Retail seasonality
A Socratic walk-through of retail seasonality — reasoned out one step at a time, not lectured.
The question we started with
THE QUESTION #Why does a shop earn most of its year in six weeks and then discount the same goods?
A toy shop, a jeweller, a department store: each takes a strikingly large share of its annual takings in the weeks between late November and Christmas. In the United States the winter holiday period is reckoned at roughly a fifth of the whole year's retail sales, and for gift-heavy categories it is far more than that. Then, in the second week of January, the same shop cuts the price of the same objects by a third.
Two things about that deserve suspicion. If the goods were worth the December price, why are they not worth it in January? And if the shop can sell so much in six weeks, why does it spend the other forty-six comparatively idle rather than spreading the work out?
Reasoning it through
REASONING #Begin with the calendar, because everything here hangs off it. Demand is not merely uneven — it is phase-locked. It arrives at the same point in the cycle each year, driven by dates nobody in retail controls: a holiday, a school term, a change of weather. Ask what a business must do when the rhythm of demand is fixed externally and cannot be persuaded to move.
It must size itself to the peak. Shelf space, stockroom, checkout lanes, the loading bay, the seasonal staff — all of it has to be adequate for the busiest week of the year, and almost all of it is paid for across the whole year. So the shop's cost structure is set by the peak while its revenue is spread unevenly around it. That already answers part of the puzzle: the quiet forty-six weeks are not a missed opportunity the shop declined to take, they are the trough of a cycle whose amplitude it cannot choose.
Now the discounting, which is the more interesting half. Consider a box of Christmas decorations on 27 December. What is it worth? Not what it cost, and not what it fetched three weeks earlier. The willingness to pay for it has just fallen off a cliff, and it will not recover for eleven months — during which the shop pays to store it, ties up cash in it, and risks it going out of style. The goods are perishable, not because they rot, but because their value is attached to a point in the cycle that has passed.
That reframes the January sale. The shop is not admitting the December price was fake. It is comparing two numbers: what it can get now, and the salvage value eleven months and one storage bill later. When the second is lower, cutting the price is the profitable move even at a loss on the item.
Why not simply order less, then, and avoid the whole business? Here is the constraint that makes seasonality genuinely hard. For a seasonal good with a long supply lead time, the shop commits to a quantity before it can observe demand — often months before, from a factory that will not take a second order in time. That single-shot decision has an asymmetric penalty, the structure inventory theory calls the newsvendor problem: order too few and you lose the margin on sales you could have made in your best six weeks; order too many and you lose only the difference between cost and clearance price. When the margin is fat and the clearance loss is thin, the arithmetic tells you to over-order deliberately. Planned leftovers are not a forecasting failure. They are the cheaper of the two mistakes.
One more turn, and it is the one that keeps the cycle sharp. Once shoppers learn that January brings markdowns, some of them wait — which pulls demand out of the peak and into the trough, and tempts retailers to start their promotions earlier to catch it. The rhythm is therefore not purely external: the industry's own response to it feeds back and reshapes it. How far that goes is contested, and it varies enormously by category.
The analogy
THE ANALOGY #Think of a fishing port during a run that comes at the same weeks each year.
The harbour must have berths, ice houses and hands enough for the peak of the run, and it pays for them in the empty months too. When the boats land, the catch must be sold quickly, because a fish held is a fish worth less tomorrow — so the price on the quayside falls through the day as the unsold stock ages, not because the morning price was a fiction.
A fish spoils on its own in hours, whereas a sweater is physically fine in January — what expired was the occasion for buying it — which means a retailer really does have the option to warehouse and re-offer next season, an option the fisherman never has, and some do exactly that with staple lines.
Clarifying the model
THE MODEL #Discounting is not a confession about the original price. It is tempting to read the January sale as proof of a December markup. But the same object genuinely commands different amounts at different points in the cycle, because what is being bought in December includes timing — the gift must exist before the day. Price tracking value-in-context is ordinary, not deceptive.
Not all seasonal goods are equally perishable, and the discount depth follows that. A fashion item's value decays fast; tinned food's does not. Look across a store in January and the markdowns are steepest exactly where the seasonal clock ran hardest, which is what the account predicts.
The peak is not free money. Concentrating a year's margin into six weeks means the year's result depends on a short window, and a single bad December — weather, a supply delay, a soft consumer — cannot be made up later. Seasonality raises the variance of the business, not just its rhythm, and that is why seasonal retailers borrow heavily in autumn and repay in January.
The falsification test. If markdowns are driven by the decay of value across the cycle rather than by inflated original prices, then goods with no seasonal clock should not follow the same markdown calendar. Staples, replenishable basics and everyday consumables largely do not — their prices move with cost and competition instead, which is the comparison that separates the two explanations.
A picture of it
THE PICTURE #How to readThe bars are units sold, the line is the average price actually realised, both indexed so the December peak reads 100 on the bars. Trace the bars up to the December spike and then watch the line: it holds nearly flat while demand climbs, then falls away exactly when the calendar occasion passes and the remaining stock has nowhere else to go. The shapes are illustrative of the pattern, not measurements of a particular retailer.
What became clearer
WHAT CLEARED #Seasonality is not a scheduling problem a better-run shop would smooth away. It is a cycle imposed from outside, and nearly everything odd about seasonal retailing follows from two consequences of that: capacity has to be bought for the peak and paid for all year, and stock has a deadline after which its value collapses. The January sale, the deliberate over-ordering and the frantic December are not three separate quirks — they are one rhythm seen from three angles.
Where to go next
ONWARD #- How markdown-optimisation systems decide the timing and depth of each cut.
- Whether promotional creep genuinely shifts demand or merely relabels it.
- The same cycle logic in airline seats and hotel rooms, where the stock expires at a known instant.
Key terms
TERMS #| Term | What it means |
|---|---|
| Newsvendor problem | the classic single-order inventory decision under uncertain demand, balancing the cost of ordering too few against the cost of ordering too many. |
| Salvage value | what unsold stock can be recovered for once its selling season has passed. |
| Markdown cadence | the planned schedule of successive price cuts used to clear seasonal stock. |
| Phase-locked demand | demand tied to fixed calendar events rather than moving with a business's own decisions. |
Every term the collection defines is gathered in the glossary.