Transform a refined idea and market intelligence into a rigorous business model that will actually sustain a company. This phase takes the subjective promise of Phase 2 and builds objective, testable business assumptions.
Workflow
1. Input Requirements
Before starting, you must have:
REFINED-IDEA.md — The validated problem, solution, and core insight
If these don't exist, ask the user to complete Phases 1 and 2 first.
2. The Nine-Cell Lean Canvas
The Lean Canvas is not a creative exercise—every cell must be challenged and grounded in market reality. Build it section by section:
Problem
Extract the top 3 problems from REFINED-IDEA.md. Do not invent new problems.
For each problem, cite market evidence from MARKET-INTEL.md
Be specific: "Hard to find X" is not a problem; "Managers spend 4+ hours/week on manual X with 87% error rate" is a problem
Skills relacionados
Order by severity (impact × frequency)
Call out which customer segments experience which problems most acutely
Solution
Define top 3 features (not solution details, actual deliverables users interact with).
Map each feature directly to one of the top 3 problems
Be specific: "Slack integration" is not a feature; "One-click bulk import of Slack messages into decision log" is a feature
Note the minimum viable form of each feature (can we launch with 60% of the full vision?)
Challenge: "What if we launched with only the #1 problem solved?"
Key Metrics
Name THE ONE metric that proves product-market fit at this stage—not the eventual north star, but the signal you'll track obsessively in months 1-3.
Example for B2B: "Monthly active users" (not DAU, not revenue, not seats)
Example for marketplace: "Completed transactions per week" (not listings, not sign-ups)
Example for SaaS: "Churn rate by cohort" (not total users, not gross revenue)
Call out vanity metrics to avoid: growth that doesn't translate to revenue, free users, referral spam
Define: What's the success threshold at 3 months? Be concrete: "50 MAU" not "high engagement"
Unfair Advantage
What will genuinely prevent competitors from replicating your success? Be brutal.
"First mover advantage" is not an advantage (competitors catch up)
"Network effects" only count if you have critical mass (0 users = 0 network)
"Founder expertise" is weak; "founding team previously scaled a $50M+ business in this space" is stronger
Real advantages: exclusive partnerships, proprietary data, unique technical capability, regulatory moat, brand loyalty (only if earned)
If you can't name 1-2 real advantages, that's a red flag—note it
Channels
Name 3-4 specific acquisition channels with concrete reasoning.
Not "social media" or "marketing"—be specific: "Twitter outreach to indie hackers," "ProductHunt launch," "Direct outreach to 50 target companies," "SEO for 'X + Y' keywords"
For each channel, estimate: early-stage CAC, realistic monthly volume, and why it fits your customer segment
Identify which channels are fastest to validate (usually 1-2 can start in week 1)
Note: cost-per-customer, not total cost
Customer Segments
Describe the early adopter profile in extreme detail (not just "SMBs" or "freelancers").
Age range / career level / company size / revenue / pain point intensity
Where do they hang out? (Slack communities, forums, subreddits, conferences)
What's their budget for solving this problem? (This informs pricing strategy)
How do they currently solve it? (Manual workarounds count)
Early adopter quirk: they're willing to accept a rough product if it saves them time/money. Who are these people?
Cost Structure
Build an honest monthly burn estimate.
Infrastructure: AWS/Vercel/Heroku/hosting costs
Tools & services: payment processing, analytics, CRM, design tools, etc. (itemize all subscriptions)
People (if any): full salaries or contractor costs for developers, designers, marketers
Founder opportunity cost: What's the market rate for your role? If you'd earn $150K as an engineer, that's $12.5K/month cost
Other: legal, accounting, customer support time
Be conservative; assume 50% of estimates as "surprise costs"
Monthly total: This is your real runway clock
Revenue Streams
How does money actually flow into the business?
Subscription (SaaS): monthly, annual, or both? Tiered pricing?
