User Segmentation for Research Template

Start with what you have. CRM data gives you company size, industry, and plan tier. Billing data gives you revenue per account. Signup surveys give you role and primary use case. Even basic data lets you define 2-3 segments. As you add analytics, refine your segments with behavioral criteria. The [Product Analytics Handbook](/analytics-guide) covers instrumentation for tracking the behavioral data you need.

Segments and personas serve different purposes. Segments are data-driven groupings used for research recruitment and analysis. Personas are narrative characters used for design empathy. You can create personas from your segments, but the segment definition should come first. A persona without a segment is fiction. A segment without a persona is still useful for research.

Some overlap is normal. A user can be both a "Power Collaborator" and a "Team Admin." When this happens, assign users to their primary segment based on the dominant behavior. If overlap exceeds 30%, your segmentation criteria are not distinct enough. Add a behavioral dimension that separates the groups more clearly.

For qualitative research (interviews, usability tests): 5-8 per segment. For quantitative research (surveys, A/B tests): 100+ per segment for statistical significance. If you are running discovery interviews across 2 segments, plan for 10-16 total interviews. See the [Product Discovery Handbook](/discovery-guide) for detailed sample size guidance.

Re-evaluate your segments every 6-12 months or when you make a major product change (new pricing tier, new feature category, new user type). If your product adds a free tier, you likely need a new segment definition. If churn patterns shift significantly, your "at-risk" segment definition may need updating. ---