Chapter 5
Score the evidence without fooling yourself
Turn score the evidence without fooling yourself into one evidence-backed research decision in 30–45 minutes, with a concrete output and a clear next action—without carrying a guess into the rest of the launch.
What you’ll get
- Primary decision: Use an evidence scorecard.
- Working asset: Interpret the pattern, not just the total.
- Quality check: Keep a contradiction log.
Before you begin
Bring the current offer hypothesis, the exact decision this research must change, and any existing evidence. Set a 30–45 minutes timer. Separate what you observed from what you inferred; record contradictions instead of averaging them away.
| Decision | Evidence you used | What could change it | Next action |
|---|---|---|---|
| Score the evidence without fooling yourself | Past behavior, workaround cost, commitment, or paid behavior | Stronger contrary evidence from a relevant participant | Complete the chapter and assign the next test |
Conversion checkpoint
Research improves conversion when it makes the offer more specific, more credible, or easier to act on. If a finding cannot change the audience, promise, price, proof, or CTA, it may be interesting but is not decision-critical.
Use an evidence scorecard
Score each dimension from 0 to 3. The score is a decision aid, not a statistical proof.
| Dimension | 0 — Missing | 1 — Weak | 2 — Useful | 3 — Strong |
|---|---|---|---|---|
| Problem consistency | No repeated problem | General frustration | Same problem in several relevant contexts | Repeated specific problem with clear triggers |
| Consequence | No meaningful cost | Mild annoyance | Material time, money, or outcome impact | Cost is visible and causes action now |
| Current behavior | No action | Consumes related free content | Uses a workaround or compares options | Already spends meaningful time or money |
| Audience clarity | “Everyone” | Broad creator category | Identifiable behavior-based segment | Segment is reachable and shares decision context |
| Offer clarity | People misinterpret it | Outcome understood after explanation | Most understand audience and result | People restate value accurately and ask buying questions |
| Commitment | Compliments only | Click, save, or basic waitlist | Qualified reply, call, or deposit | Completed purchase/preorder with clear terms |
| Delivery confidence | Undefined product | Format chosen, scope vague | Small version can be delivered | Fulfillment and support path tested |
Interpret the pattern, not just the total
A high total with a zero in delivery confidence means you may sell a promise you cannot fulfill. A high problem score with a low offer-clarity score means the opportunity may be good while the positioning is weak. A low commitment score tells you to improve and retest the offer—or stop—before building more.
Use these actions:
- Proceed: No critical zeros; problem, current behavior, clarity, and commitment reinforce one another.
- Narrow: Strong evidence appears in one segment or situation but not others.
- Repackage: The problem is strong, but the proposed format or bundle creates friction.
- Retest the message: The audience describes the problem consistently but misreads the page.
- Stop or park: The problem is not prioritized and people accept current alternatives.
Keep a contradiction log
Create a second table for evidence that challenges the idea:
| Assumption | Contradictory signal | Possible explanation | Next test |
|---|---|---|---|
| Editors want a large template library | Buyers ask for one finished proposal example | Choice creates setup work | Compare single workflow vs. large bundle |
| Faster proposals are the main value | Experienced editors focus on scope control | Different segment maturity | Separate pages by experience level |
The contradiction log protects you from turning only the positive comments into strategy.
Finish line
- The research decision is written in one sentence.
- Observations and interpretations are stored separately.
- Contradictory evidence remains visible.
- The next test has a stopping rule.
Build it in Tapmy
Turn the strongest observed language into a focused offer in Tapmy, use one proportionate CTA, and record behavior rather than collecting another abstract opinion.
