Evidence must match the next decision

Weak directional evidence may justify five more interviews. A three-month build, hire or advertising budget requires stronger action, transaction or adoption evidence. The goal is not certainty; it is keeping investment risk proportional to evidence strength.

The WOW51 E0–E3 evidence ladder

  • E0 — Untested hypothesis: AI inference, internal belief or an assumed customer profile.
  • E1 — Attention and statement: visits, saves, surveys or interview claims.
  • E2 — Action: a booking, submitted information, completed trial or decision-maker introduction.
  • E3 — Transaction and adoption: deposit, payment, valid order, real use or repeat purchase.

Four common evidence errors

  • People liked the post, therefore demand exists.
  • Interviewees said they would buy, therefore the price works.
  • One person paid, therefore the market is repeatable.
  • AI said the idea is viable, therefore investment should increase.

A practical evidence record

  • Original hypothesis: who buys in which situation?
  • Observed fact: record behavior without optimistic interpretation.
  • Evidence level: assign E0, E1, E2 or E3.
  • Counterevidence: record refusals, churn, refunds and alternatives.
  • Next test: run one experiment that can raise or overturn the level.

Continue, change or stop

  • Continue when the same buying situation repeats and produces payment or adoption.
  • Change when people act but consistently stop at scope, price, trust or payment.
  • Stop when qualified outreach produces no action across repeated tests, or viable pricing stays below delivery cost.