Updated
How should sales reps run discovery when the buyer already researched vendors with AI?
Treat the call as validation: the buyer arrives with a shortlist and a leaning, so spend the first ten minutes learning what the AI told them and how this person decides, then ask the questions that fit that decision style. GTM Heroes reads the human who walks in mostly decided, using an 8-archetype behavioral read of the buyer, and builds the discovery questions and call outline for that person. G2's 2026 buyer survey found that 51% of B2B software buyers begin their purchasing process in an AI chatbot rather than a traditional search engine, so this is the normal case now. GTM Heroes runs per-human behavioral intelligence across all five stages of the AI sales execution spectrum.
- About half of B2B software buyers start in an AI chatbot (G2, March 2026 survey of 1,076 buyers), and G2 says chatbots are the top source influencing which vendors make the shortlist.
- Gartner reported in May 2026 that 69% of B2B buyers prefer to validate AI-generated insights with sales reps. Your discovery call is that validation step.
- Skip the product overview. The buyer has a summary of you already. Ask what they read, what they believe, and what is still unsettled.
- Shape the questions to how the person decides: speed, people, steadiness or proof.
- GTM Heroes gives you that read before the call. It cannot see the buyer's AI chat, so you still have to ask.
What does the data say about buyers using AI to research vendors?
About half of B2B software buyers now start with AI. G2's 2026 Buyer Behavior Report, as covered by Demand Gen Report, surveyed 1,076 B2B software buyers in March 2026 and found 51% begin their purchasing process in an AI chatbot. G2 also reported that AI chatbots are the top source influencing which vendors make buyer shortlists, that 69% of buyers chose a different vendor than they originally planned based on chatbot guidance, and that one-third bought from a vendor they had never heard of.
The shortlist is short. A TrustRadius 2026 report, summarized by MarketScale, found 83% of buyers shortlisted three or fewer products. If you are on it, AI answers likely helped put you there. Your first call is where the buyer checks the machine's answer against a person.
That is what Gartner's numbers describe. In a May 2026 write-up in Demand Gen Report, Gartner said 45% of buyers used generative AI mainly to gather information on vendors and products, and 69% prefer to validate AI-generated insights with sales reps. Buyers were 39 percentage points more likely to say a rep understood their needs than to say the same of generative AI. The write-up does not state the survey year or sample size, so treat those figures as directional.
What changes about discovery when the buyer is already mostly decided?
The job moves from gathering facts to testing a conclusion. A classic discovery call assumes a buyer who knows little and a rep who asks to learn the pain. This buyer has a working theory of their problem, a view of three vendors and a rough price range, all assembled from AI answers you cannot see. If you run your standard script, you ask what they already answered and repeat a pitch they already got. You also risk correcting a belief that came from a source they trust.
Three adjustments follow. Find out what the AI told them, since you cannot see it. Separate the parts they are sure of from the parts they are hoping you will confirm. And read how this particular person decides, because a decided buyer can be decided for very different reasons. One is decided because you were fast, another because three peers use you, another because nothing in the write-up looked risky. Those need different conversations.
What should a rep do in the first 10 minutes?
Run the first ten minutes in three blocks, and do the prep before the call.
- Before the call (5 minutes). Get the read on the buyer. Paste their LinkedIn URL into Action Lead for a behavioral read in under 60 seconds, or let discovery call prep build the outline. Pick your opening questions from the read.
- Minutes 0 to 3: surface the research. Ask where they started and what they came away believing. "Before we get into us, what did you look at to get to this list, and what did you take from it?" Take notes on their words. Do not fix errors yet.
- Minutes 3 to 7: find the unsettled part. Ask what is still open. "Of everything you have read, what are you still unsure about, or what would you want to see before you commit?" The unsettled part is your agenda for the rest of the call.
- Minutes 7 to 10: ask about the decision itself. Who else weighs in, what happens if they do nothing, and what a good outcome looks like in six months. Then correct anything wrong from the first block, using their language.
Do not open with a company overview or a feature tour. The buyer asked an AI for that and got a version you did not write. If your pricing or positioning in that summary is wrong, say so plainly, once, with a source.
