Deciding When to Cut the Price on Clothes That Stop Selling
A seasonal retailer must clear inventory before the season ends, and every day it waits is a day of lost margin or lost sales. Markdown optimisation is the mathematics of that tradeoff.
A Perishable Product That Does Not Rot
Seasonal merchandise has a deadline. Winter coats have almost no value in March, swimwear has almost none in October, and a fashion item outside its season sells at a fraction of its price.
The retailer bought the inventory months earlier, based on a forecast, and cannot reorder or return it. What remains is a decision about price over a shrinking window.
That decision is markdown optimisation, and it is one of the clearest applications of quantitative methods in retail.
The Two Errors
The problem is symmetric and both errors are expensive.
Marking down too early gives away margin on units that would have sold at full price. A retailer that discounts in week four when demand would have cleared the stock by week eight has funded a discount for nothing.
Marking down too late leaves inventory at the end of the season, when the only options are deep clearance, holding it for a year at a carrying cost, or selling it to a liquidator for cents.
| Timing | Margin per Unit | Units Sold |
|---|---|---|
| Early markdown | Lower | Higher |
| Late markdown | Higher on early sales | Leftover stock at salvage |
The optimal markdown is not the one that clears the inventory. It is the one that maximises total margin across the season, which frequently means accepting leftover stock rather than discounting deeply enough to sell every unit.
What the Model Needs
An optimisation requires three inputs and each is estimated rather than known.
Price elasticity at this point in the season, meaning how much additional demand a given discount generates. This differs by category, by item, and critically by how far into the season it is, since a shopper in week two behaves differently from one in week ten.
Remaining demand at the current price, forecast from sales to date, comparable items, and prior seasons.
Salvage value of unsold units at the end, which sets the floor on how much a markdown is worth avoiding.
The model then solves for a markdown schedule, frequently a sequence of discounts at defined dates rather than a single decision, maximising expected total revenue.
The Constraints That Make It Harder
Real implementations operate under restrictions that are commercial rather than mathematical.
Price ladders. Retailers use defined price points rather than continuous prices, so the model chooses among discrete options.
Markdown cadence. Prices cannot change daily without confusing customers and destroying trust in the price, so changes happen weekly or at defined intervals.
Brand and channel consistency. A price cut in one region or channel frequently must apply elsewhere, which removes the ability to optimise locally.
Vendor agreements. Suppliers may fund markdowns through allowances, and the terms of that funding influence when a markdown makes sense for the retailer independent of the arithmetic.
Why Vendor Funding Distorts It
That last point deserves attention because it is where the mathematics meets the negotiation.
Markdown allowances are payments from a supplier compensating the retailer for discounting its product. They exist because a supplier would rather subsidise a markdown than have the retailer stop buying the line.
A retailer whose markdown is partly funded by the vendor faces a different economic calculation from one bearing the full cost, and will rationally discount earlier and deeper on subsidised lines.
The consequence is that the observed markdown behaviour in a store reflects supplier negotiations as much as demand estimation, which is invisible to the customer and material to the supplier.
The Alternative Approaches
Not every retailer optimises markdowns, and the alternatives are deliberate strategies rather than failures.
Everyday low pricing avoids the cycle entirely by setting prices that do not require clearance, which sacrifices the promotional response and produces steadier demand and lower inventory risk.
Scarcity models deliberately under buy, selling out at full price and accepting lost sales, which protects margin and brand and requires accepting that some demand goes unmet. Several fast fashion and luxury operators run this way.
Off price channels divert unsold inventory to a separate discount format rather than marking it down in the primary channel, protecting the price integrity of the main brand.
The Bottom Line
Markdown optimisation solves for the price path that maximises margin across a season, balancing the cost of discounting units that would have sold against the cost of ending with inventory worth almost nothing. The mathematics is straightforward and the inputs are all estimates, particularly elasticity, which changes as the season progresses. What actually drives observed behaviour in many retailers is not the model but who is paying for the markdown, which is a negotiation rather than a calculation.