How much has my CTR actually dropped because of AI Overviews?

It depends on how often AIO appears for your top queries. If 100% of your queries return AIOs, expect roughly 58% lower CTR on top-ranked pages based on Ahrefs' 300K-keyword study (December 2025 data). If about 16% of your queries return AIOs (the Semrush November 2025 average), the effective sitewide impact is much smaller, closer to 9%. Run your own query set through an AIO-tracking tool to get an accurate read for your specific traffic mix.

Should I opt out of AI Overviews for my content?

Usually not. Since August 31, 2026, Google's Search Console offers a control that excludes a site from AI Overviews and AI Mode without affecting ranking elsewhere in Search. But opting out doesn't remove the AI Overview from the results page; it only removes you from it. The clicks don't come back to your blue link, and you lose the citation surface. It can make sense for publishers negotiating content deals, less so for B2B sites that want visibility.

Are there query categories where CTR hasn't dropped?

Yes. Commercial-intent queries (comparisons, alternatives, vendor selection), niche long-tail buying queries, and queries requiring real-time data (stock prices, sports scores) have been much less affected. AIO summaries don't satisfy these intents well, so users still click through.

How do I track whether my content is being cited inside AI Overviews?

Tools like Otterly.AI and Profound automate this across major engines. For DIY tracking, define 30–50 representative seed queries and run them weekly through Google AIO, ChatGPT Search, and Perplexity. Record which sources appear in the citation chips. The data is noisy at first; signal emerges after a few weeks of tracking.

Will AI Overviews coverage keep growing?

Direction is up but not monotonically. Semrush's data shows coverage rising through mid-2025, then declining to 15.69% by November as Google tuned the system. BrightEdge's February 2026 data had AIOs on about 48% of tracked queries across nine industries, up from 30% a year earlier. The likely long-term trend is toward higher AIO presence on informational queries and lower presence on transactional and time-sensitive queries.

What about Bing and DuckDuckGo?

The same citation-first tactics apply to Bing's Copilot and DuckDuckGo's AI answers. For most B2B teams, start tracking with Google AIO, ChatGPT Search and Perplexity, then add Copilot if your buyers live in the Microsoft ecosystem.

Is the CTR drop affecting B2B SaaS as much as publishers?

Generally less. News and generic informational sites have been hit hardest because their content is most easily summarized. B2B SaaS, particularly companies with strong commercial-intent landing pages and product comparison content, has fared better. The direction is the same; the magnitude is smaller.

AI Overviews Killed Your CTR: The New Playbook

Yes, on informational queries AI Overviews have cut clicks roughly in half or worse. When an AI Overview appears, the top-ranked page earns 58% fewer clicks, per Ahrefs' study of 300,000 keywords using December 2025 data. Pages the AIO doesn't cite fare worst. The new playbook: get cited inside the AI Overview instead of chasing the click below it, restructure top-of-funnel pages for extraction, defend commercial-intent pages where AI summaries can't replace evaluation, and measure mention rate alongside sessions.

What the data actually says

The studies disagree on magnitude but not on direction. A short ledger of the published results:

Ahrefs, published February 2026 (source): analyzed 300,000 keywords from Ahrefs' Keywords Explorer database using December 2025 Search Console data. When an AI Overview is present, the top-ranked page's average CTR is 58% lower than on comparable informational keywords with no AIO. An update to their April 2025 study, which had measured a 34.5% drop.
Seer Interactive, April 2026 (source): 5.47 million queries across 53 brands, January 2025 to February 2026. On informational queries, per million impressions, pages earned about 33,500 organic clicks with no AIO, 20,743 when cited in the AIO, and 9,445 when not cited. That's roughly 38% fewer clicks when cited and 72% fewer when not.
Pew Research Center, July 2025 (source): 900 US adults' browsing in March 2025. Users clicked a traditional result on 8% of visits when an AI summary appeared versus 15% without one, and clicked a link inside the summary on just 1% of visits.
Authoritas, July 2025 (Press Gazette coverage): across 3,500 UK news queries collected in April 2025, CTR fell 47.5% on desktop and 37.7% on mobile when an AI Overview was present. Its headline estimate: a #1-ranked site could lose about 79% of its traffic for a query when it sits below an AIO. Submitted to the UK's Competition and Markets Authority; Google called it "inaccurate and based on flawed assumptions and analysis."
Semrush AI Overviews study (source): not a direct CTR measurement, but tracked AI Overview prevalence on 10 million keywords through 2025. Coverage rose from 6.49% in January to 24.61% in July, then settled at 15.69% by November. BrightEdge's February 2026 data put AIOs on about 48% of tracked queries across nine industries, up from 30% a year earlier.

