Merchandise Allocation

The unit count starts from a presentation floor rather than from demand, and the floor is set by the smallest position in the size run rather than by the average. A size has to reach each door in enough depth to be presentable: a single unit sells on day one and is then absent, or sits as the only one of its kind on the fixture. Worked backwards — minimum buy = (minimum presentable quantity × doors) ÷ smallest curve percentage — an illustrative six-size run whose smallest position takes 5%, at a minimum presentable quantity of two, needs (2 × 30) ÷ 0.05 = 1,200 units to reach 30 doors at all. Only the units above that floor are free to follow demand. A 1,600-unit buy clears it with headroom: if ten of the thirty doors carry half the style’s demand they take 800 units, or 80 each, and the other twenty share 800, or 40 each, where a flat split puts about 53 in every door — starving the ten and overstocking the twenty while showing the identical total. A 900-unit buy does not clear the floor at any split, so the style has to go to fewer doors, carry a shorter run, or not be placed. The minimum-buy-depth guide below works the floor and its four levers in full.

The size decision carries a failure mode of its own, because a chain size curve is an average of doors that do not sell the same curve. Give a door whose demand runs larger the chain curve of 10/20/30/25/15 across XS to XL and it sells its M and L, holds its XS and XL, and reads at season end as a style that did not work. The units are not missing from the chain — they are in the wrong doors. Size-profiled allocation clusters doors by the curve they actually sell and gives each cluster its own curve, which is why a door-level size history is worth more to an allocation than any chain-level total.

Allocation is not a single event, and the sequence is the cheapest lever it has. An initial allocation goes out before any sell-through exists, so it is a forecast. A holdback — units kept in the distribution centre rather than pushed to doors — is the reserve that lets the first real sell-through read send stock where demand actually turned up. Holding units back costs floor presence while they sit and nothing else, and every correction available after it spends margin instead. What allocation hides is its own contribution to the season: a style bought right and allocated wrong produces the same end-of-season markdown line as a style bought wrong, so a markdown report cannot tell the two apart. Separating them takes sell-through by door against the units that door received, not a chain total.

The grain changes by vertical. In footwear the unit is the size run in pairs including half sizes and widths, so a door sent a broken run has pairs it cannot present at all. Accessories and bags have no size dimension, so the decision is colorway depth by door — hero colors held to a steady floor presence, fashion colorways concentrated where they sell. Home and furniture allocates at container level and by showroom floor slot, so the lever is which containers route where rather than unit counts per door. In outdoor and sporting goods, dealer prebooks commit units before an allocation exists, and a model-year changeover forces a closeout route for whatever is left. In health and beauty, shade ranges behave like size curves — a door’s shade demand is its own — and testers consume allocated units that never produce a sale. Toys and games allocate against a licence window and a fourth-quarter concentration, so depth belongs where it can clear before the sell-off period closes rather than where an annual rate would put it. Baby and juvenile hard goods carry high cube per unit, so door capacity limits depth before demand does. In jewelry and watches, allocation is piece-level, and a single piece in the wrong door is a large share of that door’s category stock.