Read the result in the context of the question, audience, sample, and method. This guide explains the useful signal from each survey type and the checks to make before you act.

What You Will Be Able to Explain

How to read results from all six survey types
What signals to look for (and what to ignore)
When the result is strong enough to guide a decision and when it needs another check

Treat a survey as evidence, not an automatic verdict. The cost and reversibility of the decision determine how much evidence you need.

How to Read Each Survey Type

FastPoll

FastPoll shows how answers are distributed across one set of options. Read the size of the lead together with the response count.

A clear lead — One option is meaningfully ahead and stays ahead as more responses arrive.
A close split — The audience may be divided, the options may be too similar, or the sample may still be small.
An option with little support — Check that its wording was clear before you remove it from consideration.

See the FastPoll referencefor details on response data format.

UserChoice

UserChoice uses Conjoint Analysis to reveal which product trade-offs matter most. Look for patterns across repeated choices, not one isolated comparison.

Preference strength — A wider, stable difference is more informative than a narrow one.
Consistency across comparisons — Cyclical or unstable choices can mean that preferences depend on context or that the options need review.
Segment differences — Compare meaningful groups only when each group has enough responses to support the comparison.

See the UserChoice referencefor details on preference data.

PricePoint

PricePoint can use Van Westendorp to estimate perceived price boundaries or Gabor-Granger to test demand at stated prices. For a Van Westendorp result, start with:

Too-cheap boundary — The lower end of the range where price may begin to raise quality concerns.
Too-expensive boundary — The upper end where price resistance becomes important.
Acceptable price range — A research input to test against costs, positioning, segments, and observed buying behaviour; it is not an automatic final price.

See the PricePoint referencefor details on pricing analysis output. For deeper methodology context, read the Van Westendorp pricing guide.

FeaturePriority

FeaturePriority supports Kano, MaxDiff, and Bradley-Terry. Read the output according to the selected method rather than forcing every result into the same categories.

Kano — Separate basic expectations, performance needs, and delighters using the method's classifications.
MaxDiff — Compare relative importance across the feature set and pay attention to stable gaps, not just rank order.
Bradley-Terry — Use pairwise outcomes to estimate relative preference between features.
Differences between groups — A feature can matter to one audience and not another. Check the sample before treating that split as a segment strategy.

See the FeaturePriority referencefor ranking data details. For Kano methodology context, read theKano model guide.

OpenFeedback

OpenFeedback captures explanations in people's own words. Read the comments for recurring themes, specific evidence, and useful exceptions.

Recurring themes — Group comments that describe the same need or problem, while keeping important differences visible.
Context — Note where, when, and for whom the issue happens. Frequency alone does not tell you its impact.
Specific evidence — Concrete descriptions are more useful than vague praise or frustration. Treat requested solutions as clues to the underlying need.

See the OpenFeedback referencefor response data format.

Reaction

Reaction captures a lightweight signal with expressive faces, thumbs up/down, or a heart. It can show where to investigate, but it does not explain why someone reacted that way.

Mostly positive — The item appears to work for the people who responded. Check whether that audience matches your goal.
Mixed reactions — Look at context or pair the signal with OpenFeedback before deciding what to change.
Mostly negative — Investigate the cause and test a revision; the icon alone does not identify the problem.

See the Reaction referencefor reaction data format.

Checks to Make Before You Decide

Match evidence to the risk. A reversible copy change can need less evidence than a pricing or roadmap commitment.
Look for patterns without hiding exceptions. Repetition matters, but one severe accessibility, security, or data-loss issue can still deserve immediate attention.
Combine quantitative and qualitative. Numbers tell you what's happening. Open feedback tells you why.
Consider timing. A launch, outage, promotion, or product change can shape the responses. Compare another period when that context matters.
Share with your team. Different people notice different patterns. A designer might spot UX issues in feedback that a PM would miss.

Interpretation Basics