Semrush published a 12-month study on October 21, 2026 tracking 400 US commercial keywords. Google showed an AI Overview for 88% of them in August 2026, up from 45.75% in September 2025, and for 97.08% when the keywords were weighted by search volume.

In Short

Google now puts an AI-written summary above the normal search results for almost all of the shopping-style searches Semrush watched, including "X vs Y", "best X for Y" and "X review" searches. That matters because those are the searches people run just before they buy, and the summary often quotes other websites, videos and review sites rather than the brand's own page. So the first thing a shopper reads about a product is increasingly a Google summary built from many sources, not the brand's own comparison page.

What Semrush measured

Semrush, which now describes itself as "an Adobe company", published the research on its blog on October 21, 2026 under the title "AI Overviews have nearly doubled across five commercial search modifiers". The study is written by Luke Harsel, who leads content for Semrush's product-led growth team, with Nikola Manojlović listed as contributor. A pitch email sent to PPC Land on October 21 and a follow-up on October 22 pointed to the same study; the follow-up said the study "has just gone live", while the page itself carries the October 21 byline.

The design is narrow and fixed. Semrush built a panel of 400 keywords split across three industries - banking, software and ecommerce - and five modifier groups that searchers attach to a query when they are close to a decision. According to the study, those groups are:

  • Alternatives, such as "sofi alternatives" or "crowdstrike alternatives"
  • [Brand] vs. [brand], such as "keurig vs nespresso" or "hubspot vs salesforce"
  • Best X for Y, such as "best credit union for car loan" or "best smartwatch for women"
  • Pricing and cost, such as "coffee maker under 100" or "price antivirus software"
  • Reviews, such as "amazfit smartwatch review" or "chase checking account review"

Using Semrush US desktop data, the company checked every month from September 2025 through August 2026 whether each keyword triggered an AI Overview. The main metric is what Semrush calls keyword-weighted coverage: "the share of keywords in each fixed group that triggered an AI Overview." A second metric, volume-weighted coverage, weights each keyword by its estimated monthly search volume.

Two limits follow directly from that description. The data covers US desktop results only, so mobile results and other countries sit outside the sample. And the study records whether an AI Overview appeared, not what happened next - it contains no click, traffic or conversion data. The study does not publish how many of the 400 keywords sit in each modifier group or each industry.

Coverage nearly doubled

Across all 400 keywords, AI Overview coverage reached 88% in August 2026, up from 45.75% in September 2025, according to Semrush. That is the basis for the "nearly doubled" headline: by PPC Land's calculation, the August figure is 1.92 times the September one.

Every modifier group ended the period above 80%. In September 2025 the spread ran from 30.14% for reviews to 63.41% for brand-versus-brand queries. By August 2026 it had narrowed to between 82.05% and 96.12%. Semrush's own table gives the full picture:

Modifier September 2025 August 2026 Change
Alternatives 48.44% 82.81% +34.37 points
Best X for Y 42.72% 96.12% +53.40 points
Price/cost 43.59% 82.05% +38.46 points
Reviews 30.14% 89.04% +58.90 points
Brand vs brand 63.41% 86.59% +23.18 points

Review queries moved the furthest. "A year ago, the searches we tracked that included 'reviews' saw an AI Overview just 30% of the time," the study states. "Now, AI Overviews appear in those search results nearly 90% of the time." The table puts the end figure at 89.04%, and the study's key takeaways round it to "up to 90%". Reviews went from the least likely modifier to trigger an AI Overview to the second most likely, behind best-X-for-Y queries at 96.12%.

Brand comparison queries tell a different story. They started highest, at 63.41%, and grew least, by 23.18 points. That 23-point figure, and the 86% end point, were the lead in the pitch email sent to PPC Land, which described Google AI as appearing in "86% of searches". The underlying number is narrower: 86.59% of the tracked brand-versus-brand keywords in the panel, on US desktop.

One labeling slip in the study is worth flagging. Its key takeaways describe the alternatives figure as "appearing on more than 82% of brand comparison SERPs", although alternatives and brand-versus-brand are separate groups in the methodology and in the table. The 82.81% belongs to alternatives keywords.

Industries converged

The gap between verticals has closed even faster than the gap between modifiers. In September 2025, coverage ranged from 32.62% in ecommerce to 68% in software, a spread of more than 35 points. By August 2026 all three industries sat above 85%, with less than six points between highest and lowest, according to Semrush. The study's chart shows software at 91%, banking at 89% and ecommerce at 85%.

