How to Run a Discovery Call

The structure below isn't a script, and it isn't a framework to memorise. It's roughly what a useful discovery conversation tends to look like when it goes well — and the parts sellers most often skip.

Before the call: do enough homework to be worth talking to

Preparation isn't about reciting a company's funding history back at them. It's about arriving with a specific hypothesis: given what this company does, who this person is and what's changed recently, here's what I suspect might be hard for them. A hypothesis gives the conversation somewhere to start and gives the buyer something to correct, which is far more useful than an open-ended "so, tell me about your business."

Open by agreeing what the call is for

Thirty seconds spent setting up the conversation saves the whole call. Say what you'd like to understand, say what you'll do at the end depending on what you find — including the possibility that this isn't a fit — and ask what they want out of the time. It also gives you permission to ask harder questions later, because you agreed at the start that understanding the problem was the point.

Follow the problem, not the question list

The most common failure in discovery is moving on too early. A buyer gives a general answer — "reporting is a bit manual" — and the seller ticks it off and asks the next unrelated question. The interesting part is always one or two follow-ups deeper: how manual, who does it, how long it takes, what it stops them doing, what they've already tried.

If you have a list, treat it as a checklist for after the call, not a running order during it. Notes are for confirming you covered what mattered, not for driving the conversation.

Find the consequence, not just the pain

Problems that cost nothing don't get funded. That doesn't mean forcing a number out of someone; often the consequence is a name rather than a figure — a team that's stuck, a launch that slipped, an executive who keeps asking a question nobody can answer. Until the consequence is on the table, you have a topic of interest rather than something a buyer will spend money and political capital solving.

Understand how a decision would actually get made

Not as an interrogation about budget and authority, but as a practical question: if you decided this was worth doing, what would happen next in your business? Who'd need to be involved, what would they want to know, when has something like this been done before and how did it go? Buyers usually answer this openly, because you're asking about their process, not testing their seniority.

Be willing to conclude there's nothing here

The hardest skill in discovery is saying, out loud, that what you've heard doesn't sound like a problem worth solving right now. Sellers who can do that get more honest conversations, because buyers can tell when someone is genuinely trying to work something out rather than steering toward a demo regardless of the answers.

Close by playing it back

Summarise what you understood — the problem, what it costs, who's affected, what happens if nothing changes — and let the buyer correct you. Getting it slightly wrong is useful; the correction is often the most valuable thing said on the call. Then agree a next step that matches what you actually found, rather than the next step you always propose.

• Did I understand what this costs them, specifically?
• Could I explain their problem to a colleague without using our product's language?
• Did the buyer learn something about their own situation?
• Do I know what would have to be true for this to go ahead?

If a call fails that check, more questions wouldn't have fixed it. Staying on one of them longer would have.