Most manufacturers segment inventory by value — the familiar ABC split, where a small proportion of items accounts for most of the spend. It’s a genuine improvement over treating everything alike. It’s also only half the picture, and the missing half is the one that determines how much stock you actually need.

What ABC Tells You

ABC ranks items by annual consumption value. A items are the high-value minority, C items the low-value majority, B in between. The management implication is straightforward: concentrate attention where the money is.

What it doesn’t tell you is anything about predictability. Two A items with identical annual value can behave completely differently — one consumed steadily every week, the other in three unpredictable bursts. They need different stock policies, and ABC alone puts them in the same bucket.

What XYZ Adds

XYZ classifies by demand variability. X items are stable and predictable. Y items show pattern with variation — seasonal or trending. Z items are erratic and difficult to forecast.

Variability is what drives safety stock. A perfectly predictable item needs almost none regardless of value. An erratic one needs a substantial buffer even if it’s cheap. Classifying on this axis tells you where your inventory investment is genuinely required rather than merely traditional.

The Nine-box View

Combining them gives nine segments, and the corners are where the useful decisions sit.

AX — high value, predictable. The ideal segment. Tight control, minimal safety stock, frequent small replenishment. These are your best candidates for just-in-time arrangements and scheduling agreements.

AZ — high value, erratic. The hardest and most expensive segment, and where management attention pays off most. Options include supplier flexibility agreements, customer collaboration to reduce the erraticism at source, or accepting higher stock deliberately with the cost understood.

CX — low value, predictable. Automate entirely. Larger order quantities, generous buffers, minimal human involvement. The carrying cost of over-stocking a cheap predictable item is trivial compared to the time cost of managing it closely.

CZ — low value, erratic. Hold a reasonable buffer and stop thinking about it. These items consume disproportionate planning attention relative to their value, and the right answer is usually to buffer generously and automate.

Doing it Properly

Use at least twelve months of consumption data, ideally twenty-four to capture seasonality. Measure variability with the coefficient of variation — standard deviation divided by mean — which is comparable across items of different volumes.

Reclassify quarterly. Items move between segments as demand patterns change, and a classification from two years ago is describing a business that no longer exists. Watch the movements as well as the positions — an item drifting from X to Z is telling you something about a customer or a market before it shows up anywhere else.

And check the segment sizes. If eighty percent of your items land in Z, your data probably has gaps or your classification thresholds need adjusting rather than your business being uniquely chaotic.

ticktick.ai classifies items on both axes from live consumption data and applies stock policy by segment automatically.