Ecommerce posted the largest increase of the three, from 32.62% to 85.11%. "This suggests that whether a searcher's comparing options for a coffee maker, a CRM solution, or a credit card, they're just as likely to see an AI Overview that consolidates and summarizes the commercial information for them," the study states.

Fifteen cells, three at 100%

Crossing three industries with five modifiers gives 15 combinations. By August 2026, 12 of them had keyword-weighted coverage of at least 80%. Three reached 100%: best-X-for-Y in banking, alternatives in software and best-X-for-Y in software. Three stayed below 80%: alternatives in banking at 65%, price/cost in ecommerce at 71.43% and reviews in software at 75%.

The remaining August 2026 values in Semrush's heat map are: banking brand-versus-brand 89.3%, banking price/cost 87.2%, banking reviews 89.7%, ecommerce alternatives 84.6%, ecommerce best-X-for-Y 88.9%, ecommerce brand-versus-brand 80%, ecommerce reviews 96.4%, software brand-versus-brand 91.7% and software price/cost 83.3%.

All 15 combinations finished higher than they started. Ecommerce price/cost keywords rose from 4.76% to 71.43%, and banking review keywords from 13.79% to 89.66%.

Those percentages also say something about cell size. With 400 keywords spread across 15 cells, the average cell holds fewer than 27 keywords. By PPC Land's calculation, 4.76% and 71.43% are consistent with 1 and 15 out of 21 keywords, and 13.79% and 89.66% with 4 and 26 out of 29. At that size, a single keyword moves a cell by three to five points. The study itself makes the same caveat implicitly when it describes the 100% cells: its sentence that "every searcher saw an AI Overview" in those cells reads more into the data than a keyword count can carry, since the metric counts keywords, not searchers.

Volume weighting pushes the figure to 97%

The second metric produces a higher number. Weighted by estimated monthly search volume, AI Overviews appeared on 97.08% of commercial searches in the panel in August 2026, against 88% when every keyword counted equally. According to Semrush, the 12% of keywords that did not trigger an AI Overview accounted for less than 3% of total search volume.

The pattern held in every modifier group. Semrush's chart gives these August 2026 pairs, keyword-weighted first and volume-weighted second:

  • Alternatives: 83% and 86%
  • Best X for Y: 96% and 98%
  • Brand vs brand: 87% and 99%
  • Price/cost: 82% and 96%
  • Reviews: 89% and 94%

"Keyword-weighted coverage shows how widespread AI Overviews are. Volume-weighted coverage shows how much search demand they reach," the study explains. Its reading is that "the highest-demand commercial searches in the dataset were more likely to surface an AI Overview." The volume figures rest on Semrush's own search volume estimates, which are modeled rather than measured counts of searches.

Who gets cited: the Mailchimp and Klaviyo example

The pitch email framed the study around a single example: control of the comparison narrative. For the query "mailchimp vs klaviyo", the study shows a screenshot of the AI Overview. According to Semrush, it "cites Klaviyo's own guide for the option 'Choose Klaviyo if...' However, it cites a third-party website (Zapier) when it says 'Choose Mailchimp if you run a general small business...'." The study adds that "the other prominent citations are a YouTube video and another third party site."

The screenshot itself shows slightly more than the caption. The opening sentence of the summary is attributed to Mailercloud. The Klaviyo bullet carries a "Klaviyo +2" source chip and the Mailchimp bullet a "Zapier +1" chip, so each recommendation draws on more than one source. The side panel lists a Klaviyo page titled "Klaviyo vs Mailchimp (2026): Compare Email Marketing Platforms" and a YouTube video from a channel labeled Consumer Research Studios. The first organic result below the summary is an EmailToolTester article dated September 21, 2026. The study does not say when the screenshot was captured.

The asymmetry is the point Semrush draws from it. "The answer people see when searching '[your brand] vs. [competitor]' can draw on YouTube videos, reddit threads, and review sites, as well as your own site and your competitor's," the study states. One brand's own page made the case for that brand; a third party made the case for its rival. A single screenshot cannot establish how common that pattern is, and the study offers no citation-share figures across the 400 keywords.

What Semrush says it means

Semrush's commentary pushes in one direction. "Ranking for keywords is still a critical goal. But now that AI Overviews makes the first impression, you also need to target and track that answer," the study states. It also says that "ranking number one likely won't lead to as much traffic as it did this time last year", although no traffic data from the panel supports that sentence.

