How AI Prepares Reps for Discovery Calls

AI is most useful before the discovery call, not during it. I use it to build a brief for every prospect: who they are, what they are working on, and the questions most likely to uncover friction in their business. A rep who walks in with that brief asks better questions and wastes less of the buyer's time.

What goes into a good discovery brief?

Start with what is public. I plug in the person's professional profile and the company website, and the brief comes back with 8 to 10 solid questions about potential friction areas I should dig into. For a medical practice, that might be five or six questions around patient acquisition. For a physician, the brief can line up the clinical trials they have been part of. For a hospital system, it can list the major contracts they already hold.

There is also a lot of public data in US healthcare. The Sunshine laws require that payments from industry to physicians are published, so you can see who gets paid what from whom. A good brief pulls that in, so the rep knows the relationships before the first meeting.

Once the process is set up, a call brief takes about three minutes. That is the real change. Prep that often never happened is now something every rep can do before every call.

How do you prepare for a hospital deal with many buyers?

A health-system purchase is not one buyer. You are dealing with a whole group, from IT security to purchasing to nursing to finance. I use AI to map each of those personas and build a separate discovery brief for each one inside the system. It tells the rep, for example: here is the chief of nursing, here is her likely view of your initiative, and here is what you should be trying to discover.

This matters because value points do not travel between buyers. The clinical value point is not going to mean anything to procurement. You can talk about the best quality of care all day, and the procurement or finance person will think that has nothing to do with them. The brief helps the rep connect the right value point to the right person.

Should reps pitch on the discovery call?

No. I see so many sales organizations that discover and pitch in the same call. Most customers would rather have you come back with something meaningful than watch you start shooting left and right in the first meeting. Use the discovery call to learn. Then go away and build something specific to them.

That is where AI helps a second time. I use transcription on my calls, then plug the transcript into a proposal document that has a growth plan for the customer. The proposal reflects what they actually said, not a generic deck.

What can't AI do in discovery?

AI doesn't fix anything unless you tell it what to fix. You have to be really prescriptive with it and ask it the right questions. I sort sales work into four boxes: high value or low value, and highly repetitive or not. The highly repetitive, low-value work goes to AI. The high-value human work, like building trust and closing deals, stays with the rep. This is about augmenting people, not replacing them.

Where to start

Pick your next five discovery calls. Before each one, give an AI model the prospect's profile and website and ask for the likely friction points and the questions to test them. After each call, feed the transcript back in and draft the follow-up proposal from it. Then compare those calls with your usual ones.