The honest summary: studies measuring the per-query click impact mostly land between roughly 40% and 70% fewer clicks when an AIO is present, and the loss is steepest for pages the AIO doesn't cite. Studies measuring overall traffic loss for sites are noisier because they depend on how often AIO actually appears for the queries that drive the site's traffic, and that varies enormously by category.

Why the studies disagree on the headline number

Anyone who's spent time in SEO data knows that "the CTR drop is X%" is always shorthand for a much messier underlying question. Three reasons published numbers vary:

Different keyword sets. Ahrefs sampled across categories. Authoritas focused only on UK news queries. Seer measured client portfolios skewed toward specific verticals. Pulling a per-query measurement from any of them and applying it to "all queries" overstates the effect.

Different baselines. Some studies compare AIO-present vs. AIO-absent CTR for the same query over time. Others compare CTR for queries that have AIO vs. queries that don't. The two answer different questions and produce different numbers.

Different definitions of "click." Citation chips inside the AIO themselves are clickable, and some studies count those clicks toward the source's CTR while others don't. This single methodological choice can move a study's headline number substantially.

The takeaway is not that the CTR drop is fake. Every independent study finds it. The takeaway is that the precise number matters less than the direction, and the direction is unanimous: clicks to the un-cited blue link below an AIO have collapsed.

What has actually changed structurally

Three real shifts underlie the data, and understanding them is what makes the recovery playbook coherent rather than reactive.

1. Top-of-funnel traffic flipped from clicks to mentions

For informational queries (definitions, "what is X," "how does Y work"), users now get satisfied by the AIO and don't click. Your blog post that used to get 5,000 monthly visits from a definitional query may now get 800 visits and 4,000 brand mentions inside AI Overviews with no corresponding click.

That isn't worthless traffic. Brand mentions in AI answers do build awareness; the user reads "according to Veritas, ..." and registers the brand, even without clicking. But it doesn't show up in Google Analytics. If you measure success in sessions, your top-of-funnel "traffic" appears to be dying. If you measure in mention rate, it may actually be growing.

2. Commercial-intent traffic largely survived

The CTR collapse is not uniform across query types. Pages targeting commercial-intent queries (comparisons, alternatives, "best X for Y," vendor selection) have been much less affected, for a structural reason: AI summaries can't satisfy a buying decision the way they satisfy a definitional question. The user wants to evaluate, not just learn. They click through to vendor pages, comparison sites, and review platforms.

This is why categories like e-commerce, B2B SaaS, and review-driven verticals report smaller traffic losses than news and generic informational sites. The queries that pay are exactly the queries AI summaries are worst at replacing.

3. Uncited pages are functionally invisible

Here is the asymmetry that catches most teams off guard: when an AIO appears for a query and your page is not cited inside it, the page is functionally invisible regardless of rank. The user reads the AIO, sees the citation chips, considers them, and rarely scrolls past. Whether you ranked #2 or #20 below an AIO that didn't cite you barely matters. In Seer's April 2026 data, uncited pages earned less than half the clicks of cited ones.

This collapses the "long-tail SEO" strategy in any category where AIO has high coverage. It's no longer enough to rank; you have to be in the AIO's source chips, or you might as well not be on the page.

The recovery playbook

The playbook divides cleanly into top-of-funnel adaptation, commercial-intent reinforcement, and measurement reorientation.

Top-of-funnel: optimize for citation, not click

For informational pages where AIO has eaten your CTR, the goal shifts from "get the click" to "be the cited source inside the answer." This is a different optimization problem with a different set of tactics. We covered the full citation playbook in How to Get Cited in ChatGPT, Perplexity, and Google AI Overviews, but the short version:

Add statistics, quotations, and source citations. The Princeton GEO paper (KDD 2024) found these three methods lifted visibility in generative answers by 30–40% on its word-count metric, and citing sources raised a fifth-ranked page's visibility by 115.1% (Aggarwal et al.).
Lead every section with a definition or claim. Bury the narrative; surface the answer.
Add a substantive FAQ section with FAQPage schema. AI engines lift these directly into answers.
Build entity presence on Reddit, YouTube, LinkedIn, and Wikipedia, the four most-cited domains in Peec AI's 2026 analysis of 30 million sources cited by ChatGPT, Google AI Mode, Gemini, Perplexity, and AI Overviews (per Search Engine Land's coverage).

