Interview Script·45 min·10 questions
Discovering why enterprise CSMs miss churn signals despite strong client relationships
You're exploring a product opportunity around churn prediction, but you need to validate whether CSMs actually experience blind-spot losses. You have a hypothesis that earlier warning signals could prevent churn, but you're not sure if CSMs regularly encounter unexpected account losses or if better signals would actually change outcomes in their day-to-day reality.
Why standard questions fail here
Direct questions about churn prediction needs often get aspirational answers rather than real behavioral patterns. This script anchors CSMs in specific past losses they didn't anticipate, then reconstructs their actual decision-making process and available information at key moments. By working backward from surprise churn events, you'll uncover whether earlier signals would have genuinely altered their actions or resource allocation.
Sample Questions
Grounded in The Mom Test and Jobs-to-be-Done.
Jobs to Be Done: establish the user's job context and experience level
Let them talk freely to build rapport. Note their experience level for context on later answers
- What does a typical week look like for you?
- How many accounts do you typically manage?
- Generic job descriptions - probe for what they actually do day-to-day
Mom Test principle: ask about specific past behavior rather than hypotheticals
Use the 5W technique: get Who, What, When, Where, Why details. Don't rush - let them tell the full story
- When did you first realize they might churn?
- What had your interactions been like in the weeks leading up to that?
- How did you find out they were leaving?
- Vague generalizations like 'it just happens sometimes' - push for specific incident details
JTBD: uncover struggling moments and desired outcomes for early detection
Use reflective listening - repeat back what they say to encourage deeper thinking
- What would those signs have looked like in practice?
- How far in advance do you think you could have spotted those signals?
- Hindsight bias answers like 'I should have known' without specific signals
Pattern identification through multiple data points reduces single-incident bias
Look for emotional responses - note frustration, resignation, or confidence in their voice
- Which of these surprised you the most?
- How often would you say this happens to you?
- Generic industry wisdom rather than personal experience patterns
Understanding current workflow reveals gaps and workaround behaviors
Ask for screen sharing or detailed step-by-step walkthrough of their actual process
- What tools do you use for this?
- How much time does this take you each week?
- What's the most frustrating part of this process?
- Ideal process descriptions rather than what they actually do day-to-day
Identify successful patterns and desired outcomes to understand what 'good' looks like
Contrast technique: compare this success story to the earlier failure stories
- What made you realize they were at risk?
- How much lead time did that give you to intervene?
- What actions did you take once you identified the risk?
- Luck-based saves rather than systematic early detection
Get your problem validation interview script
Generate a complete interview guide tailored to your specific product and audience — with rationale, techniques, and follow-up prompts for every question.
Create free account & generate