Last updated September 2026. Next refresh: December 2026. Every figure below names its source, its sample size where the publisher disclosed one, and its publication year.
Call it AI-assisted buying: the growing share of research, comparison, and vendor narrowing that now happens inside a chatbot instead of on a search results page or a vendor’s own website. It is not a niche behavior anymore. It shows up in enterprise software surveys, national polls of ordinary adults, car-dealership tracking data, and holiday retail traffic logs, all pointing the same direction.
This page merges five separate studies into one ledger: G2’s March 2026 survey of B2B software buyers, Forrester’s 2026 Buyer Insights research, Pew Research Center’s February 2026 poll of U.S. adults, Cox Automotive’s AI in Auto Retail Tracker, and Adobe’s retail and travel referral data. Each row below is a self-contained claim: a number, a named source, and a year, built to stand on its own.
How to read this stat bank
A few conventions to note before you cite a number:
- Every stat repeats its full source in the row, so you can quote any single line without losing the citation.
- Studies define “using AI” differently. Some count any generative AI touchpoint at any point in a purchase. Others count only chatbot sessions, or only the first step of research. Read the Metric column, not just the headline percentage, before you compare two rows.
- Statistics from single-vendor studies are marked (vendor). They are real data from that company’s platform, but have not been independently replicated by a third party.
B2B software and technology buyers
G2’s “Answer Economy” report surveyed 1,076 B2B software buyers and decision-makers in March 2026, across North America, EMEA, and APAC. It is the largest single dataset behind this page.
How B2B buyers use AI chatbots to research
| Stat | Metric | Source | Year |
|---|---|---|---|
| 51% of B2B software buyers now start research in an AI chatbot rather than Google | Where research starts | G2, “The Answer Economy” (1,076 buyers, March 2026) | 2026 |
| 71% rely on an AI chatbot at some point in their research, up from about 60% seven months earlier | Chatbot use anywhere in research | G2, same survey | 2026 |
| 93% say an AI chatbot has fundamentally changed how they research software | Perceived change in process | G2, same survey | 2026 |
| 80% still use Google somewhere in their buying journey, even after adopting a chatbot | Google still in the mix | G2, same survey | 2026 |
| 86% increased their AI chatbot use for software research over the past year | Year-over-year usage growth | G2, same survey | 2026 |
| 63% spend six or more hours a week using AI tools for work | Weekly usage intensity | G2, same survey | 2026 |
| 41% regularly use a Deep Research-style AI tool; only 2% are unaware such tools exist | Deep Research adoption | G2, same survey | 2026 |
| 33% open their first AI research session with a category question, 31% with a head-to-head competitor comparison, and 22% by asking about requirements | First-prompt intent | G2, same survey | 2026 |
| ChatGPT holds 62% of B2B software buyer chatbot usage in 2026, up from 47% in 2025; Gemini holds 14% | Chatbot share among buyers | G2, same survey | 2026 |
| ChatGPT’s share of buyer attention peaks at 73% at the discovery stage and stays above half through consideration, decision, and retention | Chatbot share by funnel stage | G2, same survey | 2026 |
What that usage does to vendor selection
| Stat | Metric | Source | Year |
|---|---|---|---|
| 53% say software research is more productive with AI, up from 36% in August 2025 | Perceived research productivity | G2, same survey | 2026 |
| 80% say an AI chatbot sped up their purchase decision | Decision speed | G2, same survey | 2026 |
| 83% felt more confident in their final vendor choice after using AI | Decision confidence | G2, same survey | 2026 |
| 69% chose a different vendor than they originally planned, based on AI chatbot guidance | Vendor switching | G2, same survey | 2026 |
| 33% bought from a vendor they had never heard of before, after AI chatbot guidance | New-brand discovery | G2, same survey | 2026 |
