6sense and G2 Disagree on How Much AI Chat Shapes Vendor Choice
6sense says 21% of buyers credit AI tools before they contact a seller. G2 says chatbots are software buyers' top shortlist influence, at 54%. The two surveyed different kinds of buyers.
TL;DR
- 6sense says 21% of buyers credit AI tools before contacting a seller. G2 puts chatbots first among shortlist influences at 54%.
- Both surveys find buyers mostly use chatbots to compare vendors they already know. Only 26% in 6sense's data used one to find new names.
- Demand gen leaders should fund AI visibility for vendor comparison first, and test discovery against their own win-loss data before spending more.
6sense published its 2026 Global B2B Buyer Experience Report on November 2, and it concludes that buyers have not handed vendor search to chatbots. In its data, 81% of buyers used an LLM somewhere in an evaluation and 26% used one to find new vendors. Only 21% named AI tools as an influence on vendor preference before contacting a seller, behind their own existing knowledge (34%) and internal discussions (32%). The report's line is that "use is not the same as influence." G2 came away with a louder message when it surveyed 1,076 software buyers in March. AI chatbots were the top influence on shortlists at 54%, ahead of review sites at 43%, and 51% of buyers start research with a chatbot more often than with Google.
A demand gen leader setting a 2027 budget can cite either one. Neither settles the question, and both vendors have a stake in their answer. We think the useful information sits in where the two agree and in who each one asked.
The two surveys agree on use and comparison, then split on influence
They're closer than their headlines. G2 puts chatbot use at 71% at some point in research, below 6sense's 81%. G2's top use case is comparing vendors' strengths and weaknesses, with initial vendor identification third. "Buyers aren't asking AI chatbots to orient them in a new category. They already know what they're looking for," G2 writes, in a line that could sit in 6sense's report. On finding new names, 6sense's 26% isn't far from G2's 33% who bought from a vendor they had never heard of. The questions differ, though, and G2 doesn't say whether that 33% covers everyone or only the 69% who said a chatbot changed their choice.
The real split is on influence, at 21% against 54%. The trend lines diverge too. 6sense says LLM use fell from 94% to 81% in a year, and its methods note that cross-year results don't follow the same respondents. G2 says starting research with a chatbot rose from 29% to 51%. Those are different measures, so neither one disproves the other.
Who was asked and what they could answer explain much of the gap
The buyers differ. Software was 30% of purchases in 6sense's main survey, behind services at 44%, with physical goods at 27%. Its methods place the detailed LLM questions in a separate AI survey. There, software was 24% of purchases, 56% of respondents were C-level or VP, and nobody came from APAC. Every purchase in both 6sense surveys was worth at least $25,000 a year. G2's sample is software buyers from individual contributors to the C-suite, across North America, EMEA and APAC. 6sense says its drop in LLM use was steeper among individual contributors, so seniority alone could move the number. We found no split of its LLM figures by purchase type, so any effect of purchase type is our reading and wasn't measured.
The answer choices differ as well. G2's published list has no option for what buyers already knew or for internal discussion, the top two answers in 6sense's data. Set those aside and 6sense still has AI tools at 21%, level with vendor web content and just behind analysts or consultants (24.5%) and review sites (23.8%). G2 has them first by 11 points. The answer choices explain only part of the gap.
Neither source is neutral. 6sense sells intent data, and in July, before this data came out, a 6sense research analyst shared a post by head of research Kerry Cunningham in the company's forum. It was titled "How B2B Buyers Use LLMs to Discover Your Business: They Don't." G2 sells reviews and profiles, and its report calls review sites "the trust layer in AI search." We used G2's 51% when we argued for measuring AI visibility, and that figure carries the same stake. Every figure in both reports is self-reported, which 6sense's report states directly: "all survey measures are self-reported." We didn't find an independent dataset that measures AI's role in vendor discovery.
What this means for demand gen leaders planning 2027 budgets
Match your deal to the sample before you cite either number. If you sell software to buyers at mixed seniority levels, G2 is closer to your buyer. If you sell services or larger purchases to senior teams, 6sense is closer. Ask any vendor selling you AI visibility which survey its pitch rests on, and who that survey asked.
Fund AI visibility first for the job both surveys support, which is showing up accurately when a buyer asks a chatbot to compare vendors by name. Give discovery more budget only when your own deals show buyers finding you there.
Get that evidence by adding two questions to every win-loss review. How did the buyer first hear of you, and was a vendor already favored before first contact? 6sense's headline finding is that the pre-contact favorite wins 82% of the time. But buyers named that favorite after the deal, and they may have named the vendor they ended up choosing. Two quarters of your own answers will tell you more about your market than either survey, so hold back part of the 2027 line until you have them.
Put the two together and the likeliest reading is that chatbots do most of their work once a buyer is comparing vendors.
Check your own win-loss records this quarter for whether the buyer already favored you before first contact, and size next year's between-cycle budget line on what they show.
