Planning stock for a season you cannot forecast

A season is a calendar, not a forecast. Demand for Ramadan stock, school supplies or summer goods arrives because the date arrives, and its size is decided by households you will never see. So the useful plan is not a number — it is a shape: what you commit to, what you keep cancellable, when the reorder rules move, and what you will do with what is left.

Where does a season plan come from, if not a forecast?

From last season's actuals, per item and per week. Not total revenue, which hides the mix: two items can trade places and the total looks identical while the shelf fills with the wrong one.

Then read the sell-outs honestly. An item that emptied early tells you what you could supply, not what people wanted: its sales figure is a ceiling. The same goes for anything that moved only after a discount: it sold at a price you would not repeat.

How is a season different from an ordinary month?

The date is known and the size is not. That asymmetry is what makes a season plannable: the calendar goes into the system, and orders and staffing follow it. An ordinary month is planned from a rate of sale; a season is planned from a date plus a range, and the range is managed week by week.

The second difference is that everyone else is buying at once, so lead times stretch exactly when you need them short.

What is the difference between stock you commit to and stock you can still cancel?

That distinction is the main lever, and it is a purchasing decision.

Kind of stock	What it costs you now	What you can still do
Firm order, delivered	cash and shelf space	sell it, or discount it
Ordered, inside a cancellation window	a promise, sometimes a deposit	cut it before the date
An indication to the supplier	attention only	change the quantity weekly

So the buy needs a firm base and a cancellable upside. The base is what you would still sell in a weak season; the upside is ordered as a right rather than as stock — an agreement with a date before which you can cut the quantity. Committing the whole season in one order to win a better price trades that flexibility for a margin you cannot measure until the season is over — and the forecast error is usually larger.

How much cash ends up in the wrong item?

Every unit you over-bought is cash wearing a shelf. It is worse than an empty shelf: that loss is visible, this one is invisible and consumes space and attention.

Ask how much of the buy sits in the few items that carried the previous season. If the budget is spread evenly to be fair to every product, the winners run out early and the rest waits for the discount. An honest plan looks uneven.

When does the reorder point have to move?

Before the season opens, as a dated change rather than a remembered one. A reorder point tuned on an annual average is wrong in both directions: too high in the quiet months, parking cash; too low during the season, emptying the shelf. It is also built on the wrong lead time, because the supplier is serving everyone at once.

So the season setting has two parts: the expected daily rate, and the lead time to assume while the supplier is busy. Both belong in a version of the rule with a start and an end date, so the system switches and nobody has to remember. The same applies to the target cover, meaning the days of stock you accept holding.

What do you do with the leftovers?

Decide before the buy, not in the final week: the exit for a category is part of the purchase.

The options are not equal. Carrying stock into the next season costs storage and risks a style that has moved on: right for stable goods, wrong for anything tied to a date. A discount at a planned date is cleaner than an improvised one, because it can be announced and budgeted. A supplier return is the cheapest exit, negotiated when the order is placed, while you still have leverage. Writing it off is sometimes the honest answer, and cheaper than the shelf space plus the labour of moving the cartons again.

What to avoid is discovering the leftovers. They are not proof the buy was wrong; they are proof the exit was never planned.

What order do you plan this in?

1. Pull last season per item and per week. Mark which items sold out and when; those figures are a ceiling, not demand.
2. Split the buy into a firm base and a cancellable upside. The base sells even in a weak season; the upside is a right you can abandon.
3. Put the cancellation window and the delivery dates on the order. Both belong in structured fields, not in a message thread — UBL's order document carries them as part of the order.
4. Move the season settings before the season, dated. Reorder point, lead time, target cover: a version with a start and an end.
5. Reserve cash for the upside, not the base. The base is spent; the upside is still a decision.
6. Set the exit for each category before the buy is placed. The date, the discount authority, and the person who decides.
7. Manage weekly on rate of sale, not on stock value. Stock value says what you own; rate of sale says what to do next.

Where this goes wrong

• The plan is built from total revenue. The total looks fine while the mix is wrong.
• Last season's sell-out is reordered as it stands. It repeats the shortage on purpose, because the number that emptied the shelf was supply, not demand.
• The reorder point stays on its annual setting. The opening weeks empty the shelf, and the rule reacts after the fact.
• Leftovers are stored as though storage were free. The real cost is the cash and attention that cannot go into the next buy.

What this does not cover

This is not statistical forecasting and makes no claim to predict a peak. It assumes last season exists at item level; a business opening a category for the first time has no anchor, and should plan smaller, keep more cancellable and set the exit earlier rather than guess harder. It does not cover promotional pricing, supplier negotiation, or the financing of the buy, which for many small businesses is the binding constraint. Religious and national dates are treated strictly as demand windows: the calendar is what makes the plan possible, not a marketing opportunity.

We build the seasonal plan — item-level actuals, split commitments, dated reorder rules and a weekly rate-of-sale view — so a season is run from dates and evidence instead of a guess.