Google’s AI Overviews have quietly become the biggest variable in search visibility, and on August 29, 2026, Semrush pushed out the clearest public breakdown yet of what it takes to show up inside them. The company’s research arm has spent the past several months tracking how Google’s AI-generated answers behave across the web, and the numbers it’s now circulating paint a search landscape that looks nothing like the one marketers optimized for even two years ago. Coverage of the findings has spread quickly, picked up by outlets including Revista Merca2.0 alongside US-based search trade press, which is itself a sign of how global the anxiety over AI-driven search has become.
The headline figure driving the conversation: Semrush’s expanded 2026 AI Visibility Index, released June 26, 2026, analyzed 126 million U.S. AI search prompts across ChatGPT, Gemini, Google AI Mode, and Google AI Overviews, collected between January and April 2026. That’s a massive jump from the 2,500-prompt pilot Semrush ran when it first started measuring AI search behavior, and the scale of the new dataset is exactly why the company’s newest guidance is being treated as a rulebook rather than a suggestion. For an industry that built two decades of practice around ranking in ten blue links, the shift to optimizing for an AI-generated paragraph is forcing a rewrite of first principles.
Don't miss new tech stories on Google
Add Tech Insider once in the Google app and our stories appear in your news suggestions.
What Semrush’s 2026 AI Visibility Index Actually Found
Semrush’s index is built by running a large, repeated panel of real-world queries through Google and the major AI assistants, then logging which sources get cited, summarized, or ignored. According to Semrush’s own AI Overviews explainer, the feature now reaches roughly 2 billion monthly users, and Semrush Sensor data puts AI Overviews appearing in about 12.95% of search queries on average in the U.S. market. Other Semrush-linked figures from 2025 tracked a steeper early trajectory: AI Overviews showed up in 13.14% of all queries in March 2025, up from just 6.49% in January of that year, a doubling in roughly ten weeks.
Separate data compiled from Ahrefs paints an even more aggressive picture for the U.S. market specifically: AI Overviews were present on 48% of all U.S. Google Search queries as of March 2026, up from 34.5% in December 2025. That’s nearly half of all searches now returning an AI-generated summary before a single traditional listing appears on screen. Whichever dataset is closest to reality for a given site, the direction is the same, and it’s why Semrush’s May 19, 2026 post on Google’s own guide to optimizing for generative AI search landed with more urgency than typical SEO commentary.
One detail that hasn’t gotten as much attention: Semrush’s 2025 research also tracked a category most marketers ignore, navigational searches, meaning queries where someone is looking for a specific brand or site by name. AI Overviews started appearing on those too, climbing from 0.74% of navigational searches in January 2025 to 10.33% by October. That’s a meaningful shift because it means AI Overviews are no longer confined to broad informational questions, they’re inserting themselves into searches where users already know exactly where they want to go.
How Often AI Overviews Now Appear, Market by Market
AI Overview prevalence isn’t uniform, and that unevenness is itself part of the story. A Semrush study of European search behavior in 2025 found wide swings by country: AI Overviews reached 17.5% of queries in Portugal and 16.8% in Belgium, while Germany stayed under 1%, a gap Semrush attributes to Google’s cautious, market-by-market rollout pace tied to language support and regulatory posture. In that same European dataset, 88.1% of the queries that triggered an AI Overview were informational in nature, reinforcing that “how,” “why,” and “what is” queries remain the feature’s core territory even as it expands into navigational search.
| Market / Metric | AI Overview Presence | Source | Period |
|---|---|---|---|
| United States (Ahrefs dataset) | 48% of queries | Ahrefs | March 2026 |
| United States (prior baseline) | 34.5% of queries | Ahrefs | December 2025 |
| United States (Semrush Sensor) | 12.95% of queries | Semrush | 2026 average |
| Global estimate | ~13% of queries | Industry recap of Semrush data | Q1 2026 |
| Portugal | 17.5% of queries | Semrush | 2025 |
| Belgium | 16.8% of queries | Semrush | 2025 |
| Germany | Below 1% of queries | Semrush | 2025 |
The spread between the Semrush Sensor figure (12.95%) and the Ahrefs-derived figure (48%) for the same U.S. market is large enough that it’s worth flagging directly rather than glossing over: different crawlers, panels, and query sets produce different percentages, and neither company claims to be measuring the exact same query universe. What both agree on is the trend line, up, and fast, which is the part that matters for anyone deciding how to allocate SEO budget for the rest of 2026.
