Yes. In G2's March 2026 survey of 1,076 B2B software buyers, 51% said they now start researching vendors in an AI chatbot more often than in Google, up from 29% a year earlier. Gartner's own May 2026 survey found 45% of 645 B2B buyers used generative AI during a recent purchase, mainly to research vendors and products.
If the shortlist is already forming inside a chatbot, a website built only for Google is answering half the buying process. The data below is from two independent surveys and our own AI search statistics, not a guess.
About half, and the share is growing fast. G2 polled 1,076 B2B decision-makers and purchase influencers across North America, EMEA and APAC in March 2026 for its report "The Answer Economy." 51% said they now begin their research in an AI chatbot more often than in a traditional search engine, up from 29% in G2's 2025 Buyer Behavior Report. 61% said they use AI search and Google in tandem, not one instead of the other.
| When asked | Share starting research in an AI chatbot more than Google |
|---|---|
| G2 2025 Buyer Behavior Report | 29% |
| G2 "The Answer Economy," March 2026 | 51% |
Proof: G2 — The Answer Economy: How AI Search Is Rewiring B2B Software Buying.
Why it matters: a jump from 29% to 51% in one year means a site tuned only for a Google results page is already missing roughly half of the buyers who will consider it.
Yes, and that is the part a "ChatGPT replaces the website" story leaves out. Gartner's survey of 645 B2B buyers, fielded August through September 2025 and published May 20, 2026, found buyers used an average of seven information sources per purchase. 45% used generative AI, mainly to research vendors and products, but 69% still turn to a sales rep to validate what the AI told them.
Proof: Gartner — Survey Finds 69% of B2B Buyers Turn to Sales Reps to Validate AI-Generated Insights.
Bottom line: AI chatbots are one of seven sources a buyer mixes, not a replacement for the rest of the funnel. Being named early in that mix still matters, because it decides who gets validated later.
G2 reports that it does for most buyers, not a small minority. In the same March 2026 survey, G2 found that most buyers said an AI chatbot surfaced information that led them to consider, or choose, a different vendor than the one they originally planned to buy from. The chatbot is not a neutral research step, it actively reshuffles the shortlist.
Why it matters: the question is no longer only "do we rank on Google for this term," it is "does ChatGPT name us when a buyer who already has a plan asks for alternatives."
That being named is uneven even for a well-known brand, which is why "we'll probably come up" is not a strategy. Reflexa's AI Visibility Index asks ChatGPT, Claude and Perplexity the same set of buyer questions for a category and counts who gets named. In the September 2026 project-management edition, after excluding questions that already named a brand outright, the category leader was recommended in 359 of 882 eligible answers, 41% (Reflexa AI Visibility Index, September 2026). A brand that assumes it will be "the obvious choice" is still absent from roughly 6 in 10 answers.
The full method behind that number, including why anchored questions get excluded, is in what buyers actually ask AI before choosing software.
Do this: run the same kind of check on your own category before assuming your brand is the exception.
The shortlist often forms before the website gets a visit, so the site's job starts earlier than most teams plan for. When a buyer asks an engine for "alternatives to X" or "best tool for a team of 20," the answer is built from pages the engine can fetch, trust and quote, not from a pitch the vendor controls. A site that cannot be read by GPTBot or OAI-SearchBot, has no clear Organization entity, or never answers the buyer's exact question in plain text is not a candidate for that answer, whatever it ranks for on Google.
Why it matters: AI search statistics keep showing the same pattern across categories: visibility in the chatbot answer and ranking in Google are two different funnels, decided by two different sets of sources.
None of this requires a budget to start checking:
Bottom line: the surveys say the shortlist is already moving into AI chatbots. The only way to know where your own company stands in it is to ask the engines the question a buyer would ask, this week, not after the next renewal cycle.
Yes. G2's March 2026 survey of 1,076 B2B software buyers found 51% now start their research in an AI chatbot more often than in Google, up from 29% in G2's 2025 report. Gartner's own May 2026 survey of 645 B2B buyers found 45% used generative AI during a recent purchase, mainly to research vendors and products.
No. The same Gartner survey found 69% of buyers still turn to a sales rep to validate what an AI chatbot told them, and G2 found 61% use AI search alongside Google rather than instead of it. The chatbot forms an early shortlist; people still confirm it.
Yes. Ask ChatGPT, Claude and Perplexity the buyer questions a prospect would type, such as "best project management tool for a remote team," and read who gets named. Reflexa's free check runs this across all three engines and shows the actual answer, not just a score.
Because Google ranking and AI recommendation are different funnels built from different sources. A page can rank on Google and still be invisible to an engine that cannot fetch it, does not trust its entity, or is simply answering from a roundup that never named the site.
Weekly is the honest minimum. Reflexa's AI Visibility Index reruns the same buyer questions across engines and tracks how the named vendors shift from edition to edition, rather than treating one run as a permanent score.
The free check asks ChatGPT, Claude and Perplexity your buyer's own question and shows exactly who they named. 3 minutes, evidence included.
See the same numbers for your site: every AI visit counted per page, beside Search Console · free first check · på svenska