How to plan a peak-concentrated season

Measure the concentration before planning against it

Most teams know their season is peaky and very few have written down how peaky. The measurement is straightforward: take the last two or three comparable seasons, express weekly sales as a share of the season total, and look at the shape. What you want out of it is a concentration profile — how many weeks carry the majority of the volume, and how sharply the curve rises into them — rather than a single headline number.

Do it per channel. Channels rarely peak on the same date or with the same shape, and a blended curve describes none of them accurately. This is the input everything downstream depends on, and it is usually the cheapest analysis in the whole process because the data already exists.

Phase against the curve, not against the calendar

Even phasing with safety stock absorbing the peak is a defensible default when demand is steady. Against a concentrated curve it is not a compromise — it is two errors at the same time. Inventory sits through shoulder weeks that cannot sell it, and the peak still arrives under-served. Phasing against the measured curve fixes both with one change, and it does not cost anything the plan was not already going to spend. See how to plan receipt flow for the identity that turns a phased sales plan into dated receipts.

The last-receipt date decides the end-of-season position

Find the last date a receipt can land and still sell through at full price inside the peak. Subtract the full lead time — production, transit, and any approval, inspection or compliance duration that sits on the critical path — and that is the last order date.

This date matters more than the first markdown date, and by a wide margin. Goods arriving after it are structurally destined for clearance regardless of how well the season is traded. Most end-of-season overhangs are created upstream, by receipts authorised weeks before anyone was thinking about markdowns at all — which is why a run-out conversation that opens with a markdown schedule has usually already accepted a position it could have avoided.

A reserve you cannot deliver with is not a reserve

Holding back a share of the budget for in-season response is standard practice, and in a concentrated season it needs an arithmetic test rather than a conventional percentage. If the full lead time is longer than the weeks left to sell, the reserve cannot become sellable inventory this season. It is next season’s buy. That is sometimes the right decision — but it should be named as that, because a reserve described as a chase fund creates a false sense that the season is still steerable when it is not. Size it against deliverable weeks, and if the answer is zero, spend it in the original buy where it can at least be phased.

Pre-decide the shoulder, line by line

After a peak, lines that look identical in a stock report behave completely differently. An evergreen item should return to its normal coverage and will sell at full price for months. Occasion-specific or dated stock has a genuine reason to clear, and holding it costs more than clearing it. A single post-peak markdown applied across both converts planned margin on the evergreen half into an avoidable loss, and it happens constantly because the decision gets made on “post-peak inventory” as one category, under time pressure, with the position still settling. Deciding the two cadences before the peak costs an hour and is worth considerably more than that.

Set the re-forecast date in advance

In a concentrated season the first genuine read arrives late and is worth a great deal — and everything before it mostly measures anticipation rather than outcome. Fix the re-forecast date ahead of time. Doing so protects the decision from the pressure of the moment, and it forces an honest answer to the question that follows: by that date, which levers are actually still open? Usually the answer is allocation between channels and doors, and the exit cadence — which is precisely why steps three to five matter more than any in-season heroics.

• ✓A peak-concentrated season is defined by the absence of a correction opportunity, not by the height of the peak.
• ✓Measure concentration from your own history, per channel — a blended curve describes no channel accurately.
• ✓Even phasing against a concentrated curve produces two errors at once: overstocked through the shoulder and short at the peak.
• ✓The last-receipt date, not the first markdown date, is what most determines the end-of-season position.
• ✓If lead time exceeds the weeks left to sell, an in-season reserve is next season’s buy — name it as that.
• ✓Pre-decide two exit cadences: evergreen lines return to coverage, dated stock clears. One blended markdown loses margin on the evergreen half.

• How to plan receipt flow →
• How to set open-to-buy →
• How to read a WSSI →
• How to build a buy plan →
• Retail math formulas →

Frequently asked questions

What is a peak-concentrated season?

A season in which a large share of total demand arrives in a small number of weeks — gifting periods, holiday trade, and occasion-driven categories are the common cases. What makes it a distinct planning problem is not the size of the peak but the ratio of committed decisions to available corrections: the receipts serving the peak are usually ordered before any read exists, so the plan is close to the whole outcome rather than a starting position to steer from.

Why is even receipt phasing wrong for a concentrated season?

Because it produces two errors at once rather than a compromise between them. Spreading receipts evenly and letting safety stock absorb the peak means carrying inventory through the shoulder weeks that the shoulder cannot sell, and still arriving at the peak short of what it needs. A measured phasing curve costs nothing extra to build — the history is already there — and it addresses both errors with the same change.

What is the last-receipt date and why does it matter more than the markdown date?

The last-receipt date is the final date a receipt can land and still sell through at full price inside the peak window. It matters more than the first markdown date because it is upstream of the problem rather than downstream: goods arriving after it are structurally destined for clearance no matter how well they are traded. Most end-of-season overhangs are created by receipts authorised weeks before anyone was thinking about markdowns.

How big should the in-season reserve be?

Sized against what can actually be delivered inside the remaining window, not against a conventional percentage. The test is arithmetic: if the full lead time is longer than the weeks left to sell, a reserve cannot become sellable inventory in this season at all. Holding one anyway is a decision to carry budget into next season, which is sometimes right — but it should be named as that rather than described as a chase reserve it cannot fund.

When should you re-forecast in a concentrated season?

At the first read that reflects real demand rather than the build-up, on a date set in advance. Re-forecasting earlier mostly measures anticipation and produces confident moves on noise. Setting the date ahead of time also protects the decision from the pressure of the moment, which in a concentrated season is considerable — and by the read, the only levers still open are usually allocation between channels and doors, and the exit cadence.

What should be decided before the peak rather than after it?

The shoulder plan and the exit cadence, line by line. After a peak, lines that look alike behave very differently: an evergreen item should return to its normal coverage and will sell at full price for months, while occasion-specific or dated stock has a genuine reason to clear. A single post-peak markdown applied to both converts planned margin on the evergreen half into an avoidable loss, and that decision is far better made before the pressure than during it.