AI Search & GEO Statistics for B2B SaaS 2026: Primary Research and Benchmarks

This reference brings together primary research published in 2026 on AI-assisted B2B buying, Google AI Overviews, organic clicks, AI referral traffic, brand mentions, citations, and common GEO tactics. It is aimed at B2B SaaS content teams trying to decide what to measure and what to improve.

The studies do not all measure the same population. Forrester and Gartner surveyed business buyers; Pew surveyed U.S. adults; Semrush, Ahrefs and Similarweb examined search results, traffic and AI answers. I have kept those distinctions explicit rather than labelling every number a “B2B SaaS benchmark.”

AI search and GEO statistics for 2026 at a glance

• 94% of business buyers reported using AI in their buying process (Forrester, Jan 2026).
• 45% of surveyed B2B buyers used generative AI during a recent purchase (Gartner, May 2026).
• 69% preferred to validate AI-generated insights with sales reps (Gartner, May 2026).
• 60% of U.S. adults said they ever read AI summaries in search results (Pew Research Center, Jun 2026).
• +71% growth in commercial-intent search results showing AI Overviews, in Semrush’s tracked sample (Semrush, Jul 2026).
• −58% estimated difference in position-one click-through rate associated with AI Overviews (Ahrefs, Feb 2026).
• +66% growth in AI referral traffic during 2025 across sampled websites (Semrush, Apr 2026).
• <0.15% share of total web visits attributable to AI referrals in Semrush’s sample (Semrush, Apr 2026).
• 770.7M average monthly AI referral visits worldwide in Similarweb’s measurement period (Similarweb, 2026).
• 37.9% of AI Overview-cited URLs also appeared in the first 10 SERP blocks for the same query (Ahrefs, Mar 2026).
• 61.7% of domain appearances were citations without a brand mention (Semrush, Jun 2026).
• 97% of valid llms.txt files received zero requests during May 2026 (Ahrefs, Jun 2026).

2026 AI search statistics: master reference table

Publication year refers to the research release, while the scope column records when the underlying behavior or traffic was measured. Some reports published in 2026 analyze 2025 data.

AI has become part of the B2B buying process

Buyer surveys offer the closest evidence for B2B SaaS, although neither study is restricted to software purchases. The useful question is where AI enters evaluation, and which steps buyers still want to validate elsewhere.

Forrester’s State of Business Buying 2026 reported that 94% of business buyers used AI during the buying process. The study drew on nearly 18,000 buyers surveyed in 2025. Forrester also found that buyers turn to peers, experts and analysts to check AI output. This is evidence of reported usage, not a measure of purchases caused by AI.

Gartner’s May 2026 survey of 645 B2B buyers, conducted in August and September 2025, reported a narrower measure: 45% used generative AI in a recent purchase, chiefly to research vendors and products. Respondents used an average of seven information sources, and 69% preferred to validate AI-generated insights with a sales representative. In the same study, 70% preferred a completely digital, self-service buying experience.

These figures are not contradictory. “AI at any point in the buying process” and “generative AI in a recent purchase” are different questions from different samples. For a SaaS team, they support auditing the product information buyers can independently verify: positioning, use cases, integrations, limitations, security details, pricing explanations, demonstrations and customer evidence.

Google AI summaries are familiar to many searchers

Consumer research helps explain the broader discovery environment, but it is not a direct measure of B2B purchasing. Pew Research Center surveyed 5,119 U.S. adults from February 17 to 23, 2026 and found that 60% said they ever read AI summaries at the top of search results. The share was 72% among adults ages 18–29 and 38% among those ages 65 and older.

“Ever read” describes reported exposure, not how often someone reads summaries, whether they trust the answers, or whether they click a cited result. Those are separate behaviors and need separate measures.

AI Overviews are appearing more often on commercial searches

Commercial searches matter because they include the comparison and evaluation work common to software buying. Semrush analyzed more than 600,000 U.S. desktop keywords across 10 industries between November 2025 and April 2026. In that sample, the share of commercial-intent results showing an AI Overview grew 71%, while the equivalent measure for transactional searches fell 5%.

