Google AI Overviews CTR is recovering: what the data says

Google AI Overviews CTR is recovering: what the data says

The early verdict on Google AI Overviews was grim: organic clicks were evaporating, publishers were furious, and SEO practitioners were quietly updating their resumes. But a cluster of recent studies suggests the story is more complicated. Across multiple independent analyses, CTR under AI Overviews appears to be recovering, partially, unevenly, and with meaningful implications for how brands structure their content and data.

Here is what the research actually shows.

Search Engine Land: the headline recovery finding

A Search Engine Land report from April 2026 found early signs that organic click-through rates beneath AI Overview results are beginning to recover after the sharp declines observed in mid-2024. The study noted that recovery is not uniform: informational queries with clear AI-generated answers continue to suppress clicks, while navigational and transactional queries show stronger CTR rebounds. This distinction matters enormously. If your content strategy is built around answering generic questions, you are still in the suppression zone. If you are optimizing for queries with clear commercial or navigational intent, the picture is measurably better.

BrightEdge: visibility gap by content type

BrightEdge research has consistently shown that structured, citation-ready content earns disproportionate placement in AI-generated summaries. Their data indicates that pages with explicit schema markup, clear entity relationships, and well-defined authorship signals appear in AI Overviews at significantly higher rates than unstructured content. The implication for schema practitioners is direct: the structured data you spent years arguing about in Slack is now doing actual defensive work. Proper schema markup is not just a rich-result play anymore. It is an AI visibility play.

Moz: authority signals still predict AI citation

Moz has documented that domain authority and topical authority signals remain strong predictors of which sources get cited inside AI Overviews, even as CTR patterns shift. Their analysis of citation patterns found that established, high-authority domains are cited roughly 3x more often than newer domains with equivalent keyword rankings. This is worth sitting with. You can rank for a query and still not get cited in the AI Overview that appears above your result. Authority signals and ranking signals are diverging, and GEO requires optimizing for both.

SparkToro: what users actually do after an AI Overview

SparkToro's audience research has examined post-AI-Overview user behavior and found that branded searches are more resilient to the zero-click problem than generic topic searches. When users already know what brand they want, they click through at rates close to pre-AI-Overview baselines. Unknown brands embedded in AI summaries show lower click conversion. This reinforces something uncomfortable: AI search rewards brands that already have recognition. If you are trying to build brand awareness through organic search, AI Overviews actively complicate the path.

Backlinko: query volume shifts under AI search

Backlinko's ongoing analysis of search behavior found that total query volume for informational long-tail searches is declining as users shift to conversational AI tools directly. This creates a compounding effect: fewer queries, lower CTR per query, and AI Overviews capturing intent that previously flowed to organic results. The recovery in CTR may partly reflect survivor bias. If low-intent queries are migrating off Google entirely, the remaining queries may simply be higher-intent ones that were always more likely to click.

Gartner: the enterprise measurement gap

Gartner has flagged that most enterprise marketing teams lack the measurement infrastructure to track AI-generated impressions and citations separately from traditional organic impressions. This creates a reporting blind spot: brands see declining organic traffic and attribute it to algorithm changes, when the actual cause is AI Overview suppression of specific query types. Measurement is the foundational problem. Tools like winek.ai exist specifically to make AI visibility measurable across engines, but the Gartner finding is a reminder that most teams are not using them yet.

Google Search Central: what Google itself says about AI Overviews and citations

Google's own documentation on AI Overviews emphasizes that structured data, clear page authorship, and E-E-A-T signals influence which pages get surfaced and cited. Google is explicit that AI Overviews are not a separate algorithm but an extension of core search quality systems. This means the fundamentals are not wrong. They are just insufficient on their own. You need structured data, authority signals, and content that is genuinely citation-worthy, not just keyword-optimized.

Key findings mapped to GEO implications

Source	Key finding	Implication for GEO practitioners
Search Engine Land	CTR recovering for navigational and transactional queries	Prioritize structured, intent-clear content over informational blogging
BrightEdge	Schema markup increases AI Overview citation rates	Schema is now a defensive GEO tool, not just a rich-result enhancement
Moz	High-authority domains cited 3x more often in AI Overviews	Authority building is now directly tied to AI citation share
SparkToro	Branded queries retain higher CTR post-AI Overview	Brand recognition is a GEO performance variable, not just a marketing metric
Backlinko	Informational query volume declining as users shift to AI tools	Informational content strategy needs fundamental reassessment
Gartner	Most enterprises cannot measure AI-generated impression share	Measurement infrastructure is a prerequisite, not an afterthought
Google Search Central	E-E-A-T and structured data influence AI Overview citation	GEO and SEO fundamentals overlap more than practitioners assume

