By Safock Procurement Research Team7 min read
Two items in the same storeroom are worth $359,000 and $413,000 a year. One is jasmine rice: the site uses between 216 and 272 bags every month. The other is a high-grade flour: nothing for three months, then 707 and 837 bags, and later 1,008 in a single month. Rank them by value alone and they get the same treatment. They need opposite stocking rules.
ABC analysis tells you which items matter most by value. XYZ tells you which ones you can predict. Together they put every item into one of nine classes, each with its own way of being stocked.
For each item, multiply yearly usage by unit cost to get its consumption value. Sort from highest to lowest and add up the shares as you go down the list:
The ABC-XYZ tool's sample is a year of usage for 60 items at a restaurant group, worth $3.86 million:
| Class | Items | Share of items | Yearly value | Share of value |
|---|---|---|---|---|
| A | 16 | 27% | $3,105,925 | 80.5% |
| B | 12 | 20% | $561,964 | 14.6% |
| C | 32 | 53% | $191,591 | 5.0% |
The sample follows the familiar Pareto pattern: a minority of items carries most of the value. Counting accuracy, supplier negotiation and tight stock control pay off most on the A list.
The cut-offs are a choice. Set A at 70% and B at 90% and the sample has 12 A items. Set them at 85% and 97% and it has 19. Pick cut-offs that give an A list short enough to manage closely.
Value doesn't say how steadily an item is used. For that, take its monthly usage and work out the coefficient of variation (CV): the standard deviation divided by the average.
These are common cut-offs, and some guides use tighter ones for X. Three items from the sample:
| Item | Monthly usage | CV | Grade |
|---|---|---|---|
| Jasmine rice | 230, 225, 216, 233, 272, 229, 263, 239, 225, 256, 267, 246 | 0.08 | X |
| Chicken breast (site B) | 25, 82, 120, 130, 114, 72, 19, 20, 31, 25, 25, 31 | 0.75 | Y |
| Tipo 00 flour | 0, 0, 0, 707, 837, 101, 0, 135, 1008, 569, 945, 61 | 1.14 | Z |
The sample splits into 28 X items, 13 Y and 19 Z.
The chicken shows a limitation of the measure. Its usage rises from February to May and falls back by July, so its CV is high, but a seasonal pattern like that is predictable. Seasonal items are better planned to the season than buffered with extra safety stock. Some teams remove the seasonal pattern before measuring CV for exactly this reason.
Crossing ABC with XYZ gives nine classes. In the sample:
| X (steady) | Y (variable) | Z (erratic) | |
|---|---|---|---|
| A | 9 items, $1,835,567 | 4 items, $386,368 | 3 items, $883,990 |
| B | 5 items, $277,176 | 2 items, $67,000 | 5 items, $217,788 |
| C | 14 items, $112,020 | 7 items, $24,721 | 11 items, $54,850 |
The tool suggests a stocking policy for each class:
| Class | Policy |
|---|---|
| AX | Continuous review, lean safety stock and frequent deliveries. Candidates for vendor-managed or consignment stock. |
| AY | Continuous review with moderate safety stock. Review the forecast every week and watch for trends. |
| AZ | Order against confirmed demand where you can; keep a small strategic buffer and negotiate short supplier lead times. |
| BX | Automatic replenishment on a fixed weekly cycle with standard safety stock. |
| BY | Periodic review with standard safety stock, adjusted seasonally. |
| BZ | Higher safety stock or order on demand. Check whether demand can be consolidated across sites. |
| CX | Min/max or two-bin replenishment with larger, less frequent orders to cut ordering cost. |
| CY | Min/max with generous buffers and a quarterly review. Stockouts cost more than the extra stock. |
| CZ | Order only when needed. Strong candidates to remove from stock or source from another site. |
Two cells deserve particular attention.
AZ: high value, erratic. Three items, 23% of all value. Holding enough safety stock to cover their swings would tie up a great deal of cash; holding too little means running out of something important. The better answer is usually upstream: order against confirmed demand, such as booked events or production plans, and shorten the supplier's lead time so less buffer is needed.
CZ: low value, erratic. Eleven items worth $54,850 between them, 1.4% of value. Each takes shelf space, counting time and a slot in every order. For a multi-site business, many can be held at one site or a central store and transferred when needed, or dropped from the range.
Click a cell to see its stocking policy and an example from the sample. Below it, paste one item's monthly usage to see its CV and grade. The default is the erratic flour.
The tool's 60-SKU sample · click a cell
| Xsteady | Yvariable | Zerratic | |
|---|---|---|---|
| A | |||
| B | |||
| C |
AZ · High value, erratic
Order against confirmed demand where you can; keep a small strategic buffer and negotiate short supplier lead times.
3 SKUs, 22.9% of yearly value. Example: Tipo 00 flour 50 lb, CV 1.14.
Average 363.6, CV 1.14, grade Z
Free tool
ABC-XYZ Analysis Tool
Upload usage by item and month, or annual totals, set your own cut-offs, and get every item's class, a Pareto chart, the nine-cell matrix and a stocking policy for each class, downloadable as a CSV.
Runs in your browser. No sign-up.
Ranking items by yearly consumption value (usage × unit cost) and grouping them: A items make up the first 80% of value, B the next 15%, and C the last 5%. A small share of items usually accounts for most of the value, so A items get the most attention.
Grading items by how predictable their demand is, using the coefficient of variation (standard deviation ÷ average) of monthly usage. Common cut-offs are X up to 0.5 (steady), Y from 0.5 to 1.0 (variable), and Z above 1.0 (erratic).
Value alone doesn’t say how to stock an item. A high-value item used steadily can run on lean safety stock and frequent deliveries; a high-value item with erratic demand needs ordering against confirmed demand and short lead times. Combining the two gives nine classes, each with its own policy.
A common default is 80% and 95% of cumulative value for the A and B cut-offs, and 0.5 and 1.0 for the X and Y cut-offs. Adjust them to your business: tighter A cut-offs give a shorter, more focused A list. Some guides use lower CV cut-offs for X items.
Seasonal items vary a lot from month to month, so their raw coefficient of variation puts them in Y or Z even though the pattern is predictable. Remove the seasonal pattern before measuring variation, or treat known seasonal items as plannable and stock them to the season.