The study covers several industries, not a B2B SaaS keyword set. It shows where AI Overviews are expanding in the sampled search results; it cannot tell us that 71% more SaaS comparison searches displayed AI Overviews. Teams should check their own high-intent query groups rather than assume the aggregate applies uniformly.

How commercial search growth differed by industry

Semrush’s commercial-intent dataset spans 10 industries. Its 71% average growth measure masks considerable variation across the categories surveyed. The examples below are changes in AI Overview appearances for commercial-intent queries between November 2025 and April 2026, not shares of all searches and not SaaS-specific estimates.

In the same research, transactional-intent AI Overview appearances fell 5% on average. Semrush also found a difference in average cost per click for some categories: finance keywords with AI Overviews averaged $4.84, compared with $2.14 for finance keywords without them. These comparisons describe the sampled keywords, not paid-media costs for an individual SaaS business.

AI Overviews can change organic click-through rates

Ahrefs’ February 2026 CTR update used 300,000 keywords: 150,000 with AI Overviews and 150,000 informational keywords without them. Comparing aggregated Google Search Console click-through rates from December 2023 and December 2025, it estimated that the presence of an AI Overview corresponded to a 58% lower average click-through rate for the top-ranking page.

This is a modeled comparison, not a guarantee that an individual page will lose 58% of its clicks. But it makes a useful distinction for reporting. Ranking position, search impressions, organic visits and appearance inside an AI-generated answer are no longer interchangeable indicators of visibility.

For pages that once relied on a top organic position, compare impressions and clicks over time, segment queries with AI Overviews, and review changes in demand and intent. A lower click-through rate might have several causes; the presence of an AI Overview is one testable factor.

Click-through differences by ranking position

Ahrefs measured position-specific click-through-rate differences for two sets of 150,000 keywords, comparing December 2023 and December 2025. The estimated reductions were not identical for every ranking position.

The percentages are study-level estimates, not observed before-and-after losses for a specific SaaS URL. The comparison is based on desktop click-through data and cannot isolate every change to search behavior across the two years.

AI referral traffic is growing from a small base

The phrase “AI traffic is exploding” is easy to misread without a denominator. Semrush’s April 2026 traffic channel study analyzed billions of visits across more than 50,000 sites in 17 industries during 2025. AI referrals grew 66%, but accounted for less than 0.15% of total visits in that dataset.

Similarweb’s 2026 AI referral analysis estimated 770.7 million average monthly AI referral visits worldwide between June 2025 and May 2026, up 117.4% year over year. This covers desktop and mobile web and measures a different universe and time period from Semrush’s study.

Both can be true. One report emphasizes the channel share across a sample of sites; the other estimates global volume and growth. They should not be combined into a single funnel, or used to project a particular SaaS company’s traffic without site-level evidence.

AI referrals vary by industry and interface

Semrush’s traffic-channel study includes a useful constraint on the overall 66% growth figure. Its computer software and development category recorded a 26% decline in AI referral traffic during 2025, while retail grew 343% and apparel 319%. These are sector-level changes in Semrush’s dataset. Computer software and development is not a statistically representative sample of B2B SaaS vendors.

The same study reports Google AI Mode growing from 1,600 visits in January 2025 to 38.2 million visits in December 2025 in its tracked data. Semrush counts AI Mode separately from conversational-AI referral traffic. Combining those series would change the channel definition and misstate the comparison.

A citation and a brand mention are different outcomes

Ahrefs examined 863,000 keyword SERPs and four million AI Overview URLs and found that 37.9% of cited URLs also appeared within the first ten search-result blocks for the same query. That count included regular results and SERP features, so it is not equivalent to saying 37.9% ranked among the first ten ordinary organic blue links.

Meanwhile, Semrush’s ghost citations study classified 3,981 domain appearances across 115 prompts in 14 countries and four AI engines. It found that 61.7% were citations without an accompanying brand mention; 13.2% combined a citation with a mention, and 25.1% were mentions without a citation.

