Most forecasting is last month plus a bit. This uses your own history including seasonality, states the assumptions, and shows how wrong the last forecast was.
What's Included
- A forecast from real history including seasonal pattern.
- Assumptions stated so the forecast can be argued with.
- Accuracy of past forecasts measured rather than forgotten.
- An assistant that explains the drivers behind it.
How To Use
- Use at least two years where you have it, so seasonality shows.
- State assumptions. An unexplained forecast is not trusted or improved.
- Measure past forecast accuracy. It is the only way it improves.
- Forecast a range, not a single number.
Turn it into staffing with seasonal labour planning, and into stock with purchase order automation.