The Click-Through Collapse Behind the New Rules
The reason Semrush’s guidance is landing as urgent rather than routine comes down to what happens to traffic once an AI Overview appears. Ahrefs’ updated research found that the presence of an AI Overview correlates with a 58% lower average click-through rate for the page ranking in position one, compared with the same position when no AI Overview is shown, as of December 2025. That’s not a marginal dip, it’s a majority of clicks disappearing from the single most valuable ranking slot in organic search.
Data from Seer Interactive, as reported by Search Engine Land, found the damage even sharper for informational queries: organic CTR fell from 1.76% to 0.61%, a 61% decline, when an AI Overview was present. Seer’s more recent tracking shows some recovery, from a low of 1.3% in December 2025 up to 2.4% in February 2026, but the underlying mechanics remain brutal for anyone not cited directly. In that same Seer dataset, CTR ran around 3.3% on queries without any AI Overview, dropped to about 2.1% when a page was cited inside the AI Overview, and collapsed further to roughly 0.9% when a page ranked normally but wasn’t cited at all. Being cited, in other words, is now worth roughly double the traffic of being merely ranked.
Pew Research’s widely cited study on AI summaries found that users clicked through to a traditional web result only 8% of the time when an AI summary appeared in their results, compared with 15% of the time when no summary was shown, effectively halving the odds of a click. Additional figures circulating in 2026 industry recaps put the zero-click rate for AI Overview searches at roughly 64.82%, and even higher, around 93%, for searches conducted inside Google’s newer AI Mode surface. Read together, these numbers describe a search experience that increasingly resolves the user’s question on the results page itself, with the underlying websites reduced to background citations rather than destinations.
| Metric | Value | Source |
|---|---|---|
| CTR drop, position 1, with AI Overview present | -58% | Ahrefs |
| CTR drop, informational queries with AI Overview | -61% (1.76% to 0.61%) | Seer Interactive / Search Engine Land |
| CTR without AI Overview | ~3.3% | Seer Interactive |
| CTR with citation inside AI Overview | ~2.1% | Seer Interactive |
| CTR ranked but not cited | ~0.9% | Seer Interactive |
| Traditional result click rate with AI summary present | 8% | Pew Research |
| Traditional result click rate with no AI summary | 15% | Pew Research |
| Zero-click rate, Google AI Mode | ~93% | Industry recap of 2026 data |
Semrush’s New Rules for Appearing in Google AI Overviews
Against that backdrop, Semrush’s guidance functions less like an optimization checklist and more like a survival kit. The company doesn’t present a single official Google ranking algorithm for AI Overviews, because Google hasn’t published one, but its recurring recommendations, drawn from its May 19, 2026 breakdown of Google’s own generative search guidance and its broader 2026 AI Visibility Toolkit documentation, cluster around a consistent set of themes. None of them replace traditional SEO. All of them sit on top of it.
Structure Content So It Can Be Summarized
Semrush’s core argument is that AI Overviews work by synthesizing multiple relevant sources into a single answer, which means content that’s already organized into clear, extractable chunks, direct answers up front, scannable subheadings, defined terms, has a structural advantage over dense narrative prose. The system isn’t reading a page the way a human does, it’s pulling discrete facts out of it, and pages that make those facts easy to isolate get pulled more often.
Strengthen Entity and Topic Coverage
Semrush frames AI visibility less around exact-match keyword ranking and more around entity and topic strength, brand mentions, topic clusters, and the density of related queries a site can credibly answer. A site with deep, interlinked coverage of a subject reads as more authoritative to a system that’s synthesizing an answer from many sources than a single well-optimized page competing on one keyword.
Win Citations, Not Just Rankings
This is arguably the most consequential shift in the guidance. Semrush’s tools now track citation frequency and cited-page data as core AI visibility signals, separate from traditional organic rank tracking, because a page can rank on page one and still never get pulled into an AI Overview, and a page ranking lower can still be cited if it’s structured and sourced well. Semrush explicitly recommends inspecting which content formats surface most often in generative answers and reverse-engineering toward those formats.
