refresh cycle into your content calendar.
Answer-first structure. The model is scanning for a clean, citable passage — ideally 40-80 words, with the answer in the first sentence. Inverted pyramid writing. Lead with the conclusion. Structure matters more than word count.
Brand authority across platforms. LLMs were trained on Reddit, Wikipedia, review sites, news. Your brand needs to exist in all of them. A Wikipedia article is the single highest-leverage asset you can earn — if your business qualifies for one, get it.
3. Diversify off Google — now, not later
Use the budget bleeding out of AI-laden informational queries to test elsewhere. Specifically:
Lower-funnel search. Transactional queries still click at near-normal rates (AI Overviews only appear on ~16% of transactional searches, vs. 80%+ for informational). Double down on "buy," "pricing," "compare," "vs" queries where intent converts.
AI platform advertising. Perplexity, ChatGPT, and Meta's AI assistant are all rolling out ad inventory. Early movers are getting cheap CPMs. Get in before auction density catches up.
Direct traffic programs. Newsletter, podcast, community, webinars, referral. Channels you own that don't depend on a middleman algorithm. Every sophisticated B2B team I know is shifting budget this direction — not abandoning paid search, just rebalancing.
The uncomfortable truth
For the last decade, being #1 on Google was the gold standard of digital marketing. It's still worth something — but it's worth half of what it was worth two years ago, and in another 18 months it'll be worth even less.
The teams winning in 2026 aren't the ones who cracked AI Overviews. They're the teams who stopped optimizing for clicks and started optimizing for citations, share of voice, and pre-qualified traffic. Different game, different scoreboard.
If your 2026 playbook looks like your 2023 playbook with slightly more "AI" branding sprinkled on, you're going to lose ground. Quietly, steadily, while all your dashboards look fine.
Want a free audit of where you rank across AI engines (ChatGPT, Claude, Perplexity, Gemini) and how much pipeline you're leaving in AI Overview citations you could own? This is what we built Vortigen for. Get in touch and we'll run the numbers for you.
