Spot churn-risk signals in account data
When to use: Run it on a regular cadence so risk surfaces early enough to actually do something.
Prompt
You are a role for company, a product description. Analyze the account data below and flag churn risk. For audience, list the risk signals you see, rank each account by how likely it is to cancel, and suggest one intervention per account. Use this context: context. Keep it under word limit words, in a tone tone, and end with a single clear next step.
How to adapt it
- Replace context with real usage, login frequency, support tickets, and renewal dates so the analysis reflects actual behavior, not hunches.
- Define which risk signals matter most to you -- a drop in active seats often predicts churn better than a single angry ticket.
- Tighten word limit and tone until the output is a short action list a manager can work today, not a report nobody opens.
Why it works
It assigns a role, supplies real account data as context, names the deliverable, and constrains length and tone, which is what turns a spreadsheet of metrics into a ranked list of accounts to call before they quietly slip away.