Quick answer. Google AI Overviews compress organic CTR by 30-50% on the queries they appear above. Most SEO teams call this catastrophic. We think they’re measuring the wrong thing. The brands losing CTR are simultaneously gaining: brand mentions inside the AI Overview itself (zero-click, infinite reach), citations in ChatGPT and Perplexity (the same content surfaces across engines), and higher-intent click-throughs (the people who do click after reading an Overview convert at 2-3× the historical rate). The CTR drop is real. The conclusion that it’s bad is not. SEOs measuring AI Overviews with 2018-era CTR dashboards are watching the wrong instrument. What replaces CTR as the headline KPI: brand-mention share-of-voice in AI surfaces, citation count across AI engines, and assisted-conversion lift from AI-influenced sessions. The brands that lose CTR and don’t measure the offsetting gains will fire their SEO teams. The brands that build the new dashboard will out-compete them.
Table of contents
- The CTR collapse is real
- Why we’re not panicking
- What you’re gaining when you “lose” CTR
- The three KPIs that replace CTR
- How to actually measure brand-mention share-of-voice
- What this means for your SEO program
- The honest caveat: this only works if you’re cited
- Stop optimizing for clicks. Start optimizing for mentions.
The CTR collapse is real
Let’s start with the data nobody disputes.
When an AI Overview appears for a query, organic CTR on positions 1-10 drops by 30-50% on average. That’s the consensus across the major published studies through Q1 2026. On informational queries it’s higher, closer to 60%, because the Overview answers the question directly. On commercial or navigational queries it’s lower, sometimes negligible, because users still need to click to transact.
Within the Resocial audit corpus across 200+ brands, we see almost exactly the same numbers. A typical pattern: a B2B SaaS client whose flagship “what is [category]” page held a stable 23% CTR for two years suddenly drops to 11% the week AI Overviews start appearing for that query. Traffic from that single URL halved.
Multiply that across the hundreds of queries where Overviews now appear, and the aggregate hit is significant. For one Resocial client in the fintech space, organic clicks dropped 28% year-over-year on otherwise unchanged rankings.
This is the data point that’s caused the panic. It is also the data point everyone keeps quoting in isolation.
Why we’re not panicking
Three things changed at the same time AI Overviews launched. CTR went down. Brand mentions went up. AI-engine citations went up. And we have a strong working hypothesis that downstream conversion rate per session went up too, though that one is harder to isolate.
The CTR drop is the loudest signal because every SEO dashboard surfaces it. The other three are invisible to anyone still running their measurement stack from 2018.
Consider what an AI Overview actually does. It reads your page (and dozens of others). It synthesizes a 2-4 sentence answer. It cites between 3 and 8 sources by name and link. Anyone who saw that Overview now knows your brand exists, knows you’re considered authoritative enough to be cited by Google’s own AI, and has the option to click through if they want depth.
The user who doesn’t click is not lost. They saw your brand name in a position of authority. That’s a brand impression that didn’t exist before AI Overviews, Google was not previously surfacing your brand name in the search results above your URL. It was just showing your URL. Now it’s showing your name, in an explicitly editorial context, alongside Google’s own answer.
The user who does click is a higher-intent user than before. They read the Overview. The Overview did not satisfy them. They want depth, comparison, examples, the things your full page can provide. They are closer to converting than the average organic visitor of 2022.
This is what the CTR-only dashboard cannot show you.
What you’re gaining when you “lose” CTR
Three measurable gains offset most or all of the CTR loss for brands that are positioned well in AI Overviews. Brands that aren’t cited get the worst of both worlds, that’s the genuine danger zone, and we’ll come back to it.
1. Brand mentions inside the Overview itself
Every AI Overview citation is a brand impression in a position no traditional SERP feature has ever offered. The brand name appears in a Google-curated editorial context with explicit attribution. This is roughly equivalent, in trust signal terms, to being quoted in a Wall Street Journal article, except it scales to thousands of queries per day at zero cost.
