Demand Forecast Calculator

Divide a month’s historical sales by your average month. If last November did 5,400 units against a 3,600-unit average month, November’s index is 1.5; a sleepy July at 2,880 units is 0.8. One prior year gives a workable index; two or three years averaged is better because it smooths one-off spikes like a viral post or a stockout that suppressed sales.

The average month-over-month growth of your last 3–6 months, after stripping obvious distortions — a big promo month, a stockout, a marketplace suspension. Compounding makes this input powerful: 5% monthly is 80% a year, so an optimistic guess quietly inflates a 6-month forecast by hundreds of units. If growth is noisy, forecast with a conservative and an aggressive rate and buy somewhere between.

Good enough for purchase orders on established SKUs, which is its job — typically within 10–20% a quarter out when demand history is stable. It is an estimate that degrades with horizon and volatility: new launches, viral products, and heavily promoted SKUs break the trend assumption entirely. The discipline that matters more than the formula is re-forecasting every month and letting fresh actuals correct you early.

It feeds every other number. Divide the forecast month by its days to get daily demand for your reorder point; multiply the horizon’s months to size a seasonal buy; use the annualized rate as the demand input to EOQ. The chain matters because ordering against last month’s sales walks you into every peak understocked and out of every peak overstocked — the classic bullwhip a simple forecast prevents.