| 85% hold a more favorable view of a vendor once they see it cited inside an AI answer | Citation halo effect | G2, same survey | 2026 |
| 54% name a generative AI chatbot as the number one source shaping their vendor shortlist, ahead of review sites at 43% | Top shortlist source | G2, same survey | 2026 |
| 45% say a review-site citation is the single most confidence-inspiring element inside an AI answer | What builds trust in an answer | G2, same survey | 2026 |
| 64% say they hit an AI hallucination “often” or “very often” while researching software | Hallucination frequency | G2, same survey | 2026 |
| 40% self-identify as a daily AI “power user” | Power-user self-identification | G2, same survey | 2026 |
The buying committee: Forrester, Gartner, and Wynter
| Stat | Metric | Source | Year |
|---|---|---|---|
| 94% of B2B buyers used AI somewhere in their most recent purchase, up from 89% a year earlier | AI use in most recent purchase | Forrester, 2026 Buyer Insights: Buyer Usage of GenAI | 2026 |
| 61% of B2B buyers use a private, company-supplied AI tool instead of the free consumer version | Private vs. public AI tool use | Forrester, same research | 2026 |
| B2B buyers are twice as likely to use ChatGPT and four times as likely to use Microsoft Copilot at work as consumers are in daily life | Work vs. consumer chatbot use | Forrester, same research | 2026 |
| More than half of ChatGPT and Copilot users at work use their company’s private, licensed version rather than the free public tool | Enterprise vs. free-tier use | Forrester, same research | 2026 |
| Generative AI and conversational search are named the single most meaningful information source in B2B buying, chosen twice as often as any other channel | Most meaningful info source | Forrester, same research | 2026 |
| The typical B2B buying decision now involves 13 internal stakeholders and 9 external influencers | Buying group size | Forrester, “The State of Business Buying, 2026” | 2026 |
| More than 60% of business buyers now use a trial before they buy, rising to 78% among buyers spending $10 million or more | Trial-before-buy behavior | Forrester, “The State of Business Buying, 2026” | 2026 |
| 67% of B2B buyers say they would prefer a purchase process with no sales rep at all | Preference for rep-free buying | Gartner, sales survey of 646 B2B buyers, March 2026 | 2026 |
| By 2028, Gartner projects that 90% of B2B buying will be intermediated by AI agents, routing more than $15 trillion in spend | Future AI-agent-intermediated spend | Gartner, IT Symposium/Xpo press release, November 2025 | 2025 |
| 68% of B2B SaaS CMOs at $50 million-plus companies now start vendor discovery in AI tools before traditional search | CMO vendor research shift | Wynter, survey of 101 B2B SaaS CMOs, January 2026 | 2026 |
Consumer buyers
Pew Research Center surveyed 5,119 U.S. adults from February 17 to 23, 2026, as part of its American Trends Panel. This is the broadest, least commercially filtered dataset on this page: it covers the general public, not just people already shopping for software or a car.
| Stat | Metric | Source | Year |
|---|---|---|---|
| 49% of U.S. adults have used an AI chatbot, up from 33% in 2024 | Overall chatbot adoption | Pew Research Center (5,119 U.S. adults, Feb. 17-23, 2026) | 2026 |
| 25% of U.S. adults use an AI chatbot every day | Daily chatbot use | Pew Research Center, same survey | 2026 |
| 42% of chatbot users say they use it to search for information, the single most common use case | Top chatbot use case | Pew Research Center, same survey | 2026 |
| 38% of employed adults who use a chatbot use it for tasks related to their job | Chatbot use for work tasks | Pew Research Center, same survey | 2026 |
| 44% of U.S. adults use ChatGPT, the most-used chatbot; 24% use Gemini; 17% use Copilot | Chatbot brand share, general public | Pew Research Center, same survey | 2026 |
| Adults under 50 use AI chatbots at a 63% rate, versus 23% among adults 65 and older | Adoption by age | Pew Research Center, same survey | 2026 |
Vertical spotlights
Buyer behavior does not move at one speed. These three categories show how far AI-assisted research has already traveled in three very different purchases: a car, a hotel room or holiday gift, and a college.