Monitor Freshness and Movement Daily
Because AI Overview citations can shift quickly as Google’s systems re-crawl and re-synthesize, Semrush recommends tracking visibility gains and losses on a day-to-day cadence rather than the monthly reporting rhythm most SEO teams are used to. A page cited today isn’t guaranteed a citation next week, and Semrush’s Position Tracking now includes a dedicated AI Overview filter for exactly this kind of monitoring.
Keep Pages Technically Accessible to AI Crawlers
Semrush’s Site Audit tooling now flags whether firewalls or bot-management rules are inadvertently blocking the crawlers that feed AI Overviews and other generative engines, a technical SEO check that didn’t exist in most audits two years ago. Sites that lock down aggressively against scraping can end up invisible to the very systems increasingly deciding what gets shown to searchers.
Track Presence Across Every AI Surface, Not Just Google
The 2026 AI Visibility Index doesn’t stop at Google. It spans ChatGPT, Gemini, Google AI Mode, and Google AI Overviews as distinct, separately tracked surfaces, and Semrush’s toolkit extends that same competitive-gap tracking to Perplexity as well. The implication is straightforward: a brand optimizing only for classic Google rankings is now missing several other places where an AI system might be answering a question about it, or about a competitor, instead.
Google’s Side of the Story: AI Mode, Top Stories, and Ads
Google hasn’t been static while third parties measure its systems. Its own generative AI search documentation lays out how AI Overviews and AI Mode are meant to work, and its developer guidance on AI features gives site owners the technical controls available to them today. Search Engine Land reported that Google rolled out Top Stories modules inside AI Overviews in July 2026, giving news publishers a narrower, more visible lane inside the AI-generated answer rather than being blended anonymously into the summary text. Semrush’s own coverage also flagged that Google was testing advertising placements inside AI Overviews in early 2026, a move that, if it scales, would mean the AI summary box competes for revenue with the very organic listings it’s already suppressing. Layered on top of all this is Google’s continued push of AI Mode, a more conversational, multi-turn search experience running alongside classic Search and built on Gemini, which Semrush’s index now tracks as a separate destination with its own citation patterns and its own, even steeper, zero-click behavior.
None of this is framed by Google as a departure from its stated mission to send valuable traffic to the web. But the practical effect, as measured independently by Ahrefs, Semrush, Seer, and Pew, is a search experience where fewer users need to leave the results page to get an answer, which is precisely the tension driving the current wave of publisher concern.
Competitive Comparison: How the Major SEO Platforms Are Responding
Semrush isn’t alone in racing to build measurement tools for a search landscape it doesn’t control. Ahrefs has focused its public research heavily on the click-through side of the equation, publishing the position-one CTR decline figures that have become some of the most-cited numbers in this entire debate, and has built its own AI Overview detection into its rank tracking. Seer Interactive has taken a more granular, campaign-level approach, breaking out CTR by whether a client’s page is cited, ranked-but-uncited, or absent from the AI Overview entirely, data that’s proven useful for agencies trying to set client expectations. Semrush’s differentiator is scale and surface breadth, the 126-million-prompt index spans four separate AI answer engines rather than just Google, and its AI Visibility Toolkit pairs that measurement with competitor gap analysis, letting a brand see not just its own citation rate but where a rival is being cited instead.
The practical upshot for marketing teams choosing between these platforms isn’t which one is “right,” it’s which measurement philosophy matches their reporting needs: raw CTR impact (Ahrefs), citation-versus-rank granularity (Seer), or cross-platform AI visibility scoring against named competitors (Semrush). Most enterprise SEO teams in 2026 are running more than one of these tools simultaneously, because none of them alone captures the full picture of how a brand shows up across Google, ChatGPT, Gemini, and Perplexity at once.
