Supply-chain performance varies every day: supplier lead time, transit time, picking time, demand, yield and delivery performance are rarely identical from one transaction to the next. Managing only the average can hide instability. Variation analysis looks at both the center of the data and how widely results are dispersed.
Three Basic Measures
| Measure | Purpose | Example question |
|---|---|---|
| Mean | Average performance | What is average supplier lead time? |
| Range | Simple spread between minimum and maximum | How wide is the lead-time spread? |
| Standard deviation | Typical dispersion around the mean | How variable are deliveries around the average? |
Why the Mean Is Not Enough
Supplier A and Supplier B could both average 10 days. If Supplier A normally delivers in 9–11 days while Supplier B ranges from 3–17 days, the planning risk is very different even though the mean is identical. Variation influences safety stock, scheduling, capacity and customer promises.
Common Cause vs Special Cause
Statistical process control distinguishes routine variation inherent in the current process from unusual signals that may indicate a specific change or event. Control charts are designed to help teams identify those signals over time. A control limit is calculated from process behavior; it is not the same as a customer specification limit.
Supply-Chain Applications
- Supplier lead-time variability
- Daily demand variation
- Warehouse pick-rate variation
- Transport transit-time reliability
- Quality defect rates
- Order-cycle-time stability
Practical Example
If weekly supplier lead times are 8, 9, 10, 9, 8, 17, 9 and 10 days, the 17-day delivery should not simply be averaged away. Plotting the data over time helps determine whether it is a one-off special cause or evidence that the underlying process has changed.
Interview Question
Question: Why can two suppliers with the same average lead time require different inventory policies?
Answer: Because variability matters. A supplier with less predictable lead time creates more uncertainty and may require a larger buffer or different sourcing strategy even if the average is identical.
Related SCMANA Guides
References
Download the Practical Workbook
Use the SCMANA Variation Analysis Calculator to enter observations and calculate the mean, range and standard deviation for practical variation analysis.
