Coach discovery calls by checking what buyers researched with AI, which claims they raised, and how the rep responded.
Key takeaways
- Ask what the buyer researched before the call and which source shaped their view.
- Record the AI-generated claims the buyer wanted the rep to check.
- Compare the buyer’s stated need with the rep’s reply.
- Keep each research figure’s source and qualifier attached when discussing it.
Why buyer AI research changes discovery coaching
Twice as many buyers named generative AI or conversational search
The first stage of the buying journey is increasingly happening outside the seller’s funnel, according to Aviso. That shift changes what a discovery call can assume about the buyer. Your coaching should treat the buyer’s earlier research as a fact to check rather than as a blank slate for the rep to fill. A buyer may arrive with a view already formed, and your review should begin from that view. Your call notes can record the source of that view.
Twice as many buyers named generative AI or conversational search as their single most meaningful research source, compared with any other single source, including vendor websites, product experts, and sales representatives, according to Aviso. In a call, you can ask which source shaped the buyer’s view, and record the answer in your review.
Your notes should keep this figure about research sources, not use it to rank any one vendor. Coaching the rep on this point means asking what the buyer read before the first meeting. Your review can compare the buyer’s answer with what the rep assumed about that research. Where the two differ, the rep has a specific gap to address. That gap is a coaching point, not a verdict on the rep’s preparation.
The figures in this guide come from a blog page by Aviso, which reports research attributed to Forrester and Gartner. The numbers are reported figures, not independent proof that you have verified. Your team meeting can name the research next to each figure, so the source stays visible. Keep the wording of each figure when you repeat the numbers in a meeting.
Each figure has its own qualifier, and the qualifier is part of the figure. The research describes the buyers it studied, and your buyers may differ. A figure that describes one group of buyers may not describe the accounts your reps call on, so the number is best used as a question for the call.
94% of buyers use AI in their purchase process
Forrester’s 2026 research, based on nearly 18,000 global business buyers, found that 94% now use AI in their purchase process, up from 89% in 2025, according to Aviso. The sample is named in the same sentence as the number, so keep the two together when you quote it. A figure with no sample attached is harder for a team to check.
Your review can ask whether the buyer used AI before the first meeting. Your notes can record that answer beside the rep’s first question.
Your planning can treat the 94% figure as context, not as a score for the rep. A planning note can name research the buyer may have done, then leave room for the buyer to correct that picture. A planning note stays useful when the rep checks it during the conversation.
A useful review question is simple: what did the buyer read or ask an AI tool before this call? The answer is a practical starting point for the rep’s first question, not a verdict on the buyer. Your review can note whether the rep asked the question at all. If the rep did not ask, the note records that gap, and the next call can test it again.
69% still check AI insights with a rep
Among B2B buyers, 69% still turn to sales representatives to validate AI-generated insights at key decision points, according to Aviso. The phrase at key decision points is the qualifier to keep with this figure. Your notes should not stretch the number into a claim about every stage of the deal.
The verb in the finding is validate, so the moment to watch is when the buyer tests an AI insight with the rep. Your coaching can treat each AI insight the buyer brings as a claim to check, not a conclusion to accept. That framing matches the verb used in the source’s wording. The rep then has a clear job in that moment.
Your rep can list the AI-generated claims the buyer raised during discovery, then mark each claim as confirmed or corrected. A short list like that can sit at the top of a one-on-one agenda. Leave a claim unresolved when the rep has not checked it.
The 69% figure is a reported share, not a target for your team, so it should not become a quota or goal. A reported share describes what buyers did in the research, which may not match your accounts. Your review can treat the share as a reason to check that moment, not as a benchmark for reps.
Confidence and understanding: the 32 and 39 point gaps
Feeling confident and feeling understood
Gartner found buyers were 32 percentage points more likely to say a sales representative made them feel confident in their decision, and 39 points more likely to say the rep understood their needs, according to Aviso. The two gaps measure different things, so keep them separate in your notes.
Your review can check whether the buyer seemed more confident after the rep’s explanation, and note the words the rep used. The note records a specific moment in the call without treating it as a measured change. Without that moment, the figure stays a number with no place in the conversation.