Transaction fee (marketplace, payments): % of transaction or flat fee?
Freemium: what's the conversion rate assumption from free to paid?
One-time: licensing, setup fee, or upfront purchase?
B2B services: retainer, hourly, project-based?
Be clear: if you're counting on premium features or upsell, when does that realistically happen?
Unique Value Proposition (UVP)
Write one sentence a customer would repeat to a friend.
Not: "The fastest, easiest, most affordable solution"
Yes: "Slack-native task manager that reduces context switching by 80%"
Yes: "Marketplace for short-term rentals where hosts get pre-vetted guests in 24 hours"
The UVP should reflect the problem, the segment, and the benefit—all in one breath
3. Unit Economics
Unit economics define whether your business model works on a per-customer basis.
Customer Acquisition Cost (CAC)
Estimate per-channel CAC: How much do you spend (time + money) to acquire one paying customer?
If you're doing direct outreach to 100 companies and close 1, and that takes you 40 hours at your opportunity cost, CAC = (40 × $150/hr) / 1 = $6,000
If you spend $500/month on ads and land 10 customers, CAC = $500/10 = $50
Range likely spans from $50 to $5,000+ depending on channel and segment
Rank channels by CAC efficiency
Lifetime Value (LTV)
Average monthly revenue per customer (ARPU): How much do they pay per month?
Gross margin: What % of revenue is profit after direct costs? (hosting, payment processing, cost of goods)
Customer lifespan (months): How long does a customer stay? For SaaS, assume 12-24 months initially; for marketplaces, shorter
Product risk: "Technical complexity is 3x worse than estimated; MVP takes 6 months"
Example Risk Register:
1. Market Acceptance — Busy managers won't adopt yet another tool
- Likelihood: High (crowded space, switching costs high)
- Impact: Critical (no customers = no revenue)
- Priority: 1 (top)
- Mitigation: Get 5 early adopter commitments in writing (pre-sales) before building
2. CAC Too High — Social media ads cost $500+ per customer in target segment
- Likelihood: Medium (ads expensive, but direct outreach might work)
- Impact: Major (LTV:CAC < 3:1, not viable)
- Priority: 2
- Mitigation: Test direct outreach to 20 target managers before scaling paid ads
Output Format
Generate BUSINESS-MODEL.md with the following structure:
# Business Model: [Company Name]
## Lean Canvas
[9 cells, each 2-5 sentences, concrete and challenged]
## Unit Economics
**Customer Acquisition Cost (CAC):** [per channel with breakdown]
**Lifetime Value (LTV):** [with assumptions]
**LTV:CAC Ratio:** [current and target]
**Payback Period:** [months]
**Gross Margin Projection:** [month 1, 6, 12]
## Revenue Model
**Pricing Strategy:** [tiers/positioning vs. competitors]
**Revenue Projections:**
- Conservative: Month 1/3/6/12
- Optimistic: Month 1/3/6/12
**Break-Even Analysis:** [customers needed, months to break-even]
## Financial Projections
**Monthly Burn Rate:** [itemized, months 1-12]
**Runway:** [if funded, how long does it last?]
**Revenue Milestones:** [1K MRR, 10K MRR, break-even, customer milestones]
**Ramen Profitability Threshold:** [revenue needed]
## Risk Register
[Top 5 risks with likelihood, impact, priority, mitigation]
---
**Last Updated:** [date]
**Status:** [Draft / Validated / In Execution]
Tips for Success
Challenge every assumption: If you can't cite market evidence for an estimate, it's a guess. Flag it.
Be honest about weaknesses: "We don't know how to reach customers yet" is better than a made-up channel plan.
Use the Risk Register as your roadmap: Pick the top 3 risks and make them testable. That's what you validate next.
Revisit monthly: Business models are living documents. Update as you learn.
Default to boring, proven models: Subscription is easier than marketplace. B2B is more predictable than B2C. Solve this correctly before chasing innovation in monetization.