Value Proposition Canvas (VPC)

Credits: Alexander Osterwalder, Yves Pigneur, Greg Bernarda, Alan Smith — Value Proposition Design (strategyzer.com; the canvas and first chapter are free there). Ranking discipline and hypothesis framing — Steve Blank.

The canvas has two sides. Fit is the claim; evidence is the test.

Right side — Customer Profile (about THEM, drafted from evidence)

• Customer jobs — functional, social, and emotional jobs they're trying to get done (import from /jtbd if run). Their words, their situation, not your solution.
• Pains — bad outcomes, obstacles, and risks around those jobs: what costs too much time/money, what feels risky, what keeps failing.
• Gains — outcomes and benefits they want: required, expected, desired, and unexpected gains.

Rank pains and gains by intensity — with customers where possible. AI/founder rankings are placeholders: label them [ASSUMPTION — rank-test in interviews]. Interview evidence gets quoted; everything else is labeled.

Left side — Value Map (about YOU, drafted as hypotheses)

• Products & services — what you offer (the list, not the pitch).
• Pain relievers — exactly which top-ranked pains each element eliminates or reduces. Not all pains — the ones that matter.
• Gain creators — exactly which top-ranked gains each element produces.

Fit — the honest test

Fit exists when your top pain relievers and gain creators map to the customer's top-ranked pains and gains — as ranked by customers, not by you. Three levels: on paper (problem-solution fit, all assumption), in market (customers respond — early evidence), at scale (they pay repeatedly — product-market fit). State plainly which level the canvas is at.

Common failure modes to flag:

• Feature-first fit: value map written first, customer profile reverse-engineered to justify it. (Tell: every feature magically maps to a "top" pain.)
• Pain inflation: mild inconveniences promoted to burning pains without evidence. If a launch or landing page already got silence, say what that does/doesn't imply.
• Unranked everything: 15 pains all "important" = nothing ranked = nothing testable.

MVP linkage

The MVP is the smallest thing that tests the riskiest fit claim — usually top pain ↔ top pain reliever. A concierge/manual version of the service is often stronger evidence than software (real transaction, zero build). Specify: the claim under test, the MVP form, the pass/fail signal, and what you deliberately did NOT build.

Output

Both sides of the canvas in markdown, every entry labeled [FACT — quote/source] / [ASSUMPTION] / [UNKNOWN], pains/gains ranked with ranking provenance stated, the fit claim + current fit level, the MVP proposal, and a "Validate next" list routed to /customer-discovery. Get the printable canvas from Strategyzer for workshops.

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.