Skip to content
Template · used in Improve

Business test decision record

Link the decision to evidence, counter-evidence and the next question.
PreviewCT15

Record

Test question
From the validation plan
Rule set before start
Continue / change / stop / inconclusive
What actually happened
With numerators and denominators
Supporting evidence
Your answer
Counter-evidence
Your answer
Alternative explanations
Your answer
Decision
Continue / change / stop / inconclusive
Next task
One concrete action and date

When to use this record

Fill in the decision record when a test period ends. It links the decision to the evidence, forces you to write down what argues against it, and ends with one concrete next task. Kept together, your decision records become the history of how the business was actually built.

How to fill it in

  1. Copy the test question and the rule from the validation plan, unchanged.
  2. Write what actually happened, with numerators and denominators: "10 contacted · 3 replied · 1 quote request".
  3. List the supporting evidence with its level.
  4. List the counter-evidence — even when the result looks good.
  5. Write alternative explanations for the result.
  6. Record the decision: continue, change, stop or inconclusive.
  7. Write one next task with a date.

A filled-in example

FieldExample
Test questionWill small UK shops request a paid audit after seeing a sample?
RuleContinue if ≥ 2 quote requests; change if ≥ 3 replies and 0 requests
What happened10 contacted · 3 replied · 1 meeting · 1 quote request · 0 payments
Supporting evidenceQuote request for 400 rows (original attached)
Counter-evidenceOne shop says its platform already handles this
Alternative explanationThe request may be tied to a one-off promotion
DecisionChange
Next taskOffer a 400-row tier to 10 shops with 300+ products, by 20 Nov

Inconclusive is a valid result

If too few relevant people were reached, record "inconclusive" and write what would make the next test conclusive — usually a better channel or more relevant candidates. Calling a weak test a "no" kills good ideas; calling it a "yes" wastes time on bad ones.

More templates

Your first real test can start this week.

Free to start, no card. See the models that fit you, open a ready-to-run kit and plan your first 10-day test.

  • We never send messages, publish or spend money for you.
  • We don’t invent experience, clients or results.
  • Every claim is labelled: fact, estimate or hypothesis.