For GEO reporting, I would avoid collapsing these outcomes into one “AI visibility” score. Track at least four separate things: whether the brand is named, whether a page is cited, which relevant queries trigger those appearances, and whether users actually visit or convert.

What the citation and mention categories actually count

Semrush and Kevin Indig’s ghost-citations study counted 3,981 domain appearances across 115 prompts, 14 countries and four AI search interfaces. The three outcomes describe distinct classifications for a domain appearance, not conversion rates or user behavior.

A separate Ahrefs analysis distinguished ordinary organic links from all search-result blocks. Among AI Overview-cited URLs, 37.1% appeared among the first ten standard organic results, 26.2% ranked 11 to 100, and 36.7% did not rank in the top 100 ordinary organic results. These numbers come from a different query sample and should not be compared directly with Semrush’s mention categories.

What the data says about GEO tactics

Structured data is useful, but it is not a proven citation shortcut

Ahrefs tracked 1,885 pages that added JSON-LD structured data and compared them with 4,000 control pages across Google AI Overviews, AI Mode and ChatGPT. The matched analysis found no meaningful citation uplift for AI Mode or ChatGPT; the AI Overview estimate was a small decline relative to controls. This does not make structured data useless for ordinary SEO. It means the study did not show that adding schema alone causes more AI citations.

In Ahrefs’ matched schema experiment, pages adding structured data recorded an estimated −4.6% relative citation change in Google AI Overviews, +2.4% in Google AI Mode and +2.2% in ChatGPT. The last two estimates were statistically indistinguishable from zero. Ahrefs cautioned against treating the first estimate as evidence that schema itself harms citation performance.

There is little evidence that llms.txt is a priority

Ahrefs examined 137,000 domains and found that 28% published an llms.txt file. Among the valid files, 97% received no requests in May 2026. The result describes server requests in the sampled data, not every possible AI system, but it is a strong reason to avoid treating llms.txt as a guaranteed GEO lever.

The more useful work is often less novel: make product information clear and crawlable, maintain factual source pages, publish original evidence, update outdated claims, and assess what buyers actually need to validate. See my earlier breakdown of SEO versus GEO for B2B SaaS for the practical workflow.

How I’d use these statistics in a B2B SaaS content program

1. Separate buying research from traffic research. Buyer surveys help explain discovery and validation; analytics reveal what users did on your website.
2. Audit your evaluation-stage search queries. Check whether AI Overviews appear for comparisons, alternatives, integration questions, pricing and product use cases.
3. Measure citations and mentions separately. A source link that never names your brand is a different outcome from an explicit recommendation.
4. Keep publication and data dates together. A report published this year can describe responses collected last year.
5. Compare conversions, not just referrals. Growth from a small base can look spectacular while remaining commercially minor.
6. Invest in evidence you control. Product documentation, transparent methodology, screenshots, case studies and original research create material buyers can independently check.

If you are producing statistical reference pages, I have also documented how I verify statistics before citing them, including my rules for primary sources, measurement periods, and misleading comparisons.

Methodology and limitations

This edition includes findings published in 2026 by the organizations that conducted the underlying surveys, traffic analyses or search experiments. Some data were measured in 2025; those measurement windows are identified rather than relabeled as 2026 responses. The studies use different populations, sampling methods, geographies and outcome definitions. They are presented side by side for reference, not pooled into one statistical estimate. I excluded estimates from secondary roundups and claims I could not verify against the producing organization.

Primary sources

• Forrester: The State of Business Buying 2026
• Gartner: B2B Buyers Validate AI-Generated Insights
• Pew Research Center: Americans and AI 2026
• Semrush: AI Overviews and Commercial Search
• Ahrefs: AI Overviews Reduce Clicks by 58%
• Semrush: Traffic Channel Mix Study
• Similarweb: 2026 AI Referral Traffic by Industry
• Ahrefs: AI Overview Citations and Top-10 Results
• Semrush: Ghost Citations Study
• Ahrefs: Schema and AI Citations
• Ahrefs: llms.txt Study