Tracking five separate satisfaction metrics every month makes for messy meetings. A customer satisfaction index - a single weighted composite - is the executive-friendly answer. Done well, it puts a number on the overall health of your customer relationships. Done poorly, it averages signals into noise.
What goes into a satisfaction index
- CSAT - transactional satisfaction at key moments.
- NPS - overall recommendation likelihood.
- Support ticket sentiment or volume - friction signal.
- Retention rate - the behavioral confirmation that satisfaction is real.
- Portal engagement - proxy for product fit.
Different industries weight these differently. For Shopify subscription stores, we typically see weightings around 30% NPS, 25% CSAT, 20% retention, 15% support sentiment, 10% engagement. The exact weights matter less than keeping them stable so the trend is comparable month-to-month.
Established satisfaction indexes
Public indexes worth knowing: ACSI (American Customer Satisfaction Index) publishes industry benchmarks by category. UKCSI does the same in the UK. These are useful for external benchmarking, not for replacing your internal index - your business-specific composite will always be more diagnostic.
Building your own index
- Pick 3–5 input metrics you already collect reliably.
- Normalize each to a 0–100 scale.
- Assign weights that reflect business priority - defend the weights once, then leave them alone.
- Compute monthly, plot the trend, react only to multi-month direction (single-month moves are noise).
For deeper measurement work see customer satisfaction metrics and customer experience analytics.