The study closes with five recommendations: tracking AI Overview visibility alongside rankings, mapping which sources the summaries cite, keeping comparison and pricing pages current, working with YouTube creators, and earning reviews on the platforms the summaries cite. It ends with an invitation to a free seven-day trial of Semrush's tracking tools. That commercial interest is relevant context: Semrush sells the visibility tracking its own conclusions call for.

Why this matters for marketers

The finding lands on a question PPC Land has followed for more than a year: what happens to clicks when an AI summary sits on top of the SERP. Pew Research Center found in July 2025 that users clicked a link 8% of the time when an AI summary appeared against 15% without one, at a point when about 18% of Google searches in its sample produced a summary; Google called that methodology "flawed". Seer Interactive reported in November 2025 that organic click-through rates on queries with AI Overviews fell 61%, and paid click-through rates 68%, across 3,119 informational queries. Ahrefs put the reduction for top-ranked pages at 58% in February 2026, using 300,000 keywords.

Those studies largely measured informational queries. Semrush's panel moves the question to the commercial end of the funnel, where purchase intent is highest and where comparison and alternatives pages have long been built to capture buyers. Semrush's own earlier study, published on July 2, 2026 and linked from the new one, had already reported that the share of commercial-intent results with an AI Overview grew 71% on average between November 2025 and April 2026, across more than 600,000 keywords. The two studies use different panels and periods, so their figures are not directly comparable, but they point the same way.

Coverage rates also vary sharply by market and sample. SISTRIX put AI Overviews on about 20% of German keywords in March 2026, with an estimated 265 million organic clicks lost per month, across more than 100 million keywords. A 400-keyword panel built from decision-stage queries in three verticals is a very different object from a national keyword index, and an 88% figure from the former says little about the share of all Google searches.

The citation question matters as much as the coverage rate. Recrawled's study of UK B2B software firms, covered in October 2026, found that the share of companies named by at least one AI platform fell to 0% for leader-versus-rival comparison questions, and that YouTube appeared in 25% of Google AI Overview answers. Semrush's June 2026 AI visibility report counted an average of 9.2 citations per AI Overview response across 126 million prompts. Each of those citations is a slot that a brand's own page competes for against review sites, creators and competitors.

There is also an advertising dimension. Commercial queries are where search ads concentrate, and Google extended ads in AI Overviews to 11 more English-language markets in December 2025. Adthena's September 2026 data showed one-to-two-word queries taking the majority of AI Overview ad appearances in every region it measured. Semrush's study does not look at ads at all, so it cannot say whether the growth in summaries on commercial keywords changed paid placements or costs.

Finally, the source itself has changed hands. Adobe agreed to acquire Semrush for $1.9 billion in November 2025, with closing expected in the first half of 2026; the study's charts and copy now carry the "An Adobe Company" label. Research from a vendor of AI visibility tools that finds AI visibility matters more every month is not disqualified by that interest, but readers can weigh it.

What the study does not establish

Several questions remain outside the data. The study does not measure clicks, so it cannot show that the growth in AI Overviews reduced traffic to any of the pages that rank for these keywords. It does not quantify who gets cited across the panel, so the Mailchimp and Klaviyo case remains a single illustration. It covers US desktop only. It does not disclose cell sizes, though the reported percentages imply cells of roughly 20 to 30 keywords. And it gives no explanation for why Google expanded summaries across these query types between September 2025 and August 2026 - Google is not quoted in the study, and the study does not say whether it asked.

Timeline

Summary

Who: Semrush, an Adobe company, through a study by Luke Harsel with contributor Nikola Manojlović. The findings concern Google's AI Overviews and the brands, publishers and review sites that compete for citations in them.

What: Across a fixed panel of 400 US commercial keywords, AI Overview coverage rose from 45.75% to 88%, and to 97.08% when weighted by search volume. Review keywords rose from 30.14% to 89.04%, brand-versus-brand keywords from 63.41% to 86.59%, and three of 15 industry-and-modifier cells reached 100%.

When: The tracking ran from September 2025 to August 2026. Semrush published the study on October 21, 2026.

Where: Google Search results in the United States, desktop only, for banking, software and ecommerce queries.

Why: These decision-stage searches are where comparison, pricing and review pages have traditionally captured buyers. With an AI summary now above the organic results on almost all of them, and that summary drawing on third-party sources such as Zapier and YouTube, a brand's own page is one input among several in the first answer a buyer reads.