ICP Pain Extractor: Jobs-to-Be-Done from Interview Notes

Turn interview chaos into ICP clarity. Extract jobs-to-be-done, functional pains, and emotional friction from raw transcripts in minutes—no manual coding required.

Full prompt

You are a jobs-to-be-done analyst. Your task is to extract and prioritize jobs from customer interview transcripts.

Here's the interview transcript:
{{INTERVIEW_TRANSCRIPT}}

Your ICP context (optional, helps filter noise):
{{ICP_CONTEXT}}

Analyze the transcript and return a JSON object with this structure:
{
  "jobs": [
    {
      "job_statement": "[Functional/Emotional/Social] job description",
      "job_type": "functional|emotional|social",
      "supporting_quotes": ["direct quote 1", "direct quote 2"],
      "pains_triggered": ["pain 1", "pain 2"],
      "frequency": "mentioned_once|mentioned_multiple|core_theme",
      "emotional_intensity": "low|medium|high",
      "segment_signal": "which customer type mentioned this"
    }
  ],
  "top_3_jobs_by_priority": ["job 1", "job 2", "job 3"],
  "red_flags": ["assumption we should test", "contradiction in what they said"],
  "next_questions": ["follow-up to ask in next interview"]
}

Prioritize by frequency + emotional intensity. Ignore generic complaints. Focus on the *job the customer is trying to accomplish*, not the tool they mentioned.

How to use

1. Paste a single interview transcript (or 2–3 short ones) into {{INTERVIEW_TRANSCRIPT}}. 2. Add your ICP definition (company size, role, industry) in {{ICP_CONTEXT}}—this filters out noise from outlier customers. 3. Run the prompt. 4. Copy the JSON output into a spreadsheet or Notion. 5. After 8–10 interviews, sort all jobs by frequency + intensity to find your core jobs. 6. Use the red_flags and next_questions to brief your next interview.

Example input

Interview transcript:

Interviewer: What's the hardest part of your current workflow?

Customer: We spend 3 hours a week manually matching customer data across Salesforce and our billing system. It's error-prone. Last month we double-billed a customer because the sync broke.

Interviewer: How does that make you feel?

Customer: Honestly, it's embarrassing. I have to apologize to the customer, and my CEO asks why we're not automated yet. I feel like I'm doing work a script should do.

Interviewer: What would change if this was solved?

Customer: I'd have time to actually analyze churn patterns instead of babysitting data. My team would trust our numbers again.

ICP context: Mid-market SaaS (50–500 employees), RevOps leaders, $2M–$50M ARR.

Example output

{
  "jobs": [
    {
      "job_statement": "Functional: Eliminate manual data reconciliation between systems to reduce billing errors and sync failures.",
      "job_type": "functional",
      "supporting_quotes": ["We spend 3 hours a week manually matching customer data", "Last month we double-billed a customer because the sync broke"],
      "pains_triggered": ["Revenue leakage from billing errors", "Time spent on non-strategic work", "Data integrity risk"],
      "frequency": "core_theme",
      "emotional_intensity": "high",
      "segment_signal": "RevOps leader at mid-market SaaS"
    },
    {
      "job_statement": "Emotional: Restore confidence in data accuracy and reduce embarrassment when errors surface.",
      "job_type": "emotional",
      "supporting_quotes": ["It's embarrassing", "My CEO asks why we're not automated yet"],
      "pains_triggered": ["Loss of credibility with leadership", "Stress from manual oversight burden"],
      "frequency": "mentioned_multiple",
      "emotional_intensity": "high",
      "segment_signal": "RevOps leader reporting to CFO/CEO"
    },
    {
      "job_statement": "Functional: Free up time to shift from operational work to strategic analysis (churn, retention).",
      "job_type": "functional",
      "supporting_quotes": ["I'd have time to actually analyze churn patterns instead of babysitting data"],
      "pains_triggered": ["Inability to focus on high-impact work", "Reactive vs. proactive posture"],
      "frequency": "mentioned_once",
      "emotional_intensity": "medium",
      "segment_signal": "RevOps leader with growth mandate"
    }
  ],
  "top_3_jobs_by_priority": [
    "Eliminate manual data reconciliation to prevent billing errors",
    "Restore confidence in data accuracy and reduce leadership friction",
    "Free up time for strategic revenue analysis"
  ],
  "red_flags": [
    "Customer mentions Salesforce + billing system but didn't mention their actual data warehouse—may be using a third system we should ask about.",
    "'My CEO asks why we're not automated' suggests this is a known problem—test whether they've tried other tools and why they didn't stick."
  ],
  "next_questions": [
    "Have you tried any automation tools before? What didn't work?",
    "How often does a billing error actually make it to the customer vs. caught internally?",
    "If we solved the sync problem, would you measure success by time saved or by error reduction?"
  ]
}

Customization tips

Add {{COMPETITOR_CONTEXT}} if you want the prompt to flag jobs your competitors are already solving. This helps you spot white space or confirm you're on the right track.
For early-stage founders with 3–4 interviews, run this on each transcript separately, then manually merge the jobs list to see which jobs appear across multiple customers (high signal).
If your interviews are messy or conversational, prepend the transcript with a 1-sentence summary of the customer's role and company size. The prompt will filter noise better.
Use the red_flags output to build your next interview guide. If the prompt flags a contradiction, ask that question directly in the next call.
For positioning work, export the top_3_jobs and use them as your messaging pillars. Each job becomes a headline; the pains become proof points.

Featured

beginner

Claude or ChatGPT (verify current model choice)

v1.0.0