How to lift sell-through without discounting a single unit

When sell-through is soft, the reflex is to discount. It works, in the sense that a cut always moves units, and it is easy to defend, because the numbers respond immediately. It is also the most expensive lever in the building and usually the wrong one, because it treats a distribution problem as a price problem. The stock did not fail to sell because it cost too much. It usually failed to sell because it was in the wrong door, the wrong size, or arrived too late to catch its demand.

Sell-through is a function of putting the right units in front of the right demand at the right time. Price is only one input, and it is the input that costs you margin every time you reach for it. The three levers that lift sell-through without giving a point of margin away are allocation, size curve, and reorder timing. None of them requires buying more, and none of them requires a markdown. They just require putting the buy you already made where the demand actually is.

This post is not about how to read your sell-through report. It is about what to pull once you have read it: the three operational levers that raise the number, in order of how much they move it.

The biggest lever is also the most ignored. Two doors carrying the same style can post wildly different sell-through, not because one has better customers but because the allocation matched neither door's demand. Push units evenly and you stock out the door that could have sold twice as many while burying the door that could barely sell half. Both post soft sell-through, and both were avoidable with a smarter split of the same total units.

Lifting sell-through through allocation means matching each door's receipt to how that door actually trades, then moving stock between doors mid-season as the picture develops. The units that are dead in one door are often the exact units another door has stocked out of. A transfer costs you a bit of freight and lifts sell-through on both sides. No markdown, no extra buy, just the same inventory in the door where someone will pay full price for it.

The reason allocation moves the number more than the other levers is that it fixes both problems at once. A discount only helps the door that is long, and it helps by giving up margin. A transfer helps the long door by clearing its dead stock and helps the short door by capturing demand it was going to lose, and it does both at full price. You are not choosing which problem to solve. You are solving the overstock and the stockout with a single move, using inventory you already paid for. That is why it belongs at the top of the list, above size curve and well above price.

The second lever is the size curve, and it leaks sell-through quietly because the style-level number looks fine. Buy a blanket size split and you reliably sell out the middle sizes early, losing every customer who wanted a medium after week two, while the ends sit and eventually clear at markdown. The style sold through acceptably on paper. In reality you lost full-price demand on the sizes that sold and manufactured overstock on the sizes that did not.

A demand-shaped size curve, built at option and door level rather than copied across the range, lifts sell-through on both ends of the problem. The sizes that sell get the depth to keep selling at full price instead of stocking out. The sizes that do not stop arriving in volume, so they never become the residual that drags the style into clearance. For new styles with no history, inheriting a curve from the nearest historical cluster beats defaulting to the category average, which is where most size misallocation on newness comes from.

The size lever compounds with the allocation lever, which is why they belong together. The right size curve is not the same in every door: the flagship in the city runs to smaller sizes, the outlet runs larger, and a single curve pushed everywhere loses on both. So the highest-value version of this is a curve built per door archetype, matching not just which sizes sell in the range but which sizes sell in that kind of location. Get that right and a mid-season transfer becomes even more powerful, because you are not just moving a style to the door that wants it, you are moving the sizes that door wants. The units go where the demand for that exact size actually is.

That gap is the entire case against reaching for a discount first. The brands closing it are not pricing more aggressively. They are allocating to how doors trade, buying the size curve demand actually wants, and reordering the runners in time, and the sell-through follows without a single point of margin given away.

The third lever is timing, and it is the one brands congratulate themselves for getting wrong. A style that sells out early looks like a win, and it is a loss: every week it sits at zero stock is full-price demand walking out the door to a competitor. The sell-through on the units you had was 100 percent, but the sell-through against the demand that existed was far lower, and the gap is invisible unless you were watching the rate.

Lifting sell-through here means detecting the runners while their lead time still allows a reorder that lands in season, and chasing them before the demand cools. That is a speed problem, and it is where a live platform matters. It watches sell-through against plan continuously, flags the styles pacing hot enough to justify a chase while there is still runway, and proposes the reorder with the lead time and the remaining window built in, so the planner acts while it still counts instead of noticing at end of season that the winner sold out in week three.

Discounting moves the units you should not have had. Allocation, size, and timing sell the units you already own to people who will pay full price. One costs margin. The other three are free.

There is a reason brands reach for the discount first anyway, and it is worth being honest about it: the discount is the only lever that works without knowing anything. You do not need to understand why a style is soft to cut its price, and you see a response within days. The three real levers all require diagnosis. You have to know which doors are long and which are short to transfer, which sizes are leaking to fix the curve, which styles are pacing hot enough to justify a reorder. That diagnosis is work, and the discount is a way to skip it. But skipping it is exactly what makes the discount expensive, because you end up cutting the price of stock that only needed to be in a different door, or in a different size, or reordered a week earlier. You pay margin to avoid understanding your own inventory.

Reach for the discount last, not first. When sell-through is soft, the answer is almost always in the distribution: get the units to the doors that trade them, buy the size curve demand actually wants, and reorder the winners before they cool. Pull those three levers on the live signal and sell-through climbs on the buy you already placed, with your margin fully intact.