ChatGPT 5 Prompt for go-to-market sprint for a SaaS product

markdownCopyEdit# ROLE
Act as a senior B2B GTM strategist and data-driven copywriter with experience in <$20k ACV SaaS, PLG motion, and outbound testing.

# TASK
Design a 7-day GTM sprint for [PRODUCT], targeting [ICP/PERSONA] at [COMPANY SIZE / INDUSTRY]. Deliver a prioritized experiment plan, messaging, and ready-to-ship assets.

# CONTEXT
- Product: [1–2 lines on what it does + core outcomes]
- Pricing: [tiers], Free trial: [Y/N, length]
- ICP pain points: [bulleted]
- Competitors to avoid copying: [names]
- Voice/tone: [e.g., pragmatic, no hype]
- Constraints: budget [$X], channels allowed [email, LinkedIn, PPC, communities], assets available [case study Y/N, demo video Y/N]
- Success metric for the week: [e.g., 20 qualified demos booked or $5k MRR pipeline]

# REASONING INSTRUCTION
Think step-by-step:
1) Map ICP → outcomes → objections.
2) Propose 6–8 micro-experiments across 2–3 channels.
3) Score each by Impact (H/M/L), Confidence (H/M/L), Effort (hrs) and compute ICE = (I+C) – Effort.
4) Select the top 3 by ICE; justify in 1–2 sentences each.
5) Chain-of-verification: check each selected experiment against constraints, brand voice, and success metric; revise if misaligned.
6) Second pass: tighten copy using a 6-point rubric (clarity, specificity, proof, objection-handling, CTA strength, length).

# OUTPUT FORMAT
Return a concise Markdown report:
1. **Strategy Snapshot** (3 bullets: ICP outcome, primary channel, week goal)
2. **Experiment Table**

| Experiment | Channel | Audience slice | Offer/CTA | Steps | I | C | Effort(hrs) | ICE |
|---|---|---|---|---|---|---|---|---|

3. **Messaging Kit**  
   - 2 cold emails (≤120 words), 1 LinkedIn DM (≤80 words), 3 ad headlines (≤40 chars), 1 landing hero (≤12 words + subhead ≤20 words).  
4. **Day-by-Day Plan** (Mon–Sun: what to build, launch, measure)  
5. **Metrics & Guardrails** (what to track daily, pass/fail thresholds, when to kill or double-down)

# STOP CONDITIONS
- If any bracketed field is missing, ask exactly 5 crisp questions then stop.
- Keep the whole report under 900 words.
- If Confidence < “M” for any chosen experiment, flag it and suggest a safer alternative instead of proceeding.