The practice room plays the buyer back at you. Not a generic objection-handling drill — the specific company, built from the specific dossier, with the constraints that company’s public record actually supports.
This is what makes the Scout hand-off worth having. Research you read is research you half-remember. Research you had to argue against for ten minutes is research you have.
What the practice room is#
A timed voice conversation with a simulated buyer. You talk; it answers in role. It has the dossier, it does not have your pitch, and it behaves like somebody whose day this call is interrupting.
- Grounded, not improvised. Every fact the buyer asserts about their own company traces to a cited record. It cannot invent a budget cycle to make the exercise harder.
- In role, not adversarial for its own sake. The difficulty comes from the situation, not from a hostility setting.
- Repeatable. The same dossier produces the same buyer, so you can run the opening three times and hear the difference.
Starting a session#
Pick the account
From the account view, or from a booked meeting. If no dossier exists, one is requested first — the room will not open on an ungrounded persona.
Choose the scenario
First call, follow-up after silence, a specific objection you keep losing to, or a negotiation with a stated constraint. The scenario changes the buyer’s opening posture, not their facts.
Talk
Voice, in the browser. Microphone permission is requested at the start of each session and the stream is not retained after it ends.
Read the review
Coverage, objections handled and missed, and the moments where what you said did not match what the evidence supports. See scoring a rehearsal.
How the buyer is modelled#
The persona is assembled from four things, all of which are visible before the session starts.
| Input | Supplies | Source |
|---|---|---|
| Role and tenure | What they care about and how much authority they have | Dossier section 03 |
| Constraints | What they cannot do, and why | Dossier sections 02 and 04 |
| Incumbents | What they already use and how invested they are in it | Dossier section 05 |
| Open questions | What they will be vague about, because the record is vague | Dossier section 06 |
{
"session_id": "PRC-771",
"scenario": "first_call",
"duration_minutes": 10,
"persona": {
"role": "Head of Operations",
"tenure_days": 41,
"authority": "recommends, does not sign",
"constraints": [
"second depot opening in six weeks",
"no dedicated tooling budget line this year"
],
"incumbents": [
{ "tool": "spreadsheet-based dispatch", "attachment": "low", "evidence": "EV-2317" }
],
"will_be_vague_about": ["renewal dates", "headcount plans beyond this quarter"]
},
"grounded_in": ["DOS-1188"]
}will_be_vague_about is the honest part. Where the public record does not establish something, the buyer does not know it either — because a persona that confidently answers questions the evidence cannot support would teach you to expect answers you will not get.
Scoring a rehearsal#
The review is descriptive. There is no single number, because a single number for a sales conversation is a number you learn to optimise instead of learning to sell.
| Reviewed | What it tells you |
|---|---|
| Coverage | Which required qualifiers you established, and roughly when in the call you got to each |
| Objections | Every concern raised, whether it was addressed, and whether the answer landed on the actual concern |
| Evidence use | Whether the claims you made about their company match the cited record — flagged where they did not |
| Airtime and pace | Talk ratio and the longest uninterrupted stretch, shown as a fact rather than a target |
| The three moments | The three points where the review judges the call changed direction, with the exchange quoted |
The evidence-use row is the one worth taking seriously. If you asserted something on a practice call that the dossier does not support, you would probably have asserted it on the real one, and the real buyer knows their own company better than you do.
Limits#
Stated plainly, because a rehearsal tool that oversells itself is worse than none.
- It is not a predictor. A good rehearsal does not mean the call will go well. It means you will not be surprised by the obvious parts.
- The buyer is only as good as the dossier. A company with a thin public record produces a thin persona, and the room says so at the start rather than improvising the gap.
- It models a situation, not a person. It does not impersonate a named individual’s manner, and it will not attempt to.
- Voice quality varies. Latency depends on your connection; on a poor one the room degrades to text rather than producing a conversation with unnatural pauses in it.
Next: connecting a mailbox, which is the one piece of setup that has to be done exactly right.