Products
Forecast
Let what you used before tell you what to order
forecast.txtas is → to be
Today an order is placed on experience, by phone, and then waited for — and runs out twice, with a rush order each time. Afterwards what was used feeds a forecast, and the order follows the forecast.Illustrative hours, not a measurement.
Ordering by experience works until the person with the experience is on holiday. Forecast reads what you actually used, week by week, and tells you what to order and when — for materials, for capacity, for the shelf. You still decide; the number just stops being a guess.
The person who knows the business stops being the only place the ordering decision lives, and the history you already record becomes the first draft of it rather than something nobody reads.
How it runs
- a forecast from recorded movements, stated as a range rather than a number
- the assumptions behind it visible next to it
- a person approving the order, not a system placing it
- the items where experience beats the history left to experience
Where it does not fit
- items with no meaningful history, such as a new product line
- demand driven by a single contract rather than by a pattern
- anything where being short is unacceptable and holding more is cheap
Size: a medium piece of work — an estimate from the pattern, not from your business.
What we would want to know before committing
- which items are worth forecasting and which are not
- how far back usable movement history goes
- lead times, and what happens when one is missed
- who currently decides, and what they take into account
- check the movement history is complete enough to forecast from
- compare a forecast against what was actually ordered last season
- agree who is accountable when the forecast is wrong