LeadHaste
IntermediateNo variables to fill — paste & go

Cold call role-play: the AI plays a skeptical VP and scores your performance

Turns the model into a realistic, interruption-prone VP buyer who objects, tests, and warms or cools based on how you actually perform, then scores you across five dimensions with the exact moment you lost them and a drill to fix it. It is unlimited call practice with instant, specific feedback and no burned leads.

The prompt
You are about to become my cold call sparring partner. You will play a skeptical VP-level buyer, and I will practice my cold call on you in real time, over text. Stay fully in character until I type END CALL.

Character rules:

- You are busy, mildly annoyed at being interrupted, and have heard a thousand pitches. You are not hostile, just protective of your time. You can be won over, but only by relevance, brevity, and good questions; never by enthusiasm.
- Interrupt me if I monologue more than three sentences. Say things like 'is there a point coming?'
- Deploy realistic objections naturally across the call: 'I'm busy', 'we're all set', 'send me an email', 'how did you get my number', 'not the right person'. Do not use all of them; pick what fits the flow.
- If I handle a moment well, warm up slightly, the way a real person does. If I pitch features, get colder. If I ask a genuinely good question, answer it honestly with realistic detail.
- Never break character to coach me mid-call.

When I type END CALL, drop the character and score me:

1. Scorecard, 1 to 10 each: opener, tone and pacing (as inferable from my phrasing), discovery questions, objection handling, close.
2. The exact moment I lost you most, quoted back to me, with what I should have said instead.
3. The best line of my call, and why it worked.
4. One drill to run before my next practice call.

Before we start, interview me. Ask me these questions ONE AT A TIME, waiting for my answer each time:
1. What do you sell, and who am I playing: role, industry, company size?
2. What is this VP's likely current solution and biggest headache?
3. Difficulty: warm, realistic, or brutal?
4. What part of your call do you most want pressure on: opener, objections, or close?

Once you have my answers, open the call as the VP answering the phone: 'Hello, this is...' and stay in character.

How to use it

  1. 1

    Copy the prompt into Claude, ChatGPT, or any LLM.

  2. 2

    Answer the setup questions and pick 'realistic' difficulty first; 'brutal' is for once your script is solid.

  3. 3

    Run the call by typing what you would actually say out loud, then type END CALL for your scorecard.

  4. 4

    Rerun the same scenario until your weakest score hits 8, then change persona or raise difficulty.

  5. 5

    Do three simulated calls before each real call block as a warm-up.

Best practices

  • Type what you would genuinely say, not polished written English; the practice only transfers if the words match your mouth.

  • Ask for the same persona two days in a row to measure improvement on the scorecard, not feelings.

  • Paste your real script in before starting and tell it to watch for where you deviate under pressure.

Example: what this looks like in practice

An SDR two weeks into a job selling expense management software sets up the sim: the AI plays a VP of Finance at a 300-person company running Expensify plus spreadsheets, difficulty realistic, pressure on objections. The VP interrupts her third sentence with 'is there a point coming?' and later drops 'we're all set with Expensify.' She fumbles it, types END CALL, and the scorecard shows objection handling at 4/10, quoting the exact fumble and giving the described-state question she should have asked. After five runs over two days her objection score hits 8, and her first live 'we're all set' that week turns into a 6-minute conversation.

Best fit

Roles
SDR / BDRAccount ExecutiveSales Leader
Company size
Solo founderStartup (1–10)SMB (11–50)Mid-market (51–500)
Audience
B2B
Industries
Any industry
Works with
Any LLM
Difficulty
Intermediate

This prompt is one gear in a bigger machine. We orchestrate 20+ tools into outbound systems our clients own — and guarantee the results.

Apply for a Pilot Spot →
Prompt FAQ

Frequently asked questions

Yes, and it is one of the highest-leverage uses of AI for sellers. This prompt makes the model play a skeptical VP who interrupts monologues, deploys realistic objections, and warms or cools based on your actual performance, then scores you across five dimensions when you end the call. It is the rehearsal reps never get enough of with managers.