How to Run a Discovery Call

2026-06-11

title: "How to Run a Discovery Call" description: "Discovery is the most important call in the sales process and the most badly run. Here's what it's actually for, what good discovery looks like, and how to stop leaving deals on the table." date: "2026-06-11" slug: "how-to-run-a-discovery-call" category: "Sales" readTime: "7 min" author: "Numari" tags: ["discovery call", "sales discovery", "SPIN selling", "Gap Selling"]

How to Run a Discovery Call

Most discovery calls are product demos in disguise.

The seller asks a few questions, finds a problem they can solve, and pivots to showing the product. The prospect gets a demo they weren't ready for. The call ends with "I'll think about it." The seller follows up twice and gets silence.

A discovery call that actually works looks completely different. It's not a qualification checklist and it's not a precursor to a pitch. It's the conversation where the prospect diagnoses their own problem - specifically, honestly, in their own words - and arrives at the conclusion that solving it matters.

When discovery works, the close that follows is almost self-executing. When it doesn't, no amount of good pitching recovers it.

Here's what a real discovery call looks like.

What discovery is actually for

Discovery has one job: make the prospect's problem real.

Not acknowledged - real. There's a difference. A prospect who says "yeah, that's something we deal with" has acknowledged the problem. A prospect who says "honestly it's been affecting the team for six months and we've tried three things that didn't fix it" has made it real.

The first prospect might buy eventually. The second prospect is already selling themselves.

Everything in a good discovery call is aimed at moving from acknowledged to real. The questions, the follow-ups, the silences - all of it is designed to take the prospect from the surface of their problem to the felt experience of it.

This matters because people don't buy solutions to problems they've acknowledged. They buy solutions to problems they feel.

The structure of a discovery call

Open with the frame (2 minutes)

Before you ask anything, set the context for the call. What are you both trying to find out? What happens at the end if it goes well?

"What I'd like to do today is understand your situation properly - what's working, what isn't, where you're trying to get to. If it seems like there's a genuine fit, we can talk about what that looks like. If it isn't, that's fine too - I'd rather we both know early. Does that work?"

This does two things. It signals that you're not going to pitch for an hour. And it establishes that a decision or clear next step is the expected outcome - which makes the "I need to think about it" stall harder to deploy at the end.

Current state (5 minutes)

Understand their world before you try to change it. How do they currently handle the problem area? What tools, processes, or approaches are in place? Who's involved?

Keep this brief. You're establishing context, not conducting an audit. Three or four questions at most.

"How are you currently handling X?" "How long has that setup been in place?" "Who else is involved in that process?"

Problem (8-10 minutes)

This is the longest and most important stage. You're finding the friction in their current situation - and making it specific.

Most sellers stay at the first layer of the problem. The prospect says "it's inconsistent" and the seller moves on. Stay with it.

"What does inconsistent look like specifically - is it the results, the process, or something else?" "How often does that happen?" "What have you tried to fix it?" "Why didn't that work?"

The "why didn't that work" question is particularly powerful. It surfaces the real shape of the problem - the part that previous solutions missed - and it tells you exactly what your solution needs to do differently to matter.

Cost (5 minutes)

This is the stage that turns acknowledged into real.

One question, asked about the problem they've just described:

"What does that get in the way of?"

Then follow it wherever it goes. The downstream effects. The team impact. The personal frustration. The missed opportunity.

"What's the impact on X when this happens?" "What would it mean for you specifically if this were sorted?"

The prospect connects the problem to its cost - in their words, unprompted by you. That connection is the emotional engine of the buying decision. You didn't assert it. They said it.

Future state (3-5 minutes)

Once the current state and problem are real, ask the prospect to describe what they're actually trying to achieve.

"What would ideal look like?" "If this were working properly six months from now - what's changed?"

Their answer is their vision of the value of your solution. When you eventually introduce the product, you're not describing features - you're showing them the bridge between where they just said they are and where they just said they want to be.

