Claude-optimized prompt structure shaped for chatgpt.
# Task
Generate discovery questions that uncover real priorities, not just surface needs.
# Inputs- Buyer role: {buyer_role}- Company context: {company_context}- Likely pains: {likely_pains}- Sales stage: {sales_stage}- Desired outcome: {desired_outcome}- Constraints: {constraints}# Question methodology (SPIN sequence)
Organize questions in this order to build a natural conversation:
1. **Situation** (2-3 questions): Understand their current state. Keep these brief — don't make the buyer feel interrogated.
2. **Problem** (3-4 questions): Surface specific pains. Go beyond "what challenges do you face?" — ask about impact, frequency, and workarounds.
3. **Implication** (3-4 questions): Expand the problem. What happens if they don't fix it? Who else is affected? What's the cost of inaction?
4. **Need-payoff** (2-3 questions): Guide them to articulate the value of a solution. Let THEM describe the ideal outcome.
# Per question- The question itself
- **Why it matters** (1 sentence: what insight this unlocks)
- **Red flag answer**: What response indicates a bad fit, a stall, or a hidden blocker. Format: "If they say [X], probe [Y]."
- **Follow-up probe**: A natural next question if the answer is interesting.
# Rules- No generic questions ("What keeps you up at night?", "What's your biggest challenge?").
- Match question sophistication to the buyer's seniority and the sales stage.
- If early stage: more Situation and Problem. If late stage: more Implication and Need-payoff.
- Include 1-2 "permission questions" that earn the right to go deeper ("Would it be helpful if I asked about...?").
# Output1. 12-15 questions in SPIN sequence (with per-question details)
2. Recommended conversation opener (2-3 sentences to set the tone)
3. 3 questions to AVOID and why
claudethinking_medium
claude variant
Claude-optimized prompt structure shaped for claude.
<context><buyer_role>{buyer_role}</buyer_role><company_context>{company_context}</company_context><likely_pains>{likely_pains}</likely_pains><sales_stage>{sales_stage}</sales_stage><desired_outcome>{desired_outcome}</desired_outcome><constraints>{constraints}</constraints></context><task>Generate sequenced discovery questions using SPIN methodology.</task><instructions>
SPIN sequence: Situation (2-3) > Problem (3-4) > Implication (3-4) > Need-payoff (2-3).
Per question: the question, why it matters, red flag answer ("If they say X, probe Y"), follow-up probe.
No generic questions. Match seniority and stage. Include 1-2 permission questions. Early stage: more S/P. Late stage: more I/N.
</instructions><output>
Return: 12-15 questions in SPIN order with details, conversation opener (2-3 sentences), and 3 questions to avoid with reasons.
</output>
geminithinking_medium
gemini variant
Claude-optimized prompt structure shaped for gemini.
Buyer role: {buyer_role}
Company context: {company_context}
Likely pains: {likely_pains}
Sales stage: {sales_stage}
Desired outcome: {desired_outcome}Constraints:{constraints}
SPIN sequence: Situation (2-3) > Problem (3-4) > Implication (3-4) > Need-payoff (2-3).
Per question: the question, why it matters, red flag answer, follow-up probe. No generic questions. Match stage.
Based on the entire content above, return: 12-15 questions in SPIN order with details, conversation opener, and 3 questions to avoid.
Related templates
More templates in the same working category.
Email Reply
Draft a professional email reply that matches the thread tone, addresses every point raised, and drives toward a clear next step.