Small Publishers and Independent Sites Face the Steepest Climb
Large publishers and brands with dedicated SEO teams can absorb the cost of daily AI visibility monitoring and rapid content restructuring. Independent bloggers, niche affiliate sites, and small e-commerce operations generally can’t, and that gap is likely to widen the traffic divide that’s already reshaping the open web. A small site that loses 58% of its click-through rate on its highest-traffic informational page, in line with Ahrefs’ position-one figures, doesn’t have the same cushion of diversified traffic sources that a media company with a large email list or app audience can rely on. Semrush’s guidance is largely platform-agnostic and doesn’t require expensive tooling to start applying, restructuring a page so the direct answer appears in the first two sentences costs nothing but time, but the monitoring and competitive-gap analysis pieces of the recommendation do assume access to paid rank-tracking software, which puts independent publishers at a structural disadvantage in reacting quickly to citation losses.
Historical Context: From Ten Blue Links to Zero-Click Search
It’s worth remembering how recent this shift actually is. Google’s AI Overviews only rolled out broadly in the U.S. in mid-2024, following the “Search Generative Experience” testing phase that preceded it. In under two years, the feature has gone from an experimental label many users could still opt out of seeing, to a fixture appearing on roughly a third to nearly half of U.S. queries depending on whose measurement you trust, with no opt-out mechanism for site owners, according to Semrush’s own AI Overviews explainer. That pace of change is faster than any prior Google SERP feature, including the featured snippets and knowledge panels that reshaped SEO in the 2015-2018 period and that many practitioners assumed would represent the ceiling of “answer-first” search.
The featured-snippet era offers a useful, if imperfect, precedent. When Google began answering questions directly in a snippet box above organic results, publishers spent years learning to write for extraction, structured lists, defined answers up front, schema markup, specifically to win that box. AI Overviews demand a similar adaptation, but the stakes are higher because the AI Overview can synthesize from many sources simultaneously rather than lifting one snippet from one page, which is exactly why Semrush’s new guidance leans so heavily on citation tracking rather than single-position rank tracking.
Market Impact: What This Means for Publishers and Advertisers
The commercial stakes here are direct. Publishers that monetize through display advertising or affiliate commissions depend on click volume, and a 58% to 61% CTR decline on the queries most affected by AI Overviews, per Ahrefs and Seer Interactive, translates into a proportional hit to ad impressions and affiliate conversions for that traffic segment. Brands running paid search face a related distortion: if Google does expand advertising inside AI Overviews, as Semrush’s coverage suggests it’s testing, the AI summary box could become a new, separately priced inventory slot that sits above both organic results and traditional search ads, adding another line item to search marketing budgets that were already stretched by generative AI tooling costs.
There’s also a data quality wrinkle worth flagging: because AI Overview prevalence is measured differently by every vendor (12.95% by Semrush Sensor versus 48% by the Ahrefs-derived figure for the same U.S. market), marketing teams reporting to leadership need to pick one benchmark and stay consistent with it quarter over quarter, rather than mixing sources and creating the appearance of volatility that’s really just a measurement artifact.
Why Vendor Numbers Don’t Match: A Measurement Problem
Anyone comparing Semrush’s 12.95% U.S. AI Overview prevalence figure against the Ahrefs-derived 48% figure for the same market and roughly the same period will notice they’re describing very different worlds. Neither number is wrong on its own terms, they come from different query panels, different crawl frequencies, and different definitions of what counts as an “AI Overview appearance,” but the gap illustrates a real problem for anyone trying to benchmark performance. Without an industry-standard measurement methodology, similar to how Comscore and Nielsen eventually converged on shared definitions for web traffic measurement, brands risk making budget decisions based on whichever vendor’s number happens to look most alarming or most reassuring, a debate that continues to play out across SEO trade coverage week to week. Until that standardization happens, the safest practice is to pick a single measurement source, disclose it in reporting, and track relative change over time rather than treating the absolute percentage as gospel.
