The forecast window: seeing demand before it lands

You cannot buy stock for the demand right in front of you. By the time you notice a product selling faster than planned, place the order, and wait for it to arrive, weeks have gone. Every inventory decision you make is a bet on demand you cannot yet see. The forecast window is the exact size of that blind spot, and knowing it changes how you plan.

Most stores forecast on a calendar: a monthly plan, a quarterly buy. Demand does not run on your calendar. It runs on your supplier's lead time and on how often you actually sit down to reorder. Add those two together and you have the only horizon that matters, the distance between deciding today and being able to sell what you decided on.

What is the forecast window?

The forecast window is how far into the future you are forced to predict demand every time you commit to stock. It has two parts. The first is supplier lead time: the days between placing a purchase order and having sellable units on the shelf, including production, freight and inbound handling. The second is your review period, which is simply how often you place orders. If you review stock and reorder every fortnight, then on any given day the next chance to correct a mistake is up to 14 days away, on top of the lead time that follows it.

Stated plainly, the window is the sum of the two. This is why two stores buying the same product from the same factory can face very different planning problems. A store that reorders weekly sees a shorter window than one that reorders monthly, even with an identical supplier. Demand forecasting is only ever as good as the horizon it has to cover, and part of that horizon is something you control.

How far ahead can you actually see?

For most Shopify stores buying from regional suppliers, the window lands somewhere between 30 and 45 days. A fast domestic supplier can be shorter, and overseas sea freight much longer. Long enough that you are always planning around a month of uncertainty, short enough that a single soft forecast does not sink a season. Move to overseas sea freight and the window stretches past 50 or 60 days, which is why importers live or die on forecasting discipline in a way domestic sellers do not.

What the window forces you to commit

The window is not academic. It sets two hard commitments, and both are made blind. The first is inventory. To avoid a stockout you must hold enough days of cover to survive the entire window plus a safety buffer, which means your reorder point has to trigger before stock falls low enough to matter. Order too late and you sell out mid-window with no way to refill. Order too much and you tie up cash in inventory turnover that will not clear.

The same window governs media. If a product takes 44 days to arrive, then the demand you generate with advertising today has to be servable 44 days from now. Scaling spend on a hero product two weeks before it goes out of stock is how you pay to create demand you cannot fulfil. Your open-to-buy and your media plan are the same decision viewed from two sides, and the forecast window is the clock they both run on. Watch sell-through rate against the window: if a line is selling through faster than the window can refill it, that is the signal to either reorder now or pull the spend back.

How do seasonality and weather extend the window?

A flat window assumes flat demand. Real demand is not flat, and the two biggest distorters are seasonality and weather. Seasonality is predictable. If your peak is in December and your window is 45 days, your real commit date is mid-October, not December. The window effectively pushes your deadline backwards through the calendar, and missing it means arriving to a peak you can no longer supply.

Weather is the harder one, because it moves demand inside the window itself. A heatwave pulls demand forward for anything seasonal by days or weeks, and a cold snap does the reverse. For weather-sensitive ranges the window you plan against is not fixed, it breathes with the forecast. This is why Blufire runs weather-demand modelling: it reads how a store's own sales respond to temperature and conditions, so the forecast window can flex before the weather lands rather than after the stockout.

The discipline is simple to state and hard to hold: know your window, count backwards from it, and make every inventory and media commitment against the demand you will face when stock lands, not the demand in front of you today. Get the window right and the rest of the plan has something solid to stand on. The full method, including the margin arithmetic that decides how much cover is worth holding, sits in The Math.