How to Run a Product Feedback Survey (With Example Questions)

Building a product without feedback is like driving with your eyes closed. You might be heading in the right direction, but you're probably not. A good product feedback survey tells you what users actually think, what they need, and what they'd pay for. Here's how to set one up.

Define What You Want to Learn

Before you write a single question, decide what you're trying to figure out. Are you validating a new feature idea? Trying to understand why people churn? Looking for what to build next? Each goal needs different questions. A survey that tries to answer everything at once ends up answering nothing well.

Questions for Feature Validation

When you're trying to decide whether to build a specific feature:

• "How do you currently handle [task the feature would address]?"
• "How important is [this capability] to your workflow?" (not at all, nice to have, important, critical)
• "If we built [feature], how likely would you be to use it?" (1-5 scale)
• "What would this feature need to do for it to be useful to you?" (open text)

Questions for General Satisfaction

When you want a pulse check on how users feel about your product overall:

• "Overall, how satisfied are you with [product]?" (1-5 scale)
• "What's the one thing you'd change about [product]?" (open text)
• "What feature do you use the most?" (multiple choice)
• "How would you feel if you could no longer use [product]?" (very disappointed, somewhat disappointed, not disappointed)
• "Is there anything we should stop doing?" (open text)

Questions for Understanding Churn

When someone cancels or downgrades, a short exit survey gives you critical information:

• "What's the main reason you're leaving?" (multiple choice: too expensive, missing features, found an alternative, no longer need it, other)
• "What could we have done differently?" (open text)
• "Would you consider coming back if we improved [their main reason]?" (yes/no/maybe)

Structuring the Survey

Start with the easiest questions (ratings, multiple choice) and end with open-ended ones. People are more likely to abandon on open-ended questions, so put the most important data collection up front. Use a conversational layout so each question gets full attention. Cap it at 5 to 7 questions total.

Who to Send It To

Don't blast every user with the same survey. Segment your audience. Active users can tell you about satisfaction and feature priorities. Churned users can tell you what went wrong. New users can tell you about onboarding friction. Send the right survey to the right group and you'll get much more actionable insights.

Acting on the Results

The survey is pointless if you don't act on it. Review results within a week of closing the survey. Group similar feedback into themes. Share the top findings with your team. Pick the highest-impact items and commit to addressing them. Then close the loop by telling users what changed because of their feedback.