PRODUCT DISCOVERY DECISION LOG
Original ShowMeYourSaaS teaching resource | 2026-10-27
All ReviewLoop names and observations below are fictional.

COMPLETED EXAMPLE
Product: ReviewLoop, an existing app for agency draft reviews.
Decision owner: founder.
Outcome: a new agency account completes a client-submitted review within 14 days.
Unit and cohort: new agency accounts; do not substitute individual users.
Guardrails: invitation failures, preparation effort, support burden.

Observation: two people in six fictional onboarding conversations describe clients
asking which part of a draft requires a decision. This is a qualitative signal,
not a population estimate.
Interpretation: unclear instructions may be a barrier to a first completed review.
Opportunity: help the invited client understand the requested decision.
Options: a decision prompt; an annotated draft; a reminder.
Chosen test candidate: an optional one-sentence decision prompt.
Riskiest assumption: a client can explain the requested decision without help.
Other assumptions: owners can write the prompt; it does not block invitations.

Test: prepare three synthetic invitations. Observe an owner write the prompt and
a permitted participant interpret it. Ask them to show how they would respond.
Do not explain the intended answer before observing.
Decision rule: revise after any misunderstanding; assess owner effort separately.
Illustrative result: one participant reads an approval request as an editing task.
Decision: revise the wording and retest before committing to implementation.
Not established: live adoption, delivery reliability, or conversion uplift.
Next review: choose an owner and date before the next session.

BLANK RECORD
Date / owner:
Product and current customer context:
Desired customer outcome:
Measurement unit / cohort / time window:
Guardrails:
Observed situation and evidence source:
Permission boundary for customer material:
Interpretation, separated from the observation:
Opportunity:
Candidate approaches:
Riskiest consequential assumption:
Why a smaller test can answer this question:
Test inputs and method:
Decision rule written before the test:
Observed result, including intervention or failure:
Decision: discard / revise / investigate / deliver:
What remains unknown:
Next action / owner / date:
Link to the relevant roadmap decision:

REVIEW CHECK
Could the evidence contradict the preferred idea?
Does the test answer the named assumption?
Are small qualitative observations presented as qualitative observations?
Is a test result being enlarged into an unsupported market or causal claim?
Can the delivery owner explain why the decision changed?