Discovery

What Good Discovery Actually Sounds Like

Published: 2026-05-26 · ~5 min read

Every sales book on Earth has a list of "great discovery questions." Reps memorize the list, sprinkle them on calls, and wonder why prospects still don't open up. The problem isn't the questions. It's that discovery is a conversation pattern, not a checklist — and you can't read your way to a conversation pattern. You have to practice it.

Bad discovery: the interrogation

You can spot bad discovery by the rhythm. The rep asks a question, the prospect answers, the rep asks the next question on their list — regardless of what the prospect just said. No follow-up, no curiosity, no thread-pulling. It feels like a survey.

What the prospect hears: "I have a form to fill out and you are the form."

Mediocre discovery: the demo bridge

One step better. The rep asks 2–3 surface questions ("what tools are you using now?", "what's not working?"), then pivots to the pitch. The questions exist only to set up the demo. The prospect knows it.

What the prospect hears: "Pretend I care about your answer so I can show you slides."

Good discovery: the thread

Good discovery starts with one open-ended question — usually about the prospect's current state or a recent trigger event — and then follows the thread. The rep's next question is built on what the prospect actually said, not on what was next in their head.

It sounds like this:

Rep: "What's driving you to look at this right now?"

Prospect: "We had a churn spike in Q1 we couldn't explain."

Rep (bad): "Got it. How many reps are on your team?"

Rep (good): "A churn spike you couldn't explain — what's the leading theory internally? What's the gap between what you know and what you'd need to know to be confident?"

The good rep just earned ten more minutes of conversation, three new data points, and a prospect who feels heard. The bad rep just earned a team-size answer.

Why this is so hard to teach

Because the "good" response can't be scripted. It depends entirely on what the prospect said. You can teach the shape of thread-pulling — open question, listen for emotion or specificity, ask the follow-up — but the muscle of doing it in real time only builds with reps.

And here's the catch: practicing it on real prospects is expensive, because bad discovery loses deals. So most reps never get the volume of practice they'd need to develop the reflex. They stay stuck at "mediocre discovery" for their whole career.

How simulation fixes this

In SharkSale, the discovery dimension is one of the six scored categories. Every session, the rep gets a discovery score and — more importantly — a transcript they can re-read to see exactly where they failed to pull a thread.

The compounding effect is real. A rep who runs 30 simulated sessions and reviews their discovery score after each one will internalize the thread-pulling reflex in two weeks. The same rep, learning it on real prospects, takes a year. Maybe never.

A drill you can run today

  1. Run a SharkSale session and force yourself to ask only open-ended questions for the first 4 minutes. No pitching, no demoing.
  2. Pick the prospect's single most emotionally-charged answer and ask one follow-up that goes deeper on it.
  3. After the session, look at your discovery score. If it's below 60, replay the transcript and find the three moments you could have pulled a thread and didn't.
  4. Run another session. Try to do it differently.

Do this for two weeks. Your real-call discovery will not look the same.


Stop training on real customers. SharkSale puts a live AI buyer on the other end of a Discord voice call, then scores you on six dimensions and hands out certificates when you stop sucking. Start from $9.99 or jump into the Discord.

← Back to all posts