THE PLAYBOOKS / EVIDENCE

Results record worksheet

Use this after a change has been live long enough to observe. Separate measured work from public case analysis and illustrative examples. A before-and-after change on its own does not prove causality.

01Evidence label

Measured work, public case analysis, or illustrative example.

02Change date and observation period

When did the change go live, and what window did you review?

03Before: numerator / denominator / rate

For example, 20 of 100 eligible signups reached first value (20%). Write “not measured” if absent.

04After: numerator / denominator / rate

Use the same event definition, eligibility, and window as the baseline.

05Absolute difference and relative change

Absolute difference in percentage points; relative change only when the baseline is nonzero. Example only: 20% to 30% is +10 points, +50% relative.

06What else changed?

Traffic source, audience, product, offer, pricing, or timing in the same period.

07Sample and measurement limitations

Small counts, tracking gaps, definition changes, seasonality.

08Delivery and acquisition costs; guardrail results

Time and money spent, plus errors, complaints, support load, or opt-outs.

09What we can conclude, and what remains unknown

Keep these two lists separate.

10Keep, revise, or stop, and who owns the next step

State the decision rule you agreed before the test and whether it was met.

11Permission to share

Confirm claims, anonymization, and client approval before publishing.