Semrush found that AI Overviews now appear on 88 percent of a fixed set of 400 commercial keywords, nearly double the 45.75 percent recorded in September 2025. The study, by Luke Harsel and Nikola Manojlović, was covered on Oct. 22 by Search Engine Land. For teams that build comparison, pricing, and “best of” pages, it means the buying-stage SERP now opens with a generated summary rather than a ranked list.
The panel spanned banking, ecommerce, and software terms, each paired with one of five modifiers: alternatives, brand versus brand, best X for Y, price or cost, and reviews. Semrush checked U.S. desktop data monthly through August 2026. Every modifier finished between 82.05 and 96.12 percent coverage.
Reviews moved the most. Queries containing “review” triggered an AI Overview 30.14 percent of the time in September 2025 and roughly 90 percent in August 2026, a gain Semrush put at 59 points. “Best X for Y” queries ended highest at 96.12 percent, up 53.4 points. Brand versus brand queries started highest, at 63.41 percent, and grew least, up 23.18 points.
Industries converged. In September 2025, coverage ran from 32.62 percent in ecommerce to 68 percent in software, a spread above 35 points. By August 2026 all three sat between 85 and 91 percent, with under six points separating them. Ecommerce posted the largest industry gain, ending at 85.11 percent.
Three industry-and-modifier pairs reached 100 percent: best X for Y in banking, alternatives in software, and best X for Y in software. Three stayed under 80 percent: banking alternatives at 65 percent, ecommerce price or cost at 71.43 percent, and software reviews at 75 percent. All 15 combinations finished above where they began. Ecommerce price keywords climbed from 4.76 percent to 71.43 percent.
Volume weighting widens the picture. Weighted by estimated monthly search volume, coverage reached 97.08 percent in August 2026. The 12 percent of keywords with no AI Overview accounted for under 3 percent of volume. Semrush reads this as keyword counts understating exposure, because the highest-demand queries are the likeliest to show a summary.
Semrush’s own example, a “mailchimp vs klaviyo” search, shows the sourcing problem. The AI Overview cited Klaviyo’s guide for the Klaviyo recommendation, a third-party site (Zapier) for the Mailchimp one, plus a YouTube video and another outside site. A brand’s comparison page is one input among several, and rivals and reviewers write parts of the answer. Semrush recommends tracking AI Overview mentions and citations beside rankings, mapping cited sources, refreshing comparison and pricing pages, and earning reviews on the platforms that get cited.
Treat the findings with care. This is a vendor study, and Semrush sells AI Overview tracking, which it promotes at the end of the post. The sample is 400 keywords from three industries, U.S. desktop only, in a fixed panel Semrush chose, so it does not describe all commercial search or mobile results. The study also measures how often an AI Overview appears, not how many clicks organic listings lost. Semrush says a top ranking likely earns less traffic than a year ago but publishes no click figures.
The stake is reporting. A page can hold position one and still lose the first impression to a summary that quotes a competitor or a reviewer. Teams should pull their top 50 commercial queries, record whether an AI Overview appears and which sources it cites, and compare the share against Semrush’s 88 percent before planning next quarter’s comparison-page work.
Semrush blog study by Luke Harsel and Nikola Manojlović, also covered by Search Engine Land on Oct. 22, 2026.
