Common questions › Trusting the numbers
How many questions do you need to ask?
The current UK Pilot uses 20 fixed questions. That is a Pilot scope, not a final customer allowance or proof of a reliable baseline.
Relevance and completeness matter
Choose questions that reflect real buying decisions. Keep the exact wording and record where each question came from. More questions do not fix an unrepresentative sample, missing readings or incompatible settings.
Keep the ruler fixed
Compare unchanged questions under compatible conditions, separately by service. If the question basket changes, start a separate comparison. Repeated readings are not independent buyers.
Show the usable evidence
Publish the usable-answer denominator and failed observations separately. A failed reading is unknown, not absence. This Pilot has not established a complete accepted baseline. Customer allowances remain unconfirmed.
See the intended experienceExplore a static example of the intended experience. Public app access and client onboarding are not available.
Explore example report