Business Model Canvas (BMC) + Market Type + Unit Economics

Credits: Alexander Osterwalder & Yves Pigneur — Business Model Generation (strategyzer.com, canvas free). Market types, Get/Keep/Grow, customer development — Steve Blank & Bob Dorf (The Startup Owner's Manual; free course: How to Build a Startup).

Blank's framing: a startup is a temporary organization searching for a repeatable, scalable business model. Every block on the canvas is a hypothesis to test, not a plan to execute. Fill it in pencil.

The 9 blocks (each entry labeled F/A/U)

Value Propositions (import from /value-proposition-canvas) · Customer Segments (import from /jtbd + /where-to-play) · Channels · Customer Relationships · Revenue Streams · Key Resources · Key Activities · Key Partners · Cost Structure.

Rules: revenue and cost numbers need sources or [UNKNOWN + the experiment that measures it]. Identify the riskiest block (early on: usually customer segment or revenue) and propose the cheapest test for it.

Market Type — decides everything downstream (Blank)

Existing (better product, known competitors) · Resegmented (niche or low-cost carve-out) · New (no existing users — education costs dominate, adoption is slow, time-to-profitability is long) · Clone (proven model, new geography). Classify the venture, state the implications for discovery, pricing, and burn — and label the classification [ASSUMPTION] until customer behavior confirms it. Misreading a new market as an existing one is a classic startup killer.

Customer Relationships — Get / Keep / Grow

• Get: paid vs earned demand creation; for physical channels, awareness → interest → consideration → purchase; for web, acquisition funnels and viral loops. A launch without a Get plan isn't a launch — it's a coin toss with one flip. Word-of-mouth chains, referral asks, and existing-customer bases are Get channels too (often the cheapest — instrument them).
• Keep: loyalty, service quality, switching costs.
• Grow: upsell, cross-sell, referrals-by-design.

Unit economics

• CAC (per channel — only measurable from real experiments) and LTV (repeat frequency × margin × lifetime). Viability requires LTV meaningfully above CAC; both are [UNKNOWN] until measured — never estimate them unlabeled.
• Service businesses: cost-to-serve per order (travel, time, delivery) vs order value — instrument the first deliveries (km, minutes, order value) and check overhead % before scaling or hiring.
• Pricing: anchor to what customers actually paid for alternatives (behavioral), not survey answers. Named, visible surcharges beat silently-bundled costs — you learn willingness to pay only when the price is visible.

Output

The 9-block canvas in markdown (every entry labeled), market-type classification + implications, a Get/Keep/Grow plan with per-channel tracking, the riskiest-block test proposal, and a "Validate next" list. Route interview questions to /customer-discovery and desk research to /secondary-research.

Ledger discipline (v2): reference Belief Ledger IDs ([A4], [F7]) for every claim; propose new/changed items in a closing "Ledger delta" section rather than restating claims — restated claims drift.

Fill-in templates: TEMPLATES.md in this skill's folder — copy the structures into your project's discovery/ and fill with labels.