Coefficient of variation of demand (CV)
Sample standard deviation (divisor n − 1) of sales by period ÷ mean, in percent; Lab limits for XYZ: X up to 30%, Y up to 60%, then Z
- Formula
Sample standard deviation (divisor n − 1) of sales by period ÷ mean, in percent; Lab limits for XYZ: X up to 30%, Y up to 60%, then Z- Unit
- %
- Aggregation rule
- Ratio of sums
- Source system
- POS: receipts; ERP and accounting system
- Owning role
- Category manager / merchandising analyst
- Frequency
- Monthly
Passport
- Grain
- SKU or category × sales periods
- Numerator
- Sample standard deviation of sales across periods
- Denominator
- Mean sales across periods
- At zero or no data
- Mean is 0 or fewer than two periods: “—”.
- Limits
- The SD divisor is n − 1. XYZ boundaries in Lab: X up to 30%, Y up to 60%, Z beyond. Periods of equal length; seasonality and promotions are best removed.
- Check example
- Sum of squared deviations: 1000; Number of periods: 5; Mean sales, units: 100.
Math.sqrt(s/(n-1))/m*100. Result: 15.8% - Related metrics
Version 1.0 · valid from 2026-11-16 · checked 2026-11-16
Related pages
- Term: Coefficient of variation of demand
- Calculation: ABC-XYZ matrix: a SKU’s contribution to sales and demand stability
Definition
The ratio of the standard deviation of sales across periods to the mean: the higher, the less predictable demand.
There is no single standard for the limits, 30% and 60% are a starting point and are changed. Zeros from missing stock distort CV: an empty cell means “no data”.