When to Interview

Knowing when not to interview is as important as knowing how.

User interviews are expensive. Every hour spent recruiting, scheduling, and conducting conversations is an hour not spent building. The decision-first framework gives you a clear heuristic for when research is worth the cost.

The Reversibility Test

If a decision is:

• Reversible (you can undo it if it turns out wrong)
• Cheap to implement (days, not quarters)
• Backed by sufficient existing evidence (you already have analytics data or prior research)

...don't interview. Ship it and learn from real usage.

When Research Is Essential

Conversely, schedule interviews when:

• The decision involves significant engineering investment (multiple sprints or quarters)
• It affects a critical user journey (activation, checkout, core workflow)
• The team is operating on intuition alone: nobody has actually talked to users
• Existing data tells you what happens but not why

The "why" is what interviews uniquely provide. Analytics shows you a drop-off at step 3 of checkout. Interviews tell you users were surprised by the shipping cost they saw for the first time on that exact page.