Cycle Service Level vs Fill Rate: Two Numbers Often Mistaken for One

When teams talk about a '95 percent service level', each person often has a different definition in mind. Two common definitions exist and the difference is larger than it seems: cycle service level (CSL) and fill rate. Confusing them is one of the most common causes of misalignment between planning, sales and the customer.

Cycle service level is the probability of no stockout during an order cycle. If CSL is 95 percent, then in 95 percent of cycles stock is sufficient until the next order arrives. This metric only measures 'did a stockout occur?', not how large it was. A two-unit shortage and a two-thousand-unit shortage both count as one failure.

Fill rate is the percentage of demand supplied directly from stock. If fill rate is 98 percent, then out of every 100 units requested, 98 are delivered immediately. This metric is sensitive to the size of the shortage and is closer to the real customer experience. A customer who gets 98 percent of an order is not the same as one who gets nothing.

The mathematical relation between the two is not one-to-one. Typically, a given fill-rate target needs less safety stock than a CSL of the same percentage. So a company setting its target at '95 percent' may, depending on which metric it takes as the basis, hold very different safety stock and get a completely different result. Safety-stock formulas used in tools are usually based on CSL, while customers judge their experience by fill rate.

To keep definitions from mixing, ask teams to state the definition every time they quote a number: 'CSL of 95 percent at item level' or 'fill rate of 98 percent at order-line level'. Then choose the metric by real need. If rare but large failures matter to you, CSL is not suitable; fill rate or a metric based on shortage quantity fits better.

A common mistake is reporting one overall average. A 97 percent fill rate for the whole organization may include key items at 85 percent and unimportant ones at 100 percent. Break the metric down by ABC-XYZ class and customer importance, and set a different target per group, so safety-stock capital goes where it has the most effect.

Another mistake is overlooking the difference between 'complete order' and 'complete order line'. An order with ten lines where nine are complete is 90 percent at line level but, to the customer, an incomplete order (zero percent at order level). OTIF usually takes the strictest view, the order level.

A practical starting point: the next time someone reports 'service level', ask which definition and at what level (item, line or order). Then calculate both metrics for 20 important items. The gap between them often shows why customers are unhappy with the reported service level.