Inventory forecasting, without the mystique
A forecast is an estimate of how fast you will sell something. Most of the value comes from getting the basics right, not from the model.
Start with the right average
The single most common forecasting error has nothing to do with algorithms. It is dividing units sold by calendar days, including the days the product was out of stock.
Sell 30 units in 60 days, but the product was unavailable for 40 of them, and the honest rate is 30 ÷ 20 = 1.5 per day, not 0.5. Using the calendar figure teaches the system that a product which keeps selling out is a slow mover, so it orders less, so it sells out again. It is a loop that quietly buries good products.
What a forecast handles well
- Steady demand with normal variation.
- Gradual trends up or down, if the window is long enough to see them.
- Repeating seasonality, once you have more than one year of history.
What it does not handle
- Campaigns. A discount week produces a spike that is about price, not demand. Treated as normal demand, it inflates every forecast that follows.
- Anything that has not happened before. A first Black Friday, a new market, a product going viral.
- Things you know and the data does not. A supplier closing for Chinese New Year, a campaign you have already booked.
Which is why a forecast should never be the only input. The human should be able to see the deadline and act on knowledge the history does not contain — that is exactly the Black Friday case, where the numbers look calm right up until they do not.
How much history you need
Roughly: 60 days to estimate a rate, a full year before talking about seasonality with a straight face, and two before trusting it. Below that, a forecast is a smoothed average wearing a nicer name — useful, but not clairvoyant.
The part that matters more than accuracy
For an importer, being 10% off on the forecast is rarely what costs money. Missing the sea-freight deadline by a week is, because the fix is air freight at several times the price. Getting the timing right beats getting the quantity perfect.
Common questions
How much sales history do I need for a useful forecast?
About 60 days to estimate a sales rate, and a full year before seasonality means anything. With less than that, treat the forecast as a rough average rather than a prediction.
Should discount periods be included in a forecast?
Not as normal demand. A campaign spike reflects price, not underlying demand, and including it inflates every forecast that follows. Campaign periods should be identified and handled separately.
Related guides
- Looking for a Stocky alternative?
- Export your Stocky data before 31 August 2026
- Stock replenishment, end to end
Restocio plans purchasing for Shopify stores that import. It works out what to order, how much, and whether it should travel by sea, rail or air — so you pay air freight only for the units that genuinely cannot wait.