Forecasting & Demand Planning

Everything upstream in a supply chain — how much to make, how much to stock, how many people to roster — depends on a guess about the future: how much will customers want? Turning that guess into a number is forecasting, and shaping the business around it is demand planning. Done well, it is the difference between shelves that are stocked and warehouses that are not overflowing. Done badly, you get stockouts and write-offs in equal measure.

What a forecast actually is

A demand forecast is an estimate of how much will be sold in a future period — next week, next quarter, next year. The most common starting point is history: what sold last week, last month, this time last year. If you sold 500 units each of the last four weeks, a reasonable first forecast for next week is around 500. From there, planners adjust for things the raw history misses — an upcoming promotion, a holiday, a new competitor. The forecast is a blend of what happened and what you expect to be different. A supermarket planning for a long weekend will lift its bread and barbecue forecast above the plain four-week average, because it knows the calendar is about to change customer behaviour in a way last month’s sales cannot show.

Why demand is hard to predict

The honest truth of forecasting is captured in a well-known saying among planners: every forecast is wrong. The only question is by how much. Demand is hard to pin down because real life is noisy — weather, news, a viral post, a competitor’s price cut, the economy. Some patterns help, like seasonality (ice cream sells in summer, coats in winter) and trend (steady growth or decline over time). But layered on top is plain randomness no model can foresee. So the goal is never a perfect forecast; it is a forecast good enough to plan around, plus a cushion for when it misses.

Forecast error: measuring the miss

Because forecasts are always off, planners track forecast error — the gap between what they predicted and what actually sold. If you forecast 500 and sold 540, the error is 40. Tracking error over time tells you whether your forecasts are reliably close or wildly swinging, and it reveals bias: if you are always too low, your method is systematically optimistic and needs adjusting. Measuring the miss is how forecasting improves. A team that never checks its error keeps repeating the same mistake; a team that reviews it each cycle slowly tightens the gap and learns which events its model keeps missing.

Planning for uncertainty, not certainty

This is exactly where the earlier ideas connect. Because demand is uncertain, you hold safety stock to cover the times the forecast falls short, and you value short lead times because they let you re-forecast closer to the event, when you know more. Good demand planning does not pretend to know the future; it prepares for a range of futures. The forecast points you in the right direction, and the cushions absorb the difference between the guess and reality.

Every forecast is wrong — so plan for a range, track your error, and keep enough cushion to absorb the miss without disappointing customers.

Spot the forecasting concept

Read each situation and identify which forecasting idea it illustrates — then tap a card to check.

Sort the forecast challenges

Drag each situation into the bucket it belongs to — improve the forecast, account for seasonality, or use safety stock. Hit Check placement when you’re done.

How to use it

Speak about forecasts as best guesses to be improved, not promises to be defended, and you will sound like someone who has done it.

• “History says ~500 a week, but we have a promo, so I’d forecast higher.” (blends data with known events)
• “Our forecast error keeps running low — we’re biased optimistic.” (uses error to find a systematic flaw)
• “Demand here is seasonal, so compare to last winter, not last month.” (matches the pattern to the period)
• “We can’t predict the spike, so we carry safety stock for it.” (plans for uncertainty instead of denying it)

Treat the forecast as a starting point you measure and refine, and the whole supply chain has something solid to plan against.