AI Overviews CTR Just Rebounded 85%. Only Cited Brands Got the Clicks.
Seer's February 2026 data shows organic CTR on AIO queries jumped from 1.3% to 2.4%. The rebound is pointed at cited brands. Uncited pages keep losing.
AI Overviews CTR just rebounded 85%. Only cited brands got the clicks.
For 18 months the trend line ran one direction. Google launched AI Overviews. Organic click-through rates fell. Then they fell again. Then again. The narrative was simple: AI summaries answer the question on the SERP, so users stop clicking through.
In December 2025, organic CTR on AIO-present queries hit 1.3%. That was the floor.
In February 2026, it bounced to 2.4%. An 85% jump in two months. Seer Interactive published the numbers at the end of Q1 and the SEO world spent a week trying to figure out what changed.
The headline number is real. The conclusion most teams drew from it is wrong.
The rebound isn’t a recovery for the search ecosystem. It’s a redistribution to a small group of cited brands. Pages cited inside an AI Overview pulled in 2.1% CTR. Pages on the exact same SERP that weren’t cited dropped to 0.9%. The cited-versus-uncited gap is now wider than the AIO-versus-no-AIO gap used to be.
If you’ve been operating on the assumption that AI Overviews are a tide that lifts or sinks all boats, that mental model is dead. The tide already split into two currents — and one of them is moving backward.
What Seer actually measured
The dataset is large enough to take seriously. Seer analyzed 53 brands across 5.47 million distinct queries, 2.43 billion organic impressions, and 296.9 million paid impressions between January 2025 and February 2026. The methodology only counts queries with complete data across every month, which removes most of the noise that plagues smaller AIO studies.
Here is what the per-million-impression math looks like for informational queries, based on Seer’s own breakdown:
| SERP type | Clicks per 1M impressions | Effective CTR |
|---|---|---|
| No AIO present | ~33,500 | 3.35% |
| AIO present, brand cited | ~20,743 | 2.07% |
| AIO present, brand not cited | ~9,445 | 0.94% |
Being cited delivers 120% more organic clicks per impression than not being cited. That’s not a rounding gap. That’s the difference between a useful AI search channel and a dead one.
The paid side held up better than organic. With AI Overviews present, paid CTR climbed from 14.6% to 16.2%. Without overviews, paid CTR slipped from 26% to 21.8%. We covered the paid-vs-organic split in detail in our analysis of how ads, not AI Overviews, are doing most of the damage to organic search. The February data reinforces that story: ads keep collecting clicks because they sit above the AIO. Organic only collects when it survives inside it.
Which queries trigger AIO (and which don’t)
The 85% rebound number also hides which queries Seer was even measuring. AI Overviews don’t appear uniformly. They cluster on specific query shapes:
| Query format | AIO trigger rate |
|---|---|
| Comparison queries (“X vs Y”) | 95.4% |
| Question-format queries | 85.9% |
| “Near me” informational queries | 76.9% |
| Single-word informational queries | 27.3% |
| Brand navigational queries | 2-3% |
If your traffic mix leans heavily on comparison and question queries, you’re playing on a board where almost every result page is an AIO. The 0.9% uncited CTR is your default outcome. The 2.1% cited CTR is your ceiling. The traditional “rank in top 10 and collect clicks” path doesn’t exist there anymore.
If your traffic leans on brand navigational queries, you barely see AIOs at all. Most teams have both kinds of traffic, which is why aggregate dashboards keep showing mixed signals. The split is happening at the query level, not the site level.
Why the rebound happened
Seer is explicit that they cannot claim causation, and they call the trend a “leveling off” rather than a recovery. That caution is appropriate. Two months of data isn’t a trend. It’s a turn that might or might not hold.
A few plausible explanations for the December-to-February shift:
- User behavior is normalizing. The initial shock of AI Overviews drove users to accept the summary and bounce. Eighteen months in, more people have figured out that the AIO sometimes hallucinates, sometimes oversimplifies, and sometimes hides the better answer two paragraphs into a cited source.
- Google tuned the AIO down. There’s anecdotal evidence that Google scaled back AIO presence on some query categories where engagement was poor. Less coverage on borderline queries means the AIOs that remain are on queries where users actually want detail.
- Citations got more useful. Google has been iterating on how citations render and where they sit. More prominent cite chips mean clickable surface area, which directly raises CTR for cited brands.
None of these explanations are exclusive. All three could be contributing. The thing they have in common is that none of them help you if you’re not in the citation set.
What this changes for your strategy
The honest read of this data: getting cited has shifted from “valuable” to “the only path to clicks on AIO queries.” Here’s how that should change the work.
Stop benchmarking against the pre-AIO baseline. A 3.35% CTR on queries without AIOs is the only meaningful comparison if you’re trying to forecast traffic loss. But the practical bar in 2026 is 2.07% if you’re cited and 0.94% if you’re not. Your traffic model needs to account for which bucket each of your important queries falls into.
Stop optimizing pages. Start optimizing chunks. AI Overviews don’t pick “the best page.” They pick the best passage to answer a sub-query. We wrote about this shift in how Google AI Mode raids pages for parts, and the citation pattern in AI Overviews follows the same logic. A 200-word self-contained answer block in the middle of your page is doing more work than the entire page wrapped around it.
Prioritize comparison and question content. Those formats hit 95.4% and 85.9% AIO trigger rates. If you don’t have strong comparison pages and question-answering content, you’re invisible on the queries where users actually go looking. The good news: these formats are also the most concrete to write for. Pick a competitor matrix, build a clean comparison table with sources, and answer the specific question in the first 150 words.
Audit your citation gap monthly. The set of queries where competitors are cited and you aren’t is the highest-leverage worklist you have. If a competitor is getting 2.07% CTR and you’re getting 0.94% on the same SERP, the delta is real money. We covered some of the measurement framework in how Google rankings still control AI visibility — the same correlation logic applies here. Track which side of the citation line you’re on, query by query.
The risk of overreading two months of data
Seer’s own caveat deserves to be repeated: the linear regression model they use “underestimates in recovery environments.” Translation: if the rebound continues, the published numbers may understate it. If it reverses, they may overstate it. Two months isn’t a trend.
There’s also a selection effect. Higher-authority brands are more likely to be cited and more likely to rank, so the 120% click advantage for cited pages is partly a measure of which brands get to play. The citation mechanism itself probably explains some of the gap, but not all of it. Don’t read the 2.07% as a number you can hit by tweaking schema markup. Read it as the upper bound for brands that earn citations the hard way: external corroboration, content that answers specific sub-queries, and presence on the small set of source domains AI platforms actually trust.
The other thing to watch: paid CTR on AIO queries climbed during the rebound. If Google is quietly testing more aggressive ad placements above and around AI Overviews, the organic rebound could be a side effect of users hunting past the AIO to find a result they trust. That’s a fragile dynamic. The cited brand still wins, but the underlying mechanism could be Google’s monetization roadmap rather than user satisfaction with AI summaries.
The takeaway
The CTR collapse story is over. The redistribution story is what matters now. AI Overviews don’t suppress clicks evenly across the SERP. They concentrate clicks on cited sources and starve everyone else. A 2.07% CTR is what a cited brand earns. A 0.94% CTR is what an uncited brand survives on. The aggregate “85% rebound” headline is true only for the half of the SERP that won the citation lottery.
If you have a traffic forecast that assumes AIO presence affects all pages on the SERP the same way, it’s wrong. Build the forecast around citation status instead. That’s the dimension the data actually splits on.
RivalHound tracks your brand’s visibility across ChatGPT, Google AI, Perplexity, and more. Start monitoring to see where you stand.