thinkhard.SI

Prompt repository Sales

Rehearse a sales call with AI playing the buyer

You need to practice discovery, objections or a negotiation with a realistic buyer before the real call.

Registry no.
SAL-009
Collection
Sales
Laid out for
Any chat model
Score
10 out of 10
Blanks to fill
7
Added

SalesAny chat

The prompt

Score 10 of 10, Excellent

Use it when You need to practice discovery, objections or a negotiation with a realistic buyer before the real call.

  • {{call_type}}
  • {{buyer_role}}
  • {{buyer_situation}}
  • {{hidden_concerns}}
  • {{difficulty}}
  • {{what_i_sell}}
  • {{call_goal}}
Full prompt and why it works
Prompt
Let's rehearse a sales call. You play the buyer and I play the salesperson. I am preparing for a real {{call_type}}, and the point is to practice under realistic pressure, so a buyer who is too easy is no help to me.

The buyer you play:
Role: {{buyer_role}}
Their situation: {{buyer_situation}}
Concerns they hold but will not volunteer: {{hidden_concerns}}
Difficulty: {{difficulty}}

What I sell:
"""
{{what_i_sell}}
"""

My goal for the call: {{call_goal}}

How to play the buyer:
- Reply only as the buyer, in 1 to 4 sentences, the way a busy professional talks. Do not coach me or step out of character during the call.
- Answer what I ask, but reveal the hidden concerns only when my questions earn them. A vague question gets a vague answer, as it would in real life.
- Raise the objections a real buyer in this role would raise, such as price, timing, risk or the effort of switching, and push back if my answer is weak.
- Do not agree to a next step just because I asked. Agree only if I have given you a reason to.
- If anything is missing from my description that the buyer would know, such as their budget or who else decides, decide it once, keep it consistent, and tell me what you decided in the feedback.

Start by opening the call as the buyer, in one line, and then wait for me.

When I type "end role-play", step out of character and give feedback in this format:
1. Outcome: what this buyer would honestly do next, and why.
2. What worked: two specific moments, quoting my words.
3. What to fix: three specific moments, each with what I said and a better line I could have used.
4. Questions I should have asked and did not.
5. Hidden concerns: which ones I uncovered and which I missed.

The feedback should be direct and specific, the way a good sales coach gives it, without softening. Keep it under 400 words.

Variables

{{call_type}}
The call you are preparing for, for example "first discovery call", "pricing negotiation" or "renewal with an unhappy customer".
{{buyer_role}}
The buyer's job title and the kind of company they work for.
{{buyer_situation}}
What is going on for them: the problem, any current supplier, pressures and deadlines.
{{hidden_concerns}}
Worries or objections the buyer has but will only reveal if asked well, for example "burned by a failed rollout last year".
{{difficulty}}
How hard to make it: "friendly", "realistic" or "tough and skeptical".
{{what_i_sell}}
Your product or service, its price range and what it does, as you would describe it on the call.
{{call_goal}}
What you want from the call, for example "agree a demo with their IT lead".

Why it works

  • What to avoid. The buyer reveals concerns only when your questions earn them and will not agree to a next step just because you asked, which is what makes the practice worth doing.
  • Format. Feedback waits until you end the role-play and comes in a fixed order with a better line for each weak moment, so it never breaks the rehearsal.
  • If unsure. Details you left out are decided once, kept consistent and disclosed afterward, so the buyer does not change their story halfway through.

How to use it

  1. Press Fill in variables and type your details into each blank. What you type stays in this browser.
  2. Copy the filled prompt into the AI tool you already use, such as ChatGPT, Claude or Gemini, or open it in the Grader or the Prompt Tester here.
  3. Before you rely on it, try it on three or four real examples. The Prompt Tester runs them all at once and shows which ones go wrong.
Use it from code

Get this prompt as JSON from the free REST API, with your own text in the blanks, or let an AI assistant fetch it through the MCP server. No key.

curl -s https://thinkhard.si/api/v1/prompts/sales-call-role-play

With the MCP server added to Claude, ChatGPT, Cursor or VS Code, call get_prompt with {"prompt": "sales-call-role-play"}, or just ask for it by name.