Markdown timing: what waiting actually costs you

Ask a founder when they last marked stock down and you will usually hear a date that is months after the stock stopped moving. Ask why the cut was five percent and you will hear that anything deeper felt wrong. Both answers come from the same place. The markdown was treated as a concession rather than a recovery decision.

Ageing stock is a depreciating asset that nobody depreciates. Every week it holds a price the market has stopped paying, it gets more expensive to sell.

Recovery falls with age and does not come back

There are two clocks running against the same units.

The first is demand. Interest in a product decays. A colour that was current last season, a variant that has been superseded, a festive pack sitting in February. Nothing you do to the price changes the fact that fewer people want it than did six months ago.

The second is cash. Holding stock costs storage, and long term storage charges at marketplace fulfilment centres bite specifically at older units. It costs working capital that could fund a buy that actually moves. It costs shelf position and dark store facings, which are finite and which you are paying for in one form or another. And it costs attention, because someone is managing the problem every week.

Put the two together and the shape is consistent. The same unit clears at a modest cut at sixty days, needs a deeper cut at one hundred and eighty, and by three hundred often cannot clear through the main channel at any sensible price. The depth you avoid today is not saved. It is deferred and it grows.

Why the first markdown is almost always too small

Because it is sized to protect margin on the units that sell, not to clear the units that will not.

A five or ten percent cut on a product with no demand signal does nothing except lower the price for buyers who were going to purchase anyway. Rate of sale stays flat. Two weeks disappear. The stock is now older and the next cut has to be deeper than the one that would have worked at the start.

So attach a test to every markdown. Define an observation window before the cut goes live, usually two to four weeks depending on the base rate of sale, and define what movement would count as a response. If rate of sale does not change inside that window, the markdown failed. It was a price change, not a markdown, and the correct response is to escalate depth immediately rather than to wait and see for another month.

Step markdowns or one deep cut

This is the question that gets argued in every meeting, and it has an answer that depends on two inputs.

The first input is remaining time. Count backwards from the hard deadline, which is expiry, the platform minimum life cut off, the end of the season, or the launch of the replacement model. Then count how long a step ladder actually takes. Each step needs its observation window. Three steps is six to twelve weeks of calendar. If the deadline sits inside that, the ladder is not available to you. Cut once and cut properly.

The second input is demand signal. Steps work where there is genuine elasticity and enough traffic to read a response. A SKU doing forty units a week gives you a readable signal at each step. A SKU doing three units a week does not, and stepping it is just a slow way of running out of time.

There is a third consideration. Repeated small cuts train buyers to wait for the next one, and that behaviour persists after the clearance ends. One clean, time bound cut is easier to defend than a slow slide.

Write the rule before you need it

The reason markdowns are late is that each one is a fresh negotiation with the person who bought the stock. Remove the negotiation.

The rule takes a standard shape. When a SKU crosses a defined ageing bucket and weeks of cover are above a defined threshold, the first markdown fires at a pre agreed depth for that category, unless the category owner records a written exception with a reason and a review date. The exception is the control. It is not a veto, and it expires.

Set approval authority in the same document. Depth up to one level sits with the category owner. Beyond that it goes to the head of the business. Beyond that, to the founder. Nobody should be discovering the escalation path in the middle of a problem.

Do not write a universal discount table. Depth is a function of remaining life, weeks of cover, the channel you are clearing through and the category. What you should fix in advance, once per category and reviewed each quarter, is the trigger, the observation window, the escalation ladder and who signs what. The depth itself should be derived from what your own stock has actually recovered in the past, not from a number someone read somewhere.

The psychology that keeps prices high

Every reason for holding price is emotional and they are all recognisable.

Anchoring to landed cost. The stock cost a certain amount, so selling below it feels like accepting a loss. The loss already happened at the point the demand did not appear. The price only decides how much you recover.

Sunk cost dressed as patience. Waiting is framed as discipline when it is usually avoidance.

Festive hope. This is the expensive one in India. There is always another festive window ahead, so brands carry stock from March to September on the belief that the season will absorb it. Sometimes it does. Often the season absorbs the fresh, well merchandised range and leaves the tired stock exactly where it was, six months older and with six months of holding cost added.

Fear of brand damage. This one is legitimate, but it is a question of where and how you clear, not whether. Clearing in a separated way protects the main range, and that is the point at which the conversation becomes a liquidation conversation rather than a markdown one.

The honest test is a single question, and it works on founders because it removes the history. If you had this cash sitting in the bank today, would you buy this stock at this price, in this quantity, right now. If the answer is no, then you are not holding an asset at its value. You are funding a position you would not take. Price it to move.