Conventional CTR measurement counts this as a loss because no click occurred. Brand-impression measurement counts it as a win because awareness compounded. Both are correct. One of them is the metric the next decade of SEO will actually be optimized for.
2. Cross-engine citation lift
Here’s the part most teams don’t realize: the content structure that earns Google AI Overview citations is the same content structure that earns citations in ChatGPT, Perplexity, Gemini, and Claude. Schema-rich, chunked, source-attributed, fact-dense pages get cited everywhere.
When a brand starts appearing in Google AI Overviews, we typically see Perplexity citation share rise by 20-40% within 60 days, and ChatGPT-cited brand mentions (where measurable via traffic referral or self-reporting from users) rise correspondingly. These aren’t separate optimization tracks, they’re the same content surface evaluated by different ranking systems.
Losing 30% of Google CTR while gaining presence in three additional AI engines is a net expansion of brand surface, not a contraction. The CTR dashboard cannot see this. The dashboard you should be building can.
For more on how AI engines decide what to cite, see our State of AI Search 2026 research and our analysis of the most cited Reddit threads in ChatGPT.
3. Higher-intent click composition
This is the gain that’s hardest to measure directly but the one that most directly affects revenue.
The users who click through to your page after reading an AI Overview are pre-qualified in a way that 2022 organic visitors were not. They’ve already absorbed the basic answer. They’re clicking because the basic answer was insufficient, they want depth, they want to evaluate you specifically, or they want to take action.
We see this pattern in client conversion rates. For one ecommerce client, organic sessions dropped 22% after Overviews appeared. Revenue from organic dropped 4%. Conversion rate per session jumped from 1.8% to 2.3%. The composition of clicks changed: fewer browsers, more buyers.
This is not universal, sites with weak conversion paths or thin product pages don’t see this lift. But for any brand whose product page actually answers the question the visitor came to ask, the post-Overview click is worth more than the pre-Overview click.
The three KPIs that replace CTR
If CTR is no longer the headline indicator of SEO program health, what is? We propose three KPIs that, taken together, give a defensible view of AI-era organic visibility.
KPI 1: AI-Surface Share of Voice (AI-SOV)
The percentage of relevant queries where your brand is mentioned in any AI surface, Google AI Overviews, ChatGPT, Perplexity, Gemini, Claude, regardless of whether a click follows.
Measurement: maintain a tracked query set of 100-500 strategically chosen prompts. Query each engine on a monthly cadence. Count brand mentions. Express as a percentage.
This is the closest thing to a 2026-equivalent of “first-page rankings”, a measure of presence in the surfaces users actually consult.
KPI 2: Citation-Weighted Authority Score
Not all citations are equal. Being cited as the source in an AI Overview is more valuable than being one of seven sources cited. Being cited by name (with brand attribution) is more valuable than being cited only by URL. Being the only branded source among Wikipedia and Reddit is more valuable than being one of three branded sources.
Score each citation along these dimensions, weight, and aggregate monthly. Track the trend, not the absolute number, the absolute number is unstable as AI engines change their UI.
KPI 3: Assisted Conversion Lift from AI-Influenced Sessions
The hardest to measure but the most defensible to leadership. Compare conversion rate on organic sessions where the user also had a Google or AI-engine impression in the assisted-conversion window vs sessions without. The lift quantifies the offsetting revenue gain when CTR drops.
This requires GA4 conversion tracking properly configured and ideally a server-side attribution layer. It’s not free to set up. It is increasingly the only honest answer to “is SEO still working” in the AI era.
For the operational framework that ties these together with the work that actually moves them, see Our methodology and the GEO service overview.
How to actually measure brand-mention share-of-voice
This is the one most teams haven’t built yet, so a tactical aside.
