Most business documents fail for the same reason: they present information instead of driving a decision. Claude can help you write documents that get people to act — if you prompt it with the decision in mind, not just the content.
The Decision Document Framework
Every business document exists to move someone from uncertainty to action. Before you prompt Claude, answer three questions:
- Who decides? — The specific person or group making the call
- What do they need to believe? — The key insight or argument
- What should they do next? — The concrete action you want
The Amazon-Style Memo
Jeff Bezos banned PowerPoint for a reason. Narrative memos force clear thinking. Claude is excellent at this format because it requires structured reasoning, not bullet-point decoration.
Write a narrative memo (Amazon 6-pager style) for [audience] about [topic]. The memo should argue for [recommendation].\n\nStructure:\n1. Opening — state the core question in one paragraph\n2. Background — what led us here (keep it tight)\n3. The problem — what's broken or at risk\n4. Proposed approach — what we should do and why\n5. Key tradeoffs — what we're giving up and why it's worth it\n6. Next steps — specific actions with owners and dates\n\nTone: clear, direct, data-informed. No corporate fluff. Write in prose paragraphs, not bullet points. Target: 1,500-2,000 words.
Pro Tip
The best memos start with the conclusion. Don't make your reader hunt for the point. Tell them what you think in paragraph one, then spend the rest of the document proving it.
Proposals That Win
Proposals fail when they describe what you do instead of what the client gets. Claude can flip that framing if you give it the right inputs.
Write a proposal for [client/company] for [project/engagement]. Structure:\n\n1. **The Situation** — reflect back what the client told us about their problem (show we listened)\n2. **What's at Stake** — quantify the cost of inaction\n3. **Our Approach** — what we'll do, in phases, with clear deliverables\n4. **Timeline & Investment** — dates, milestones, pricing\n5. **Why Us** — 2-3 differentiators (proof, not claims)\n6. **Next Step** — one clear action to move forward\n\nClient context: [what they told you in the discovery call]. Budget range: [range]. Timeline expectation: [timeline]. Tone: confident, specific, zero fluff. Under 1,500 words.
The One-Pager
When you need to compress a complex idea into a single page — for investors, board members, or busy executives — every word has to earn its place.
Write a one-pager for [audience] about [topic]. This should fit on a single printed page (~500 words max).\n\nStructure:\n- **Headline**: Bold statement that captures the core idea\n- **The Problem** (2-3 sentences): What's broken or what opportunity exists\n- **The Solution** (3-4 sentences): What we're proposing\n- **Key Metrics** (3-5 bullets): Numbers that prove this works or project impact\n- **Ask** (1-2 sentences): What we need from the reader\n\nAvoid jargon. Use concrete numbers. Every sentence should pass the 'so what?' test.
The Strategic Brief
For cross-functional projects where multiple teams need to align on scope, goals, and boundaries.
Write a strategic brief for [project name]. This will be shared with [teams involved] to align on what we're building and why.\n\nInclude:\n1. **Objective** — One sentence. What does success look like?\n2. **Background** — Why now? What triggered this? (3-4 sentences)\n3. **Scope** — What's in and what's explicitly out\n4. **Success metrics** — How we'll measure this worked (3-5 metrics)\n5. **Key risks** — What could go wrong and how we'll mitigate\n6. **Timeline** — Major milestones and dates\n7. **Owners** — Who is responsible for what\n8. **Open questions** — What we still need to figure out\n\nTone: clear, practical. This should be a working document, not a performance.
The Board Update
Board updates are a specific art. Too long and nobody reads them. Too short and the board asks questions you should have preempted.
Write a monthly board update for [company]. Period: [month/quarter].\n\nStructure:\n1. **TL;DR** — 3 bullets: one win, one challenge, one decision needed\n2. **Key Metrics** — Table format: metric, last period, this period, target, trend\n3. **Wins** — 2-3 highlights with specific numbers\n4. **Challenges** — 1-2 honest assessments of what's not working\n5. **Strategic Decisions** — Any decisions made this period and rationale\n6. **Ask** — What you need from the board (feedback, introductions, approval)\n7. **Next 30 Days** — Top 3 priorities\n\nData: [paste key metrics]. Tone: transparent, confident. Show you're in control even when sharing bad news.
The Internal RFC (Request for Comments)
When you need buy-in from peers before making a significant change.
Write an RFC (Request for Comments) for [proposed change]. This will be shared with [audience] for feedback before we commit.\n\nStructure:\n1. **Summary** — What are we proposing, in plain language?\n2. **Motivation** — Why should we do this? What problem does it solve?\n3. **Detailed Design** — How would this work, specifically?\n4. **Alternatives Considered** — What else did we look at and why did we rule it out?\n5. **Migration / Rollout Plan** — How do we get from here to there?\n6. **Risks and Open Questions** — What could go wrong? What do we not know yet?\n\nTone: thoughtful, inviting feedback. This isn't a decree — it's a proposal. Make it easy for people to disagree constructively.
Scenario
You're presenting to your board next week and need to explain why revenue is down 15% while simultaneously asking for approval to invest $300K in a new product line.
Iteration Is the Point
The first draft Claude gives you is a starting point, not a finished product. The real value comes from iterating.
Generate the first draft
Give Claude your full context, audience, and desired structure. Let it produce a complete first pass.
Challenge the framing
Ask: 'What's the strongest counterargument to this memo? Rewrite the opening to preempt it.'
Tighten the language
Ask: 'Cut this by 30% without losing any substance. Remove every sentence that doesn't directly support the recommendation.'
Stress-test with questions
Ask: 'You're [the skeptical board member / the CFO / the engineering lead]. What questions would you ask after reading this? Add a section that addresses them.'
Final polish
Ask: 'Read this as someone who has 5 minutes and a strong bias toward inaction. Does this create enough urgency to act? If not, strengthen it.'
Pro Tip
The best technique for improving any document: tell Claude to role-play as the toughest person in the room and poke holes in your argument. Then rewrite to address those holes before anyone else finds them.