The questions that don't work

A few discovery questions that reliably produce shallow answers:

"What are your main challenges?" - Too broad. The prospect gives a generic answer that doesn't surface anything useful.

"What does your ideal solution look like?" - Too early. Asked before the problem is real, this produces a feature wish list rather than an honest expression of need.

"Are you happy with your current setup?" - Closed question, almost always answered with "yes" regardless of reality.

"What's your budget?" - Too early in discovery, feels transactional, produces defensive answers. Budget comes later, in the context of a problem they've already made real.

The pattern: questions that can be answered without engaging with the real problem don't surface the real problem. Ask questions that require the prospect to be specific.

What to do when the prospect answers too briefly

Discovery stalls when prospects give short answers. "Yeah, it's fine." "We manage." "Not really an issue."

Two moves that reliably open these up:

"Say more about that." Simple, non-threatening, invites elaboration without specifying what direction to go.

Reflect and ask. "It sounds like it's mostly working but there's something that isn't quite right - what's the part that bothers you most?" You're not putting words in their mouth. You're giving them a frame that makes it easier to be specific.

Short answers usually mean one of two things: the prospect hasn't thought about this in the terms you're asking, or they're being cautious with someone they don't fully trust yet. Both are solved by staying curious and patient rather than moving on.

When to introduce the product

Not until the prospect has described their problem specifically and articulated what solving it would mean for them.

The instinct to pivot earlier is strong - you've identified the problem, you know you can solve it, and it feels like you're wasting time by staying in questions. You're not. Every minute you spend in genuine discovery is a minute that makes the solution more relevant when it arrives.

A prospect who has spent twenty minutes describing their problem in detail and articulating the cost of it is primed for the product. A prospect who has answered three surface questions is not.

The pivot moment is when you can connect their exact words to your exact solution. "What you just described - [their words] - is precisely what this is built for." That connection is only possible if discovery was specific enough.

How Numari supports discovery in real time

Good discovery requires tracking multiple threads simultaneously - the current state, the problem, the cost, the future state - while also listening, maintaining rapport, and deciding what question comes next.

That's the cognitive load that causes sellers to rush discovery or stay shallow. They're tracking too many things at once.

Numari keeps the discovery framework live alongside the call. When the prospect names a problem, it surfaces the follow-up question that goes deeper. When the cost needs to be surfaced, it provides the question that does it. When the future state is ready to be built, it flags the move.

You stay in the conversation. The framework tracks itself. Discovery goes as deep as it should, consistently, on every call.

The call that does the selling for you

A discovery call done properly produces a prospect who has sold themselves. They've described the problem. They've articulated the cost. They've said what better looks like. Your job in the second half of the call is to connect what they said to what you do - and ask if they're ready to close the gap.

That's the discovery call that closes deals. Not the one that ends with "I'll send you some information."

Frequently Asked Questions

What is a discovery call in sales?

A discovery call is the conversation where you understand the prospect's situation specifically enough to know whether and how you can help. It's not a qualification checklist - it's the conversation where the prospect diagnoses their own problem, in their own words, and arrives at the conclusion that solving it matters.

What questions do you ask on a discovery call?

Discovery questions work in layers. Current state questions establish context. Problem questions surface friction (what isn't working, how long, what they've tried). Cost questions make the problem real ('What does that get in the way of?'). Future state questions complete the picture ('What would ideal look like?').

How long should a discovery call be?

A proper discovery call typically runs 30-45 minutes. The problem and cost stages combined (10-15 minutes) are the most important and most commonly rushed. A call that ends in 20 minutes usually stayed shallow - and shallow discovery means the close that follows will be hard.

What is the difference between a discovery call and a sales call?

Discovery and sales are stages within the same motion. A discovery call is one where the primary job is understanding the prospect's situation. A sales call is one where the primary job is connecting that situation to your solution and asking for a decision. Discovery has to come first - the close follows from it.

Numari keeps the discovery framework live in real time - surfacing the right question at each stage so discovery goes as deep as it should, every call. Try Numari →