What Comes Next: Five Predictions for AI Search Visibility
First, citation tracking will become a standard line item in SEO reporting by the end of 2026, sitting alongside keyword rank and backlink counts as a core KPI, following the direction Semrush, Ahrefs, and Seer have all already moved. Second, expect AI Overview prevalence to keep climbing toward the 50%-plus range in the U.S. market given the trajectory from 34.5% in December 2025 to 48% in March 2026 reported by Ahrefs, with Google likely extending AI Mode’s footprint in parallel. Third, if Google’s tested advertising placements inside AI Overviews go live broadly, expect a new pricing tier in search advertising within the next one to two quarters, reshaping budget allocation away from traditional position-one bidding. Fourth, cross-platform visibility tools that track ChatGPT, Gemini, AI Mode, and Perplexity together, in the mold of Semrush’s 126-million-prompt index, will keep gaining adoption faster than single-platform rank trackers, since no serious brand can afford to optimize for Google alone anymore. Fifth, expect continued volatility in navigational-query AI Overviews, the category that grew from 0.74% to 10.33% of Semrush’s tracked queries in 2025, since brand-name searches getting summarized rather than linked directly is likely to draw further scrutiny from publishers and possibly regulators concerned about disintermediation.
How Marketing Teams Should Actually Respond Right Now
Practically, the guidance boils down to a short list any team can start on this week: audit which of your existing top pages already get cited in AI Overviews using an AI Overview filter in your rank tracker of choice, restructure your highest-traffic informational pages so the direct answer appears in the first two sentences rather than buried after a long introduction, verify your technical SEO isn’t accidentally blocking AI crawlers the way it might block scrapers, and start tracking citation rate as its own metric separate from organic rank. None of this requires abandoning traditional SEO fundamentals, backlinks, page speed, and relevance still matter, but none of it works in isolation anymore either. The sites treating AI Overview citation as a bolt-on afterthought are the ones most exposed to the CTR declines Ahrefs and Seer have already documented.
Frequently Asked Questions
What is Semrush’s 2026 AI Visibility Index?
It’s a research report Semrush published on June 26, 2026, analyzing 126 million U.S. AI search prompts collected between January and April 2026 across ChatGPT, Gemini, Google AI Mode, and Google AI Overviews, expanded from an earlier 2,500-prompt pilot study.
How often do Google AI Overviews actually appear in search results?
Estimates vary by measurement method. Semrush Sensor puts the U.S. average around 12.95% of queries, while an Ahrefs-derived dataset put U.S. prevalence at 48% of queries as of March 2026, up from 34.5% in December 2025. Globally, one 2026 industry recap of Semrush data put the figure around 13% of all queries.
How much does an AI Overview reduce click-through rates?
Ahrefs found a 58% lower average CTR for the position-one result when an AI Overview is present, as of December 2025. Seer Interactive data reported by Search Engine Land found a 61% decline in CTR on informational queries specifically, from 1.76% down to 0.61%.
Can website owners opt out of Google AI Overviews?
According to Semrush’s AI Overviews explainer, there is currently no dedicated way to opt a site out of having its content summarized in AI Overviews specifically, separate from broader crawler-blocking controls.
What are Semrush’s main recommendations for appearing in AI Overviews?
Semrush’s guidance centers on structuring content for easy extraction and summarization, strengthening entity and topic coverage rather than chasing single keywords, actively tracking and improving citation frequency, monitoring AI visibility daily rather than monthly, keeping pages technically accessible to AI crawlers, and tracking presence across ChatGPT, Gemini, AI Mode, and Perplexity in addition to classic Google Search.
Is Google testing ads inside AI Overviews?
Yes, according to Semrush’s own 2026 coverage, Google was testing advertising placements within AI Overviews in early 2026, though the company has not announced a broad rollout timeline.
What’s the difference between Google AI Overviews and Google AI Mode?
AI Overviews are AI-generated summary boxes that appear above traditional results within classic Google Search. AI Mode is a separate, more conversational, multi-turn search experience built on Gemini that runs alongside classic Search. Industry data from 2026 suggests AI Mode has an even higher zero-click rate, around 93%, than standard AI Overviews.
Does ranking number one in Google still matter if AI Overviews are shown?
It still matters, but it’s no longer sufficient on its own. Seer Interactive’s data shows a page that ranks well but isn’t cited inside the AI Overview can see CTR as low as 0.9%, compared to roughly 2.1% for a page that is cited, and around 3.3% when no AI Overview appears at all. Traditional ranking remains a prerequisite for eligibility, but citation inside the AI Overview has become the more decisive factor for actual traffic.
