Design a customer health score that actually predicts churn
A practical health score model with weighted signals, thresholds, and a plan to test whether it predicts renewals.
The prompt
Help me design a customer health score for my book of business.
Product and business model: [e.g. "B2B SaaS, annual contracts", "agency retainer", "usage-based"]
Data I can access: [e.g. logins, feature usage, seats used vs. purchased, support tickets, NPS/CSAT, invoices paid late, exec engagement, CRM notes]
Number of accounts and typical contract size: [DETAILS]
What churn or downgrade has looked like historically: [EXAMPLES OR PATTERNS, if known]
Tool I'll build it in: [spreadsheet / CRM / CS platform]
Give me:
1. Recommended signals grouped into categories (product usage, relationship, support, financial, outcomes), and why each predicts risk. Drop any signal that's easy to measure but unlikely to matter.
2. A weighting scheme and scoring method (e.g. 0-100), with simple formulas I can implement.
3. Thresholds for Green / Yellow / Red, and what action each color triggers.
4. Overrides: situations that should force an account to Red regardless of score (e.g. champion left, open escalation, failed payment).
5. How to validate it: back-test against past renewals and churns, what accuracy to look for, and how often to recalibrate.
6. Common mistakes (too many signals, lagging indicators only, scores no one acts on) and how to avoid them.
Do not invent benchmarks. If a threshold depends on my data, say how to set it from my historical numbers.
Highlighted [brackets] are blanks to replace with your own details.
Fill in the blanks and use this promptMore Data & Analysis prompts
What to fill in
- DETAILS
- EXAMPLES OR PATTERNS, if known
- spreadsheet / CRM / CS platform
How this prompt is built
This prompt follows the Few-Shot structure (Few-shot prompting). A few-shot prompt includes one or more worked examples of the input and the output you want, so the model can copy the pattern instead of guessing it.
Related prompts
- Design an A/B testHypothesis, variants, metrics, sample size, and duration for a valid test.
- Break down my results into the business drivers behind themBuilds a driver tree from your results, sizes how much each driver contributed to the change, and names which ones to act on.
- Identify the leading and lagging indicators for my business or teamSorts your metrics into leading and lagging indicators, links each lead to the outcome it predicts, and fills gaps in early-warning measures.
- Find the 20% of activities driving 80% of my resultsRuns a Pareto analysis to show the few activities, customers, or products producing most of your results, and what to cut or double down on.