Intercept Survey Template for User Research

It depends on the metric. For satisfaction (CSAT), use a 1-5 scale with labeled endpoints ("Very Dissatisfied" to "Very Satisfied"). For effort (CES), use a 1-5 or 1-7 scale with labeled endpoints ("Very Difficult" to "Very Easy"). For loyalty ([NPS](/glossary/nps-net-promoter-score)), use the standard 0-10 scale. For quick binary feedback, thumbs up/down is effective. Consistency matters more than the specific scale: pick one and use it across all surveys so you can compare over time.

For a rating average, 100+ responses gives you a margin of error around +/- 0.2 on a 5-point scale. For segmented analysis (comparing free vs pro users), you need 50+ per segment. For open-ended themes, 30+ responses typically reveal the major themes. If your product has low traffic, run the survey longer rather than lowering the sample rate.

Three rules: (1) enforce a minimum gap between surveys per user (14 days is the standard), (2) set a sample rate below 100% for high-traffic touchpoints (25-50%), and (3) cap the total number of surveys any user sees per quarter (2-3 maximum). If you run multiple intercept surveys across different product areas, maintain a central registry so the frequency controls apply globally.

Generally no. Intercept surveys already achieve 40-60% response rates because they appear in context with minimal effort. Incentives introduce response bias (people answering for the reward, not because they have genuine feedback). If your response rate is below 30%, the problem is usually timing, targeting, or survey length, not lack of incentive. Fix the trigger and simplify the questions before adding incentives.

Yes, but with caveats. In-product NPS surveys tend to skew higher than email NPS surveys because they capture users who are actively engaged (by definition, they are using the product). This is fine as long as you compare in-product NPS to in-product NPS over time, not to industry benchmarks based on email surveys. Run the NPS intercept on a recurring schedule (quarterly) with consistent targeting to track trends. ---