Automotive
| Stat | Metric | Source | Year |
|---|---|---|---|
| 63% of in-market vehicle shoppers say they will definitely or probably use AI on their next vehicle purchase | Planned AI use, next purchase | Cox Automotive, AI in Auto Retail Tracker (1,500+ in-market shoppers, Q1-Q2 2026) | 2026 |
| 44% of car shoppers already used an AI tool during their 2025 vehicle-buying process | AI use, realized | Cars.com, survey of in-market shoppers, November 2025 (vendor, pre-filtered toward AI users) | 2025 |
| 24% of shoppers say AI makes them feel more prepared walking into a dealership | Buyer preparedness | Cox Automotive, same tracker | 2026 |
| Only 29% of dealers have started adapting to AI-powered vehicle search, even though 82% of dealers already use AI somewhere in their business | Dealer-side adoption gap | Cox Automotive, same tracker | 2026 |
| Dealers working with an external AI partner report far higher confidence in their AI output (30%) than dealers going it alone (8%) | Confidence in AI output | Cox Automotive, same tracker | 2026 |
Retail and travel
| Stat | Metric | Source | Year |
|---|---|---|---|
| AI-driven referrals to U.S. retail sites jumped 693% year over year during the 2025 holiday season | AI referral growth, retail | Adobe Analytics (1 trillion+ U.S. retail site visits tracked), January 2026 | 2026 |
| That surge peaked at 769% year over year in November 2025 alone | Peak monthly AI referral growth | Adobe Analytics, same dataset | 2026 |
| AI referrals to U.S. travel sites rose 3,500% year over year in July 2025 | AI referral growth, travel | Adobe Digital Insights (8 million+ visits tracked), September 2025 (single-vendor panel, not industry-wide) | 2025 |
Higher education
| Stat | Metric | Source | Year |
|---|---|---|---|
| 46% of high school students used AI in their college search, up from 26% in spring 2025 | AI use in college search | EAB, survey of 5,000+ high school students, Oct.-Nov. 2025 (vendor) | 2025 |
| 18% of students removed a college from consideration because of something an AI tool surfaced | AI influence on decisions | EAB, same survey (vendor) | 2025 |
What the numbers mean
AI research now happens before a seller knows a buyer exists. Forrester found 61% of B2B buyers already use a private, company-supplied AI tool for work, and 47% build an internal business case before ever contacting a vendor. By the time your sales team hears from a prospect, the shortlist may already be set.
Citation changes preference, not just awareness. 85% of B2B buyers hold a more favorable view of a vendor once they see it cited inside an AI answer, per G2. That is not a small nudge. It is close to the entire buying committee.
The behavior is consistent across very different purchases. A B2B software buyer, a car shopper, and a high school student choosing a college show the same pattern: rising AI use, growing trust in what the AI surfaces, and real decisions (a vendor switch, a dealership visit, a college dropped from the list) that follow from it.
Trust is not unconditional. 64% of B2B buyers hit an AI hallucination “often” or “very often,” per G2, and review-site citations remain the single biggest trust signal inside an AI answer. Buyers use AI heavily, but they still check its work.
How to check whether you show up in these buyer answers
The statistics above describe a research process that now runs mostly out of view, inside a chat window, before a buyer ever fills out a form. The practical question for your own brand is not whether this shift happened. It is whether you appear when it does.
For programs that need to bring this kind of data to leadership, Profound tracks citation and mention share at enterprise scale across nine or more answer engines, with the source attribution that survives an executive review.
Temso is a lighter, all-in-one starting point: one flat subscription from $89/mo that tracks how often your brand appears across the major AI engines, shows which competitor sources get cited in your place, and turns that gap into a content brief, with no per-engine add-on fees.
For a narrower, lower-commitment check, Peec AI and Otterly.AI offer lightweight monitoring built around prompt-level citation tracking, a reasonable first look before you commit to a full program. A full side-by-side comparison of these and other platforms is at /rankings/geo-tools.
Pick five buyer-intent prompts a real customer would type before they ever contact you. Run them across ChatGPT, Perplexity, and Google AI Overviews. Write down whether your brand shows up, and whose brand it loses to when it does not. That check takes twenty minutes and tells you more than any single row in the tables above.
Sources and methodology: every statistic on this page names its primary source, sample size where disclosed, and publication year. Rows marked (vendor) come from a single company’s own data and have not been independently replicated. This page is refreshed quarterly. Last full review: September 2026.