ABC analysis is a way of admitting that not all stock deserves equal attention. You rank every item by how much money flows through it in a year, then split the list into three classes and manage each one differently.
The pattern behind it is consistent across almost every operation. A small group of items, usually around twenty per cent, accounts for roughly eighty per cent of annual value. A much larger group contributes very little.
Once you can see that split clearly, a lot of decisions get easier. Where to count more often, what to slot near despatch, which items justify a safety buffer, and which ones you should probably stop stocking at all.
Rotation rules matter just as much for the A items. Our comparison of FIFO vs LIFO explains how each method changes stock valuation and ageing. Classify first, then decide how each class should rotate.
What ABC analysis actually measures
The ranking uses annual usage value, not unit cost and not units shipped. You multiply the annual quantity consumed by the cost per unit, which gives the total money moving through that item each year.
That distinction catches people out. A cheap fastener picked thousands of times a month can easily outrank an expensive item that sells twice a year. Equally, a costly component used sparingly may land in B rather than A.
How to run an ABC analysis
Pull twelve months of usage per SKU, multiply each by unit cost, and sort the list from highest to lowest. Then calculate a running cumulative percentage of total value down the list.
Cut the list where the cumulative value crosses roughly 80 per cent, and again at about 95 per cent. Everything above the first cut is class A, the middle band is B, and the long tail is C. Those thresholds are conventions, not rules, so adjust them if your curve is unusually flat or steep.
A full refresh once or twice a year is usually enough. Items do migrate between classes, but reclassifying monthly creates churn in your operating rules for very little benefit.
What to do differently with each class
Class A items earn tight control. Count them monthly, review their safety stock and reorder points regularly, slot them in the most accessible pick faces, and keep a close watch on supplier performance.
Class B items sit on a lighter touch. Quarterly counts, periodic buffer reviews, and standard slotting usually suffice.
Class C items should be governed by simple rules that cost almost nothing to run. Count them once or twice a year, order in larger and less frequent batches, and resist building elaborate forecasts for items you sell a handful of times.
Using ABC analysis for slotting
Travel time dominates picking labour, so putting your highest-velocity items closest to despatch pays back immediately. Here, though, you should rank by pick frequency rather than by value.
That is why many operations run two parallel classifications: one by value for purchasing and counting, another by movement for slotting. The same item can reasonably be an A for cash and a C for travel distance.
Where ABC analysis falls short
The method is blind to criticality. A low-value gasket that halts a production line when it runs out will land in class C, and a purely value-driven rule will under-protect it.
It is also backward looking. A product launched three months ago has no meaningful annual history, so it will be misclassified until a full year of data exists. Flag new lines and handle them manually until they settle.
Seasonality distorts things too. An item that sells heavily for one quarter may average out into B, even though it needs class A treatment during its peak.
Extending to a two-dimensional view
Many operations pair value with demand variability, producing a grid rather than a single list. Steady, high-value items can run on lean buffers because they are predictable. Volatile high-value items need more cover.
This is more work, but it is where the real savings sit. Treating a predictable A item and an erratic A item identically guarantees you are either overstocked on one or short on the other.
How ABC analysis fits daily operations
The classification is only useful once it drives an actual rule. Tie it to your count frequency, your buffer policy, and your slotting plan, and it becomes a quiet engine behind better decisions.
In practice that means feeding the classes into your cycle counting programme and letting them shape safety stock levels. Most inventory systems can store a class code per item and filter on it. Professional bodies such as ASCM treat the technique as foundational planning practice, and standardised product identification through GS1 makes the underlying usage data far cleaner to pull. Good classification also sharpens order picking routes.
Frequently asked questions about ABC analysis
How many classes should you use?
Three is standard and usually enough. Some operations add a D class for dead stock, which is genuinely useful because it separates slow movers from items that have not shipped at all.
Should ABC analysis use cost or selling price?
Use cost when the goal is managing working capital and inventory investment. Use margin when the goal is prioritising revenue protection. Pick one, then stay consistent so the classes remain comparable over time.
How often should the classification be refreshed?
Annually for most operations, or twice a year if your catalogue turns over quickly. More often than that creates instability in the rules built on top of it.
The practical takeaway
Rank by annual usage value, cut at roughly 80 and 95 per cent, and then actually change how you treat each class. Add a second dimension for variability when you are ready, flag critical and new items manually, and refresh the whole thing once a year.