CTR recovery by query type: what the data suggests

Query type	Estimated CTR trend	AI Overview frequency	Recommended GEO priority
Branded navigational	Recovering, near baseline	Low to medium	Maintain schema, entity clarity
Transactional commercial	Partial recovery	Medium	Product schema, review markup, clear CTAs
Informational how-to	Still suppressed	Very high	Citation optimization, not click optimization
Local intent	Recovering in some markets	Medium	Local schema, NAP consistency, entity signals
Comparison and review	Mixed, query-dependent	High	Structured comparison content, rating schema

The pattern across all this research

Taken together, these studies point to a search landscape that is fragmenting by intent type rather than collapsing uniformly. The early panic about AI Overviews destroying all organic CTR was an overstatement. The complacent counter-argument that nothing fundamental has changed is equally wrong. What is actually happening is a sorting process: queries where AI can fully satisfy intent are being absorbed by AI Overviews, while queries where users need to click somewhere to complete their task retain meaningful CTR.

The GEO implication is that content strategy needs to make a deliberate choice. You can optimize to be cited inside AI Overviews, which builds brand visibility and authority even without clicks. Or you can optimize for the query types that still drive clicks, which requires a harder look at your content mix and whether you are producing too much AI-absorbable informational content. The worst position is doing neither while hoping the algorithm sorts it out for you.

What practitioners should do next

1. Audit your content by query intent. Separate your traffic by informational, navigational, transactional, and local intent. The CTR recovery data is not evenly distributed, and your strategy should not be either.

2. Implement or audit schema markup across all high-priority pages. BrightEdge data makes a direct case that structured data increases AI Overview citation rates. If your schema is incomplete, inconsistent, or outdated, you are leaving AI visibility on the table.

3. Build a separate measurement layer for AI visibility. Organic impressions in Google Search Console do not tell you whether you are being cited in AI Overviews, mentioned in Perplexity, or surfaced by ChatGPT. Platforms like winek.ai measure AI-specific brand citation separately from traditional rank tracking.

4. Treat brand recognition as a GEO performance input. SparkToro's finding that branded queries retain CTR is not just a branding insight. It is a GEO optimization signal. Content that builds entity recognition and brand association with specific topics increases the probability that AI citations drive downstream clicks.

5. Stop optimizing purely for click volume and start tracking citation share. The success metric for AI search is not always a click. If your brand is cited authoritatively in an AI Overview for a high-volume query, that is a visibility event worth measuring, even if no click follows.

Frequently asked questions

Q: Is the recovery in Google AI Overviews CTR happening across all query types?

A: No. The recovery is concentrated in navigational and transactional queries, where users still need to click somewhere to complete their intent. Informational queries remain heavily suppressed because AI Overviews can fully answer them without a click. GEO practitioners need to segment their performance data by query type before drawing conclusions about whether their specific content is recovering or still in decline.

Q: Does schema markup actually affect whether a brand gets cited in an AI Overview?

A: Yes, according to BrightEdge research and Google's own documentation. Pages with clear structured data, explicit authorship, and well-defined entity relationships appear in AI Overviews at measurably higher rates than equivalent unstructured pages. Schema markup is no longer just a rich-result optimization tool. It directly influences AI citation eligibility, making it one of the most practical GEO interventions available to technical teams.

Q: How do you measure AI Overview citation share if Google Search Console does not show it?

A: Google Search Console shows impression and CTR data for organic results but does not break out AI Overview citations as a separate attribution. Dedicated measurement platforms like winek.ai track AI-specific brand visibility across engines by querying AI systems directly and monitoring citation patterns. Without this layer, most brands are flying blind on whether their GEO efforts are producing measurable results.

Q: Why are branded queries more resilient to AI Overview CTR suppression?

A: SparkToro's research indicates that when users already have brand intent, they click through at rates close to pre-AI-Overview baselines because the AI Overview does not fully satisfy their goal. They want a specific brand's page, not a summary. This means brand recognition functions as a direct CTR protection mechanism in AI search environments, which is a stronger business case for brand-building investment than most marketing mix models currently capture.

Q: If informational query volume is declining, should brands stop producing informational content?

A: Not exactly, but the strategy needs to shift. Informational content that is optimized purely for keyword rankings and organic clicks is becoming less viable. The same content, reframed as citation-ready reference material with strong entity signals, structured data, and genuine authorship, can earn AI Overview citations that build brand authority even without direct clicks. The goal for informational content in a GEO context is visibility share, not just traffic share.