In-Stock Rate

In-stock rate is the share of planned selling positions that actually held sellable stock when measured. In apparel a position is one door, or the DTC site, by one size of one style-color in one week. It measures availability at the grain the customer buys, which makes it the metric that tells a planner whether a style's sales history can be read as demand at all.

Two definitional choices decide whether the number means anything. The denominator is planned positions — the door and size combinations the allocation intended to stock. A door never allocated a size is an assortment decision, not an out-of-stock, and counting it buries real availability failures under deliberate ones. The numerator counts sellable units only: a position holding only a floor sample, a tester or damaged units is out of stock.

How it is measured

Measure it at door by size by week, the grain at which a customer finds or misses a size, using one of two conventions. A snapshot records whether each position held stock at a fixed point, such as the weekly count; a time-weighted rate records the share of trading days it held stock. The snapshot misses a position that sold out on Tuesday and was refilled on Friday; the time-weighted rate catches it but needs daily on-hand. Whichever is chosen stays fixed, because the trend is read more than the level. In footwear the position is the size and width inside the size run at a door; in beauty it is the shade by door, and a missing tester counts against it.

The figures below are illustrative and not drawn from any brand. A style-color allocated to 40 doors in 6 sizes has 40 × 6 = 240 planned positions for the week. At the weekly check 204 hold sellable stock, so the in-stock rate is 204 ÷ 240 = 85%. Every door still holds at least one size, so at style level the style-color reads as in stock everywhere while 240 − 204 = 36 positions are empty. Style-level availability hides the size-level stockouts that cost the sale.

Why in-stock rate censors demand

A position with no stock records no sales, whatever its demand. Sales history is therefore demand filtered by availability — censored, in statistical terms. The example's style-color recorded its week from 204 positions, not 240, so any sensed rate, size curve or reorder read from that history understates demand by whatever the 36 empty positions would have sold. The positions that stock out first are the fastest — the core sizes, the hero shade — so the censoring falls hardest where demand is highest. That is why in-stock rate is read beside sell-through before a sensed rate is trusted, as demand sensing in retail sets out.

In-stock rate vs fill rate and lost sales

Fill rate measures supply against an order: the share of units a wholesale account ordered that shipped complete and on time. In-stock rate measures presence at the point of sale. They diverge both ways: an account can receive every unit it ordered and still run empty positions because the order was too shallow or split on the wrong curve, and a door can stay in stock on a style whose orders shipped short because it sold slowly.

Lost sales are the demand the empty positions turned away. In-stock rate shows where demand was censored; a lost-sales estimate sizes how much.

RetailNorthstar holds allocation, sizing and door-level sell-through in one connected workflow, so a size-level availability gap can be read beside sell-through on the sizes still in stock. See how RetailNorthstar handles allocation →