Session 3 · team build

Practice a real call out loud, then score it.

Claude plays the buyer and makes you work for the number and the date. Say END CALL and it drops the role and scores you. Take turns. Everyone gets a go.
  1. Turn on voice mode first. This only works spoken.
  2. Copy the prompt below and paste it into a new Claude chat.
  3. Run the call out loud. One person sells, the rest listen and note what they would have asked.
  4. Say END CALL and read the score together. Then swap seats and run it again.
Paste into Claude, voice mode on
You are going to run a discovery call role-play with me, out loud, and then score me on it.

I am the seller. You are the buyer.

Before we start, ask me these one at a time:
1. What do I sell, in one sentence?
2. Who are you playing, what role and what company type?
3. How hard should you make this: easy, realistic, or difficult?

Then take the buyer role and stay in it. As the buyer:
- Answer only what I actually ask. Never volunteer.
- Have a real problem underneath, with a number and a date attached, but make me earn both.
- If I ask a vague question, give a vague answer.
- If I pitch before I have understood you, go politely non-committal.
- Do not break character to help me. We do that afterwards.

Run it as a real conversation. When I say END CALL, drop the role and score me against SPICED:

Situation is Qualified when I could describe your model, team, and tooling well enough to predict where your handoffs break.
Pain is Qualified when you described a consequence you personally absorbed, unprompted.
Impact is Qualified when you did the arithmetic out loud.
Critical Event is Qualified when there is a date with a consequence that exists whether or not you buy from me.
Decision is Qualified when I can name the signer, your criteria in your words, and the steps to signature.

Evidence has to be something I actually said. With nothing to quote, it is a Gap. Score me honestly, not kindly.

Then show me two moments. My biggest blind spot, in three lines: what you said, what I said, what I should have said. And the strongest thing I did, quoted, with why it worked.
Voice mode and dictation are different jobs

Voice mode is for the practice. In Claude or ChatGPT, turn on voice and have the spoken conversation. You hear yourself ask the question and get an answer back in real time. Reading a good question silently teaches you nothing about whether you can say it.

Dictation is for the input. A tool like Wispr Flow turns speech into text in any field, so you can dump call notes in thirty seconds instead of typing for five minutes. Use it right after a real call, while you still remember what was said.

What your team brings back

A scored practice call for each person, and the one follow-up question each of you should send.

Winning by Design · AI for GTMSession 3 · Run the Call, Prove the Impact