ABC Analysis Warehouse: Classification, Calculation, Example

80 items sit in the warehouse. A dozen of them cause almost 80 percent of the annual consumption value. That exact distribution is what ABC analysis makes visible. It shows which items need tight control, where a standard process is enough, and which positions can cost more to order than the material itself is worth.

What is ABC analysis in the warehouse?

ABC analysis splits warehouse items into three classes by one target metric: A carries the largest share of it, B a medium one, C the smallest. For classic stock and procurement control, that metric is usually the annual consumption value:

An expensive item with low annual consumption can reach the same consumption value as a cheap item that leaves the warehouse every day. That is why unit price alone isn't enough. Current stock value answers a different question too: it shows how much capital sits on the shelf today, not how much value an item moves over a full year.

SAP distinguishes, among other things, consumption value and requirement value in material classification. Oracle NetSuite uses annual consumption value, from annual consumption and unit cost, for its ABC categorization. This guide uses that value-based variant.

The Pareto principle behind it

In the late 19th century, Vilfredo Pareto observed that a small share of the Italian population held a large share of the land. That observation grew into the Pareto principle, often called the 80/20 rule.

H. Ford Dickie applied Pareto's logic to inventory in 1951, in an essay titled "ABC Inventory Analysis Shoots for Dollars, not Pennies," the origin ABC analysis still carries today. It was meant to help purchasing and warehouse teams focus their limited time on the items with the biggest economic leverage.

80/20 stays a rule of thumb. An assortment can land at 72/18/10, or a single position can already make up 40 percent of the consumption value. The distribution in your own warehouse decides where sensible boundaries sit.

ABC classes: what A, B, and C items mean

The boundaries are drawn on cumulative value. For ABC analysis the commonly cited steps are 80, 95 and 100 percent. The ranges below are the typical outcome of that, not a target you have to hit.

An ABC class counts item numbers: A covers the few that together move the most value. It says nothing about consumption quantity. A cheap box can land in A because of a high annual quantity. A rarely sold machine can land in B or C despite a high unit price.

Calculating ABC analysis: six steps

1. Set the target metric

Use consumption value if you want to prioritize procurement, tied-up capital, or planning effort. For assortment decisions, revenue or contribution margin can fit better. For storage locations and picking routes, you need a separate movement-based ABC by picks or withdrawals.

During a stock review for a repleno rollout, we set the rollout sequence with an ABC analysis; the target metric was order frequency. A supplier had sent us an item list covering all orders from the last twelve months. The switch range accounted for 31 percent of all orders, followed by distribution board construction at 23 percent and pipes/conduits at 16 percent. In that exact order, we started rolling out repleno and onboarding the items and stock. The customer had previously struggled with missing parts, so the first improvement needed to be visible quickly. Starting with the most frequently ordered product groups covers most of the daily business with limited effort, and shows fastest whether the new process holds up.

2. Clean the data: stockouts, period, pricing logic

Stockouts drag consumption down. If an item was out of stock for months, its recorded consumption is artificially low, because you can't issue what isn't there. It slides into too low a class even though real demand is higher. Check known stockouts separately and set their quantity to real demand before you calculate.

One-off outliers push it up. A project order, a promotion, or a bulk order inflates annual consumption and pushes a normal C item into A for months. Check such special cases separately and subtract returns cleanly. As the period, twelve rolling months is a workable start for many businesses; use the same price type for every item, for example the average unit cost.

3. Calculate the consumption value per item

Multiply annual consumption quantity by unit cost. Then add up all item values to get the total consumption value.

4. Sort items descending

The item with the highest annual consumption value goes on top. Only after this sort can you build meaningful cumulative shares.

5. Cumulate value shares and assign classes

As a starting point, the cumulative value decides after each row: up to and including 80 percent A, up to and including 95 percent B, above that C. Decide in advance how you handle an item that straddles a boundary.

6. Review the result

Flag new items, critical spare parts, perishable goods, bulky parts, and items with long lead times. These traits can call for their own stock rule. Update the classes every six months, or quarterly for seasonal or fast-changing assortments.

Write down the period, the pricing basis, and the calculation date before you put the sheet away. Without those three, the next run cannot be compared to this one, and purchasing ends up arguing about a class shift that only comes from a different time window.

Worked example: office supplies online retailer

A small online retailer evaluates ten core items. The "annual quantity" column holds warehouse withdrawals from the last twelve months. All values are based on average unit costs.