Build a tracked query set of 100-500 prompts. These should reflect what your prospects actually ask AI engines, not what you wish they asked. Source them from:
- Your top-100 SEMrush/Ahrefs queries by traffic value (still useful as intent proxy)
- Customer interview transcripts and sales call recordings
- Live ChatGPT logs from your team (yes, ask your own team what they search)
- Reddit and Quora questions in your category
- Your competitor’s tracked keyword lists, if observable
Query each AI engine, ChatGPT, Perplexity, Gemini, Claude, with the same prompts monthly. Record whether your brand is mentioned, in what context (cited source, mentioned in passing, mentioned negatively), and what alternatives are surfaced alongside you.
Express as a percentage of queries where you appear, broken out by engine. Track the absolute number and the trend. The trend matters more than the absolute number, absolute numbers will be noisy as the engines change.
This is what AI-SOV looks like operationally. It’s not magic. It’s just discipline applied to a measurement surface most teams haven’t built yet.
What this means for your SEO program
If you accept this framing, several things change about how you allocate SEO investment.
Stop optimizing for #1 rankings on AI-Overview queries. They’re worth less than they used to be. Rank well enough to be one of the cited sources, then move on. Marginal effort to go from rank 4 to rank 1 is mostly wasted because the Overview captures the answer regardless.
Start optimizing for citability. Schema markup, fact-density, source attribution, structured chunks, these earn AI Overview and cross-engine citations. They are the new “title tag optimization” in terms of leverage.
Diversify your traffic ambition across surfaces. Treat ChatGPT, Perplexity, and Gemini as distribution channels worth tracking and influencing, not as adjacent curiosities. Your distribution is now multi-engine.
Rebuild your dashboard. Replace CTR-as-headline with AI-SOV-as-headline. Keep CTR as a secondary metric for diagnosing query-level changes, but stop reporting it to leadership as the indicator of program health.
Communicate the shift to your CMO. This is the part SEO leaders most often skip. If you don’t reset the measurement frame with your stakeholders, you’ll be defending a CTR drop you can’t reverse against a brand-mention gain they can’t see. You’ll lose that conversation. Have it on your own terms.
The honest caveat: this only works if you’re cited
We want to be clear about the caveat that makes this whole argument conditional.
The brand that loses CTR and isn’t cited in any AI surface gets the worst of both worlds. Traffic down, brand mentions flat, AI engine presence zero. That brand should panic. Their SEO program is failing on both the old and new metrics simultaneously.
The brand that loses CTR but is cited prominently across AI surfaces is fine, often improving. They just need to rebuild the dashboard to see it.
The difference between the two outcomes is content structure, schema, entity authority, and the boring infrastructure work of being citable. It is not luck. It is also not optional anymore.
For the operational checklist of what makes a brand citable, see our analysis of the 12 most common GEO mistakes we see in live audits and the complete guide to AI search optimization.
Stop optimizing for clicks. Start optimizing for mentions.
CTR was the right headline KPI for the 2010-2024 era of search. It mapped cleanly to revenue because the click was the only available conversion event from search. In 2026, the click is one of several outcomes, and not the most valuable one for many query types.
The brands that adapt their measurement and their optimization to this reality will look like they’re winning SEO. The brands that don’t will look like they’re losing, even when the underlying business outcomes are actually fine, even when their content is being read by millions of users via AI engines, even when their brand mentions are compounding.
The CTR drop is real. The conclusion that it’s catastrophic is a measurement artifact. The honest assessment of where AI Overviews leave a well-positioned brand is: smaller traffic number, larger brand surface, higher-intent visitors, and entry into three additional citation surfaces.
We’ll take that trade.
If you want to know whether your brand is actually positioned to win the new metric, or whether you’re in the danger zone of losing CTR without offsetting gains, start with our Free SEO Audit, or describe your situation and we’ll map you to the closest service.
Yuki leads Resocial’s GEO research practice. Petros leads measurement and analytics infrastructure. They’ve spent the last 18 months rebuilding the Resocial measurement stack for AI-era SEO and apply it across 200+ brand audits annually.