Demand Variability (CV) and XYZ Classification Calculator
Before choosing a forecast method you need to know how noisy an item is. The coefficient of variation (CV) and the average demand interval (ADI) place every SKU in an XYZ class and in one of four demand patterns. This calculator does both from three numbers.
Your numbers
Average demand per period over your history window, for example the last 24 months.
Sample standard deviation of the same period series.
Total periods divided by number of periods with non-zero demand. 1.0 means demand every period.
Items at or below this CV are class X (stable). Many sites use 0.5.
Items above X limit and up to this value are class Y. Above it is class Z.
Your results
Classification cut-offs are conventions, not laws. Tune the CV limits to your own item population and review them yearly.
CV as standard deviation rises
Sweeps Standard deviation of demand from 0 to 600, holding your other inputs fixed.
Classify your full item master
Optional: we can run this CV and ADI classification across your ERP demand history and return a spreadsheet.
Optional. The tool and every export work without this. No spam.
How to use it
Pull the last 18 to 36 periods of demand for the item at the level you forecast, such as SKU by month. Enter the mean and the sample standard deviation, then enter ADI, which is the number of periods divided by the number of periods that had any demand. Set your X and Y limits and read the class and pattern. Run the same logic across the catalogue in a spreadsheet or the ERP once you trust the cut-offs.
- Use the same period bucket you plan in
- Exclude one-off bulk orders only if you will also exclude them from the forecast
- Use at least 12 periods
- Recompute quarterly
The method behind it
CV is sigma divided by mean. XYZ analysis ranks items by this CV: X items are stable, Y have trend or seasonality and moderate noise, Z are erratic. The Syntetos-Boylan scheme adds ADI and CV squared: ADI below 1.32 with CV squared below 0.49 is smooth, ADI below 1.32 with higher CV squared is erratic, ADI above 1.32 with low CV squared is intermittent and above both is lumpy. Smooth items suit exponential smoothing, intermittent and lumpy items suit Croston or SBA style methods.
Reading the result with an example
Mean 480, sigma 190 gives CV 0.40, class X. With ADI 1.1 the item is smooth, so a simple statistical model should do well and safety stock can be driven by a normal error assumption. If sigma rises to 600 the CV passes 1.25, the item becomes class Z and erratic, and you should expect poor forecast accuracy and higher safety stock, or consider make-to-order or postponement for it.
Common mistakes
Calculating CV on a series that contains a strong trend or seasonality overstates noise, because part of the movement is predictable. Detrend or compare against forecast error instead when you can. Another mistake is mixing channels or pooling distinct customers, which hides intermittency. Finally, do not treat the class as permanent: new products and phase-outs change class quickly.
Frequently Asked Questions
What is a good CV for demand?
There is no universal value, but below about 0.5 is usually considered stable and easy to forecast, 0.5 to 1.0 moderate, and above 1.0 erratic. Compare against your own portfolio and check whether trend or seasonality is inflating the figure.
What is the difference between ABC and XYZ?
ABC ranks items by value or volume, showing where money is. XYZ ranks by variability, showing how hard demand is to predict. Combining them gives nine cells, for example AZ items are valuable and unpredictable and deserve the most planning attention.
Why do ADI 1.32 and CV squared 0.49 matter?
They are the cut-offs from research by Syntetos and Boylan on when Croston type methods beat simple exponential smoothing. They are a practical rule of thumb, not exact boundaries.
Should zero periods be included in the mean?
Yes, for forecasting at the period level include zeros in mean and sigma, because they are real periods with no demand. ADI then describes how often demand actually occurs.
If you want this classification run across every item in your ERP and fed into the right forecast method automatically, Netray can build a read-only analysis over your item and demand history.
Related Tools
Croston Intermittent Demand Forecast Calculator
Apply one Croston update step to demand size and interval estimates, then read the per-period rate, the SBA corrected rate and demand over your horizon.
Supply Chain & InventoryMAD to Safety Stock Forecast Error Calculator
Use your measured mean absolute deviation of forecast error to set safety stock and reorder point instead of a guessed buffer.
Supply Chain & InventoryLinear Regression Trend Forecast Calculator
Enter six periods of demand and get the least squares trend, how well it fits and the forecast for the period you choose.
Supply Chain & InventoryStatistical vs Sales Forecast Blend Calculator
Weight the statistical forecast and the sales or customer forecast by how accurate each has been, and see the consensus number and the gap between sources.
More Supply Chain & Inventory tools
See all- Oracle Fusion Inventory Turns and Carrying Cost Calculator
- Distribution ERP Fill Rate Calculator
- Distributor Rebate Tracking Worksheet
- Branch Transfer Planning Worksheet
- Distributor Slow-Moving Inventory Review Worksheet
- Distributor Vendor Performance Scorecard Worksheet
- Distributor GMROI Calculator
- Distributor Backorder Cost and Fill Rate Calculator
- EOQ with Vendor Minimum Order Calculator