Coefficient of variation calculator

CV = (SD / mean) × 100%. Compare spread across different scales

Coefficient of variation (CV) — relative spread: (standard deviation / mean) × 100%. Lets you compare variability across different scales.

FAQ about coefficient of variation

Enter survey data

Option 1: enter a sample of numbers. Option 2: enter mean and standard deviation.

Assumptions and limitations

  • CV = (σ / μ) × 100%, where σ = standard deviation, μ = mean. For sample: SD with n (unbiased estimate).
  • When mean is close to zero, CV is meaningless — the calculator will show a warning.
Coefficient of variation
Enter data

In SurveyNinja, NPS, CSAT and more are calculated automatically. Create a survey in 5 minutes and get real-time analytics.

Start free

Data is not stored — calculation runs in your browser.

Coefficient of variation

Why use CV

To compare spread of variables on different scales: e.g. variation in revenue (in currency) vs variation in market share (%). SD has different units; CV is dimensionless %.

Interpretation

CV < 15% — low variability; 15–30% — moderate; > 30% — high. Norms vary by field: biology and economics often use different thresholds.

Limitations

Meaningful only for positive mean (or when interpreting |μ|). For zero or negative mean, CV is not used.

When to use

Comparing stability of metrics (e.g. NPS by region vs conversion by channel), assessing risk and spread in reports.

Coefficient of variation calculation examples

1 Revenue

Mean: 1,000,000, SD: 200,000. In "Mean and SD" mode enter 1000000 and 200000.
Result: CV = 20% — moderate variability relative to mean.

2 Conversion

Mean: 3%, SD: 0.9%. Enter 3 and 0.9 in Mean and SD fields.
Result: CV = 30% — high variability. Relative spread is higher for conversion than revenue.

3 NPS by region (sample)

NPS scores for 8 regions: 20, 35, 40, 25, 50, 30, 45, 38. Enter in "Sample" mode.
Result: calculator will compute mean, SD and CV. CV shows how much NPS spread is relative to average level.

4 Stable metric

Mean: 50, SD: 5. Small spread relative to mean.
Result: CV = 10% — low variability. Metric is stable from period to period.

5 Comparing two channels

Channel A: mean 100 conversions, SD 40 → CV = 40%. Channel B: mean 50, SD 10 → CV = 20%.
Result: channel A has higher relative spread — forecasts less precise. CV lets you compare channels on different scales.

6 Response time (seconds)

Mean response time 2.5 sec, SD 0.8. Enter 2.5 and 0.8.
Result: CV ≈ 32% — high variability. Some requests take longer; useful for SLA and optimization.

Interpreting CV

CV
Interpretation
< 15%
Low variability
15% – 30%
Moderate variability
> 30%
High variability

Norms depend on industry and metric; biology and economics often use different thresholds. CV is meaningful only for positive mean.

Frequently asked questions about coefficient of variation

Automate your metrics

In SurveyNinja, NPS, CSAT and more are calculated automatically. Create a survey in 5 minutes and get real-time analytics.

Start free