A citation in a Google AI Overview is associated with more clicks than appearing on an AI Overview result without one. That is not the same as proving the citation created those clicks. Seer Interactive's April 2026 update tracked 53 brands, 5.47 million queries and 2.43 billion organic impressions from January 2025 through February 2026. Its authors report 120% more organic clicks per impression for cited brands than uncited brands within AI Overview results. They also report cited results still underperformed results without an AI Overview by 38%. These are cohort associations, not an experiment in which otherwise identical pages were randomly cited.
Three comparisons, three different questions #
| Comparison in Seer's cohort | What it can tell you | What it cannot tell you |
|---|---|---|
| AI Overview present, brand cited versus not cited | The observed CTR difference between those groups on AI Overview results | The incremental clicks that obtaining a citation would cause for a given page |
| AI Overview present versus absent | How CTR differs across the tracked query groups | How the same query would perform if Google removed its AI Overview |
| Organic versus paid CTR | Whether the two channels moved together in this cohort | That source citations determine ad performance |
The distinctions matter because Google does not assign AI Overviews or citations at random. Query intent, ranking, brand familiarity and the likelihood of a click differ between the groups. Seer explicitly labels AIO status from the most recent SERP snapshot and applies that label retrospectively to every month. That makes the long trend useful for directional comparison, but not a month-by-month history of which brand was actually cited then. The sample also keeps only search terms with complete SERP, paid and organic data across the study window. Neither selection rule supports a universal CTR benchmark for every site.
The 2026 reversal changes the story #
Seer's cohort did not continue declining in early 2026 as its earlier model projected. Organic CTR on AI Overview-present queries rose from 1.3% in December 2025 to 2.4% in February 2026; CTR where no AI Overview appeared rose from 2.8% in January 2025 to 3.8% in February 2026. Two months do not establish a permanent recovery. The authors ask readers to inspect their own segments before forecasting one.
The paid side did not uniformly collapse when an AI Overview appeared. In this cohort, paid CTR on AIO-present queries moved from 14.6% to 16.2% over 2025, while paid CTR on no-AIO queries fell from 26% in January 2025 to 21.8% in February 2026. These groups differ in intent and baseline click propensity; the contrast is not proof that an AI Overview helped ads.
Exposure itself varies. Among Seer's tracked informational queries, 36% showed an AI Overview; among transactional queries, 5% did. Within the informational set, 267 of 280 comparison queries (95.4%) and 1,214 of 1,413 question-format queries (85.9%) triggered one. Those denominators are the study's tracked query sample, not rates for all Google searches. A program dominated by comparison questions therefore needs a different click expectation from one dominated by transactional searches.
Citation presence is not an entity-mass experiment #
An earlier Seer study found organic CTR of 0.74% versus 1.02% and paid CTR of 7.89% versus 11% in its comparison of brands absent from versus present in AI Overviews. Seer itself notes that branded query mix could contribute to the difference. The 2026 update expands the cohort and changes the question; it should not be spliced into the older percentages as one continuous time series.
The Machine Relations Index measures observed source selection across answer engines. Its citation rates, engine coverage and release window can show whether a domain was selected as a source. They do not measure Google Search Console clicks, paid CTR, brand recognition or the causal effect of an additional citation. Calling cross-engine citation breadth “entity mass” does not bridge those datasets. A domain could gain AI citations while click volume stays flat; Seer observed cases where impressions more than doubled while clicks stayed flat, lowering CTR without fewer clicks.
To evaluate a specific site, segment its own Search Console and ad data by query intent, AIO presence and citation status; compare clicks and impressions over the same window. Record ranking and brand versus nonbrand mix, and distinguish an observed association from an intervention's effect. Use the Index for source-selection evidence and the site's first-party analytics for clicks. Neither can substitute for the other.
FAQ #
Does an AI Overview citation increase organic CTR? #
In Seer's 2026 cohort, cited brands had 120% more organic clicks per impression than uncited brands within AI Overview results. The retrospective status label and differences between queries mean the study does not estimate how much CTR a particular page would gain if it earned a citation.
Does an AI Overview always reduce paid CTR? #
No. Seer's paid CTR for AIO-present queries was relatively stable and rose from 14.6% to 16.2% over 2025 in its tracked cohort, while the no-AIO group declined. That comparison does not isolate the AI Overview's causal effect on ads.
Can the Machine Relations Index predict click recovery? #
No. The Index measures which sources engines cited in observed runs; it does not measure Google search clicks or paid performance. Compare its dated source-selection observations with separately measured first-party click data rather than converting citation rate into a CTR forecast.