Inventory
ABC Analysis: Sorting Inventory by What Actually Matters
Ranking stock by annual value tells you where to spend management attention. Ranking it by value alone tells you to ignore the cheap part that stops the line.
ABC analysis is the observation that inventory attention should not be spread evenly. A small share of items accounts for most of the money, and those items deserve tight control, frequent counting, and a planner who knows them by name. The rest can be managed with simpler rules.
How to build it
The percentages are conventions, not laws. What matters is the shape: in most catalogues, A is 10-20% of the item count, and C is over half the items and almost none of the money.
What each class gets
• A — tight forecasting, frequent cycle counts, individually reviewed reorder points, named owner.
• B — periodic review, automated reorder with occasional sanity checks.
• C — simple rules, generous safety stock, count once or twice a year. Holding extra is cheaper than thinking about it.
The counter-intuitive part is that C items should often be over-stocked. The carrying cost is trivial and the planning time saved is not.
The single-dimension trap
Value ranking has a well-known failure: the cheap, low-volume component whose absence halts production. It lands in C and gets C-class treatment, right up until the day the line stops.
A five-cent part that stops a production line is not a C item, whatever the value ranking says.
Add a criticality axis
The fix is a second dimension. Score each item for consequence of stockout — halts production, delays a customer order, merely inconvenient — and combine it with the value class. You end up with a small grid, and the cells that matter are the high-criticality ones regardless of value. Those get A-class control on availability even when they get C-class attention on cost.
Other useful dimensions
• Demand variability — steady items are easier to automate than lumpy ones.
• Lead time — a long-lead C item needs earlier ordering than its class suggests.
• Obsolescence risk — high-value items near end of life need the opposite of safety stock.
Re-run it, and expect movement
Classification drifts as products launch and decline. Re-run quarterly and look specifically at items that changed class: a B item moving to A usually means demand grew without anyone adjusting the controls, which is exactly where the next stockout is coming from.
The criticality axis is the fix for the classic failure. We had a £0.40 sensor sitting in C class that idled a line for eleven hours. Two-dimensional grid now, and that part sits in the top-right cell.
Deliberately over-stocking C items still feels wrong to people who came up through cost control, but the planning time it frees up is the actual saving. Took a year of arguing to land that internally.
Looking specifically at items that changed class is the tip I'll steal. A B moving to A is almost always a stockout in three months' time, and it's invisible if you only look at the current classification.