The accepted reality: on heavily-AIO'd informational queries, your blog post will get cited or it will not be read. There is no third option.

Commercial-intent: defend and reinforce

Pages targeting buying-decision queries are your CTR fortress. They've been less affected by AIO and they convert at higher rates. Spend disproportionate effort here:

Comparison pages ("X vs Y", "alternatives to X") tend to be heavily clicked because users want to evaluate, not summarize.
Pricing pages for any vendor are nearly impossible to summarize satisfactorily; the user needs the actual price, terms, and feature breakdown.
Listicle pages ("best X for Y") still earn high CTR because users perceive them as evaluation tools rather than answers.
Use case pages scoped to specific buyer personas (the long-tail commercial queries) are largely unaffected by AIO because they're too niche for AIO coverage.

These pages should be the highest-priority real estate in your site. Their conversion rate makes them the actual revenue drivers; their CTR resilience makes them the most defensible traffic. Audit them for completeness and update frequency.

Bottom-of-funnel: ramp up the layer AI can't replace

Below commercial-intent comes the layer AI summarization can't touch at all: trial signups, demo requests, sales conversations, customer onboarding. The CTR collapse upstream means the absolute volume of users reaching this layer is smaller, but the qualified rate of those who do is higher (they self-selected past the AIO summary).

For most B2B teams, this means investing more in product-led growth and bottom-of-funnel content. The traditional content funnel (broad awareness post → narrower consideration post → BOFU page) is leakier at the top than it used to be, but the bottom hasn't changed. The teams adapting fastest are reweighting toward bottom-of-funnel.

Measurement: track mention rate, not just sessions

If you measure success purely in Google Analytics sessions, AIO has decimated your top-of-funnel reporting. If you measure in citation rate across major AI engines, the picture is different and often more accurate.

The metrics that should be on every content team's dashboard in 2026:

Brand mention rate across a fixed set of seed queries, tracked per engine (Google AIO, ChatGPT Search, Perplexity, Gemini, Copilot).
Cited URL rate: the percentage of seed queries where a specific URL on your domain is in the citation chips.
Share of voice vs. competitors on commercial-intent queries.
Bottom-of-funnel conversion rate from AI-mentioned sessions: when users do click through after reading an AIO that mentioned you, do they convert better? (Empirically, in many B2B categories, yes; these visitors are warmer.)

Tools like Otterly.AI and Profound automate citation tracking (how the two compare). For teams with budget constraints, a manual seed-query audit weekly is a viable starting point.

What's not a fix

Some commonly-recommended responses to the AIO CTR drop don't work, and it's worth saying so.

More content. Volume alone doesn't move citation rates. The Princeton effect sizes are about what's in the content, not how much of it exists.
Aggressive on-page SEO. Keyword density and link velocity tactics that still move blue-link rankings have minimal effect on AIO citation. The selection algorithm reads for meaning, not for keyword distribution.
Hiding from AIO. The old levers, noindex and nosnippet (Google's docs), also remove or strip your regular listing. Since August 31, 2026, a Search Console control lets any site exclude itself from AI Overviews and AI Mode without affecting ranking elsewhere. Cleaner, but still not a fix: the AIO stays on the page, you just aren't in it.
Pivoting entirely to social. Social channels have their own discovery problems and don't compound the way owned content does. Reorienting away from search instead of toward citation-first content is throwing out the moat.

Closing

"AI Overviews killed your CTR" is a useful headline but a misleading frame for strategy. What actually happened is that search has fragmented into two surfaces with different physics. The classical SERP, where rank predicted clicks, still exists for commercial intent and time-sensitive queries. The AI-summarization surface, where citation predicts visibility and clicks are mostly into source chips, now dominates informational queries.

You don't recover from this by trying harder at the old game. You adapt by recognizing you're playing two games at once, scoring them separately, and reweighting your investment toward the surface where each query type actually lives.

Veritas generates content built for the new game: structured for citation, sourced on every claim, and ready for both human and machine readers. Try Veritas free or explore SEO Intelligence.

AI Overviews Killed Your CTR. Here's the New Playbook.