Three of ten item positions generate 78.4 percent of the consumption value. That is 30 percent of the positions, more than the range above suggests: with ten items, every single position carries a lot of weight. The larger the assortment, the closer the share moves to the typical 10 to 20 percent.

Ballpoint pens and notepads both move 2,500 euros a year and still land in different classes: the 95 percent boundary falls right between them. For control, that means the same rule for both, whatever the class column says.

That's the economic view. For placement in the warehouse, it looks different: notepads might get picked more often than office chairs and belong closer to the packing table despite their C value class.

ABC analysis in Excel

A spreadsheet is enough to get started. Example with annual quantity in column B, unit cost in C, and consumption value in D:

1. Consumption value in D2: =B2*C2
2. Value share in E2: =D2/SUM($D$2:$D$11)
3. Cumulative share in F2: =SUM($E$2:E2)
4. Class in G2: =IF(F2<=80%,"A",IF(F2<=95%,"B","C"))

Sort column D descending before you drag the cumulative-share formulas down. Format E and F as percentages. With multiple warehouse locations, you also need to decide whether to combine items across all sites or classify per site. The same item can be A at the main warehouse and C in a small branch.

Which stock rules fit A, B, and C

The ABC class determines how much control effort is economically justified.

For A items, overstock ties up a lot of capital. Tighter control and reliable inventory management pay off. B items should run on the standard process. For C items, the biggest lever often sits in the process itself: placing an order, approving it, and booking goods receipt can cost more internally than the item itself, the core of any C-parts management.

Cheap still doesn't mean harmless. A two-euro gasket stops a machine just as surely as a two-thousand-euro part. Such items need a criticality flag and often a higher safety stock. ABC assesses value consumption; failure impact is a second axis, and you maintain that one yourself.

Advantages and limits of ABC analysis

Advantages:

• You see which items carry the biggest economic leverage
• A spreadsheet is enough for the first evaluation
• Purchasing, stocktaking, and planning can scale their effort by class
• Classes make large assortments easier to discuss

Limits:

• The method only ever looks at the chosen target metric
• Fixed percentage boundaries can create artificial cutoffs in small assortments
• Criticality, demand variability, and lead time are missing

ABC is the wrong tool as long as you can name your most expensive items without a table: the class boundaries then split items that behave identically.

The demand variability from that last limit is what the second analysis brings in.

Combining ABC analysis and XYZ analysis

XYZ analysis adds regularity of consumption to the value dimension. It's measured via the coefficient of variation: divide the standard deviation of consumption quantities by average consumption. The data basis is monthly consumption quantities over at least twelve months.

German materials management uses the boundaries the REFA lexicon lists:

• X: coefficient of variation up to 25 percent, constant, well-planned consumption
• Y: coefficient of variation 25 to 50 percent, fluctuating or seasonal consumption
• Z: coefficient of variation above 50 percent, irregular, hard-to-forecast consumption

In a spreadsheet, you calculate this per item over its twelve monthly values with =STDEV.S(B2:M2)/AVERAGE(B2:M2).

Two conventions, a factor of two apart

English-language sources set the boundaries at 0.5 and 1.0, or 50 and 100 percent. These are not the same classes on a different scale: an item fluctuating by 60 percent is a Z part under the German boundaries and a Y part under the English ones. Set your boundaries once and keep them consistent across every evaluation, otherwise two runs aren't comparable.

An AX item has high consumption value and stable demand. You can plan it tightly and procure it often in smaller quantities. An AZ item ties up a lot of value but is hard to forecast; supplier agreements and a deliberate choice between holding stock and ordering to job matter here. A CX item is cheap and needed regularly. Simple min-max, kanban, or two-bin rules usually fit well. A CZ item needs a case-by-case check before you build up stock.

CZ is the hardest combination. Low value and irregular consumption together mean these items neither forecast well nor ever make it to the top of anyone's review list. That's exactly why both failure modes show up here at once: dead stock on items nobody checks, and a stockout on the one CZ item a job suddenly needs. A fixed per-item rule, such as two-bin or a deliberate decision not to stock at all, is more reliable here than trying to predict consumption.

Putting the result into practice

Calculate the classes, flag critical exceptions, and give each class two or three concrete rules. For B items, that means choosing between reorder point and min-max. Then record stock value, shortages, and order effort as warehouse metrics: annual consumption and average stock value are the figures you just assembled, and without that starting point next quarter's comparison has no basis.