Your review should find the point where the buyer described a need, and check what the rep said back about that need. A rep who repeats the need in the buyer’s words gives you a clear example to discuss. Put that example next to the buyer’s original statement.
The figures describe buyers in the research, and your buyers may differ, so use the comparison as a prompt for coaching. A prompt can become a question the rep asks the buyer during the call.
Misleading information: 51% for AI and 49% for reps
51% expect misleading information from generative AI
Gartner found that 51% of buyers said they were likely to encounter misleading information from generative AI, compared with 49% who said the same about a sales representative, according to Aviso. The two figures sit close together, and the source reports them side by side for the same question. Your notes can record both numbers as reported, with the question they answer.
Your notes should not turn this comparison into a claim that reps are more reliable than AI tools. A comparison of two close numbers is a reason to ask questions, not a ranking. These figures describe expectations about misleading information, not a test of accuracy.
Your review can use the comparison to ask the rep how the buyer could check a claim made during the meeting. That question moves the discussion toward evidence. It also keeps the rep focused on the claim rather than on the rep’s own reputation.
Your review can record which source the rep pointed to for each claim. That keeps attention on the claim rather than on either party. A recorded source is something the buyer can check after the call.
How to use each figure in a call review
Your call review should start with the buyer’s research rather than the rep’s pitch. Note any AI tool or source the buyer named, and whether the buyer had compared vendors before the call. Those two notes give the rep a starting point based on the buyer’s own account. The notes also show where the rep’s pitch assumed a blank slate.
Your review can treat the 94% figure as background, then check whether your own buyer followed the same pattern. The figure describes the research sample, so a match in your pipeline is a question, not proof. Your review can record whether the buyer mentioned AI in any form.
The 69% figure gives you a reason to watch the moment a buyer checks an AI claim with the rep. Your review can mark that moment in the call notes, with the claim and the rep’s reply. A short note of this kind shows whether the rep treated the claim as a check or as a debate.
Your review can locate the point where the buyer described a need, and check whether the rep’s reply showed that the need was heard. This pairing of a need and a reply is the core of the review. Use it with the 32 and 39 point figures as a check on the call itself.
Your review can use the 51% and 49% figures to talk about where a buyer’s claims came from, not about the rep’s character. The conversation stays on sources, because sources are something the rep can name and check. A rep who names the source has given the buyer something to verify.
The table below identifies the buyer research source used throughout this guide.
| Publisher | Page | Scope |
|---|---|---|
| Aviso | AI is changing the buyer | Reports research attributed to Forrester and Gartner |
Illustrative example: review a buyer’s claim
In an illustrative discovery call, the buyer raises an AI-generated claim, but the rep does not ask where it came from. Record the claim and the missing question. Coach the rep to ask, “Which source shaped your view?” Then practice asking, “What would you like me to check?” Keep the claim unresolved until it has been checked, rather than marking it confirmed because the rep agreed.
Your next step today is to choose a recent discovery call, note what the buyer had already researched, and coach the rep with a question that confirms that research.
Questions and answers
How many buyers use AI in their purchase process?
Aviso reports that 94% of nearly 18,000 global business buyers used AI in their purchase process in Forrester’s 2026 research. Quote the purchase process wording in your notes rather than a broader phrase.
Do buyers still talk to sales reps to check AI answers?
Yes, buyers still turn to sales representatives to validate AI-generated insights at key decision points. Your coaching can focus on those moments during a call review.
Who do buyers find more likely to mislead, AI or reps?
The reported expectation of misleading information was slightly higher for generative AI than for sales representatives. Your notes can name the source a rep pointed to, rather than rank AI against reps.
What should a manager coach after a call where the buyer arrived informed?
Your review can check which claims the buyer raised and how the rep handled each one. Coaching should focus on the buyer’s starting point, not on a generic opening pitch. Your notes can list the claims, then coach the rep to check each one before answering.
Does AI research make discovery calls unnecessary?
AI research does not remove the need to check a buyer’s claims during discovery. The figures describe buyer behavior, and the 69% figure points to a check that buyers make with reps. That check is a reason to keep the call in the plan.
Can I apply these figures to my own buyers?
No, these study results do not establish how your own buyers behave. Ask your buyers about their research before assuming they follow the same pattern.