Request a callAI Overviews now appear on 88% of tracked commercial searches: what B2B teams should do
A Semrush study published on October 21, 2026 found that Google AI Overviews appeared on 88% of 400 tracked US commercial keywords in August 2026, up from 45.75% in September 2025. Weighted by search volume, coverage was 97.08%. For B2B teams, comparison, alternatives, pricing and review searches are now often answered by an AI summary before anyone reaches a vendor site, so those pages need to be measured by citations and brand mention, not rankings alone.
Key takeaways
- Semrush found AI Overviews on 88% of its tracked commercial keywords in August 2026, up from 45.75% in September 2025.
- The panel is 400 US desktop keywords across software, banking and ecommerce, so treat it as direction, not a Canadian or mobile benchmark.
- "Best X for Y" searches were highest at 96.12%, and software alternatives and software "best X for Y" cells reached 100%.
- Audit about 20 commercial queries per client and record the summary, the cited sources, brand inclusion and factual errors.
What Semrush measured
Semrush tracked a fixed panel of 400 commercial keywords every month from September 2025 to August 2026, using US desktop Google results. The keywords sit in three industries (banking, ecommerce and software) and five modifier groups: alternatives, "best X for Y", price and cost, reviews, and brand versus brand.
The headline number is keyword-weighted: the share of keywords that triggered an AI Overview. A second view weights each keyword by its estimated search volume, which is a modeled estimate rather than a measured count.
The numbers that matter for B2B
| Search type | September 2025 | August 2026 |
|---|---|---|
| All tracked keywords | 45.75% | 88% |
| Alternatives | 48.44% | 82.81% |
| Best X for Y | 42.72% | 96.12% |
| Price and cost | 43.59% | 82.05% |
| Reviews | 30.14% | 89.04% |
| Brand versus brand | 63.41% | 86.59% |
Weighted by estimated search volume, coverage reached 97.08% in August 2026. In the software segment, the alternatives and "best X for Y" groups both reached 100%.
What this study does not show
- It records whether an AI Overview appeared. It has no click, traffic or conversion data, so it cannot tell you how much traffic was lost.
- It covers US desktop only. Canadian, European and mobile results can differ.
- Groups are small. Each industry and modifier cell holds roughly 20 to 30 keywords by the arithmetic in the PPC Land coverage, so one keyword can move a cell by several points.
- It does not report citation share, and Semrush sells visibility tracking, so read the recommendations with that in mind.
Why comparison pages matter more now
Buyers use these query types late in a purchase: comparing vendors, looking for alternatives and checking reviews. If an AI summary frames that decision, the question for a vendor becomes whether its facts appear in the summary and whether its own pages are among the sources cited.
That makes pages that state clear, checkable facts more valuable: honest comparison pages, pricing explanations, customer evidence you are cleared to publish, and third-party sources that Google already cites for your category.
Run a 20-query audit for each client
- Pick about 20 commercial queries that real buyers use, split across the five groups: brand versus competitor, competitor alternatives, best category for use case, brand pricing and brand reviews.
- For each query, record whether an AI Overview appears, how it describes your brand, which sources it cites, whether your brand is included and any factual errors.
- Fix errors at the source: update your own comparison and pricing pages first, then the third-party pages the summary relies on.
- Repeat monthly with the same query set so changes are comparable.
What to measure
- AI Overview presence on the query set.
- Brand mention share and owned-domain citation share.
- Factual accuracy of what the summary says about you.
- Branded organic click-through rate and qualified traffic from the affected queries, from Search Console and your CRM.
Frequently asked questions
Does this mean organic rankings no longer matter?
No. Strong organic pages remain the base for being cited. The change is that rankings alone no longer show whether a buyer saw you.
Is the 88% figure true for Canada or mobile?
Unknown. The study covers US desktop results only, so audit your own queries in your own markets.
Can we force our brand into an AI Overview?
No. You can make your pages clear, factual and easy to cite, and correct errors in the sources the summary draws from.
How many queries should the audit cover?
About 20 per client is enough to see patterns without turning the audit into a project of its own.
Sources
Related service: Generative engine optimization. See also pricing and client results.
