Customer Feedback Survey

Surveys Fail When They Ask What People Think Instead of What They Do

The customer feedback survey is one of the most overused and underperforming tools in product management. Not because surveys are inherently broken, but because most surveys ask the wrong questions to the wrong people at the wrong time.

The Problem

The default instinct is to ask customers what they want, what they like, or how they feel. These opinion-based questions produce data that feels actionable but is not. Customers are notoriously bad at predicting their own behavior. They say they want features they will never use. They report satisfaction with products they are about to churn from.

A 2023 Qualtrics study found that 80% of customers who churned had reported being "satisfied" or "very satisfied" in their most recent survey. Satisfaction scores are lagging indicators dressed up as leading ones.

The real problem is even more fundamental. Most surveys treat customers as focus group participants when they should be treated as witnesses to their own behavior. The question is not "What do you think of our product?" but "What did you do the last time you encountered this problem?"

How This Prompt Works

This prompt generates behavior-focused survey questions that uncover what customers actually do, not what they claim to prefer. It follows the "Mom Test" principle from Rob Fitzpatrick's methodology: questions should be constructed so that even your mother cannot lie to you.

The prompt produces three categories of questions:

• Behavioral questions that ask about specific past actions and decisions
• Context questions that reveal the circumstances surrounding product usage
• Outcome questions that measure whether the product delivered the result the customer needed

Each question is paired with a rationale explaining what signal it captures and how to interpret responses at scale.

According to SurveyMonkey's 2023 benchmarks, surveys with behavioral questions achieve 34% higher completion rates and produce responses that correlate 2.7 times more strongly with actual user behavior in analytics data.

When to Use It

• Before a major product decision when you need quantitative validation of qualitative hypotheses
• After shipping a feature to measure whether it changed behavior, not just whether people noticed it
• During churn analysis to understand the sequence of events that led to departure, not just the stated reason

Common Pitfalls

• Survey fatigue. Every survey you send reduces response rates for future ones. Make each one count. According to a 2022 Medallia report, the average survey response rate has dropped from 33% in 2016 to 18% in 2022 as customers become oversaturated.
• Leading questions. "How much do you love our new feature?" is not research. It is self-congratulation disguised as methodology.
• Ignoring non-respondents. The people who do not answer your survey are often the ones whose feedback matters most. They are too busy, too disengaged, or too frustrated to participate.
• Averaging away insight. A Net Promoter Score of 7 could mean everyone is mildly satisfied or that you have equal numbers of passionate advocates and harsh critics. Distributions matter more than averages.

Sources

• Fitzpatrick, R. (2013). *The Mom Test*. https://www.momtestbook.com/
• Qualtrics. (2023). Global Consumer Trends Report. https://www.qualtrics.com/ebooks-guides/global-consumer-trends/
• SurveyMonkey. (2023). Survey Benchmarks Report. https://www.surveymonkey.com/curiosity/survey-benchmarks/