Fill Rate Calculator

The maths behind it

Fill rate is the share of demand met from stock on hand. The unit fill rate is units shipped ÷ units ordered; the line fill rate is complete lines ÷ total lines; the order fill rate is complete orders ÷ total orders.

The three give progressively lower numbers on the same operation, and quoting one without saying which is the most common way fill rate gets misrepresented.

Putting numbers to it

In a month, 1,000 order lines covering 1,740 units are placed. Twelve lines are short, totalling 38 units unshipped. Those twelve lines fall across ten orders out of 620.

Unit fill rate: 1,702 ÷ 1,740 = 97.8%. Line fill rate: 988 ÷ 1,000 = 98.8%. Order fill rate: 610 ÷ 620 = 98.4%.

Three figures between 97.8% and 98.8% on identical performance. And note the relationship to service level: a 95% cycle service level does not mean a 95% fill rate: cycle service level is the probability of no stockout in a cycle, and fill rate is the proportion of demand met, which is almost always the higher number.

Where it is unreliable

Fill rate is measured against orders received, and orders are not placed for products that are visibly out of stock. A product unavailable on the website for a fortnight generates no orders and therefore no fill rate penalty, while losing every sale it would have made.

Substitutions also flatter it. Shipping a different size or colour and counting the line as filled records a success where the customer experienced a compromise.

How to act on this

Report the line fill rate as the operational measure and the order fill rate as the customer-experience measure. The first tells you how the stock is performing; the second tells you how many customers had a bad experience, and it is always the lower number.

Then track lost demand separately from fill rate. The two together describe availability; either alone describes half of it.

Fill rate in supplier agreements

Retail and wholesale supply agreements routinely specify a fill rate, commonly 95% to 98%, with chargebacks for falling short. The definition in the contract is what matters, and it is usually the line or order version rather than the unit version, the harder ones to hit.

The measurement window matters as much as the target. A 98% requirement measured weekly is considerably harder than the same figure measured annually, because a single bad week cannot be averaged away.

For a supplier, the practical response is to size safety stock against the contractual definition rather than against a general service level, and to hold more on the specific lines that customer orders most. A blanket service level applied evenly will miss a line-level target while appearing to be comfortably above it in aggregate.

One measurement that improves the picture considerably: record attempted demand for out-of-stock items. A back-in-stock notification signup, a search with no results, or a product page view on an unavailable item are all proxies for demand the fill rate never sees.

Adding that estimated demand to the denominator produces a harsher and more honest number, and it is the version that reflects what customers actually experienced rather than what the order system happened to capture.