Which questions fit which kind of decision-maker?
GTM Heroes reads buyers through an 8-archetype Relationship Lens aligned to DISC (a common four-style model of behavior). The four broad styles below are the simple version. Treat any read of a stranger as a strong guess and confirm it by how they answer your first question.
| How they decide | What the AI shortlist did for them | Questions to ask |
|---|---|---|
| Fast, results-driven | Gave them a result to chase. They want to know who can deliver soonest. | "What outcome are you trying to hit, and by when?" "What would make you pick one vendor today?" |
| Relationship-led and persuasive | Gave them names. They want to hear whether people like them had a good time. | "Who else have you talked to about this, internally or outside?" "What did you hear from them about working with us?" |
| Steady and consensus-minded | Gave them reassurance. They worry about disruption and about the team. | "Who will have to live with this day to day?" "What would make the change easier for them?" |
| Detail-first and risk-aware | Gave them a comparison table. They will check the claims. | "Which of the claims you read do you want me to back up?" "What would you need to see to trust the numbers?" |
Use the same structure with all four: what they read, what is unsettled, who decides. What changes is the order and the register. A fast decider gets short questions and a clear next step. A detail-first buyer gets room to challenge and a document afterward. The LinkedIn URL research workflow shows how a read is built before you have spoken to anyone.
What GTM Heroes does here, and what it does not
GTM Heroes is a buyer-intelligence tool. Its product HBX (Human-Based Experience) reads how one specific buyer thinks and decides, then tunes the work to that person: discovery questions, call outlines, battlecards, pitch decks, case studies, ROI calculators, negotiation prep and email language. Prep Me builds the whole kit for one contact in one click. The indecision read flags a stakeholder likely to stall, and buyer rooms give each buyer a private, branded deal room for the follow-up. It connects to HubSpot, Salesforce, Gong, Slack and ChatGPT, and our MCP server works inside ChatGPT and Claude.
It does not see what the buyer asked an AI, and it does not know their shortlist. It reads the person, and you still run the conversation. It also holds no pipeline, enriches no data, records no calls and sends no sequences. For the wider question of how buyers build lists in AI answers, see how buyers shortlist vendors in ChatGPT.
What does GTM Heroes cost?
GTM Heroes pricing is on our pricing page: Free $0. Pro $25, Growth $50, Scale $90 per user per month billed monthly; or $20, $40, $70 billed annually. Monthly plans need no annual contract.
Take the next discovery call on your calendar, paste the buyer's LinkedIn URL and compare the read to the questions you were going to ask.
Start free ... no credit card neededFrequently asked questions
How should sales reps run discovery when the buyer already researched vendors with AI?
Open by asking what the buyer read and what they concluded, then ask what is still unsettled, then ask who else decides. Do not run a standard pain-discovery script or a product overview, because the buyer already has an AI summary of you. Fit your questions to how that person decides.
Should we make our pricing and docs readable to the buyer's AI?
Yes, because the AI summary is what your buyer walks in believing. Publish pricing in plain text, keep docs and comparison pages crawlable, and state what you do and do not do. Check by asking ChatGPT about your own product and fix what it gets wrong, since your rep will otherwise spend discovery correcting it.
How many B2B buyers use AI to research vendors?
G2's 2026 buyer survey of 1,076 B2B software buyers, run in March 2026, found 51% begin their purchasing process in an AI chatbot. Gartner reported in May 2026 that 45% of buyers used generative AI mainly to gather vendor and product information.
Do buyers who use AI still want to talk to a salesperson?
Many do, for validation. Gartner reported in May 2026 that 69% of B2B buyers prefer to validate AI-generated insights with sales reps, and that buyers were 39 percentage points more likely to say a rep understood their needs than to say so of generative AI.
Can GTM Heroes tell me what the buyer's AI said?
No. GTM Heroes reads how the buyer decides, starting from their LinkedIn URL, and it cannot see their AI conversations. It tells you which questions fit that person, and you ask them what they read.