# Run a post-mortem on a churned customer account

A blameless review of why an account churned, with root causes, missed signals, and changes to prevent it next time.

- Topic: Thinking & Decisions
- Framework: RTF
- Level: Standard
- Tags: Analyze, Plan, Fill-in Template
- Source: https://www.thinkwithprompt.com/prompts/run-a-post-mortem-on-a-churned-customer-account

## Prompt

```text
Help me run a post-mortem on a customer we lost.

**Account details:** [PRODUCT, CONTRACT VALUE, TENURE, CHURN DATE]
**Stated reason for leaving:** [WHAT THEY SAID - exit survey, email, call]
**Timeline:** [KEY EVENTS - onboarding, usage changes, support issues, stakeholder changes, renewal conversations]
**Health signals we had:** [HEALTH SCORE HISTORY, usage data, NPS, ticket volume]
**What we did to save it:** [ACTIONS AND TIMING]
**Where they went:** [competitor, in-house, no replacement, out of business - if known]

Analyze:
1. Stated reason vs. likely root cause - and whether they differ. Separate controllable causes (onboarding, product fit, service, relationship) from uncontrollable ones (acquisition, budget cut, closure).
2. Timeline of warning signs: when each first appeared and when we noticed it.
3. Decision points where a different action might have changed the outcome.
4. Whether this account should have been sold or onboarded the way it was (fit issues from day one).
5. Lessons in three buckets: CS process, product, and sales/handoff.
6. 3-5 specific changes with owners, and the signal or playbook that would catch this pattern earlier.
7. Whether to plan a win-back attempt, and when.

Keep it blameless: focus on systems and signals, not individuals. Mark gaps in the timeline as [UNKNOWN] rather than guessing.
```

## What to fill in

- PRODUCT, CONTRACT VALUE, TENURE, CHURN DATE
- WHAT THEY SAID - exit survey, email, call
- HEALTH SCORE HISTORY, usage data, NPS, ticket volume
- ACTIONS AND TIMING
- UNKNOWN

Free to copy and use. From ThinkWithPrompt (https://www.thinkwithprompt.com/), a library of 2,663 curated AI prompts.