Forecast 3

Forecast 3

1. Select a Forecast interval

   • Safety Stock & Lead Time
   • Lot sizes
   • Seasonality
   • Trends

2. seasonally index

   • Is a measure of how a particular season through some cycle compares with the average season of that cycle

3. De-seasonalizing your demand.

   • We must remove the seasonality effect on our demand in order to more accurately identify noise and trend.
   • The object here is to identify all predictable elements of your demand so you can use these in your forecast.
   • If you have a predictable element (such as seasonality), remove it and then analyze your demand to see what other predictable elements can be identified.

4. Calculating and smoothing trend.

   • Subtract the basic smoothed forecast for the previous period from the basic smoothed forecast for this period. Since we are calculating our trend from the already smoothed fore- cast, our trend inherently has some smoothing in it, so it's possible we could just use this trend for our trend-adjusted forecast.
     • This tends to work better when you are using a lower smoothing factor, but since we are using a moderate-to-high smoothing factor, we are going to smooth our trend again.
   • Adjusting for trend lag.
     • Adjusting for seasonality.
       • Promotions and other events.
         • Overrides and adjustments

5. Group forecasting logic.

   • Grouping your products to apply a specific forecasting method or configuration
     • If you take the exponential smoothing forecast, you may find that even though all your products can use the same forecasting logic, they may not have the same demand profiles and therefore may require some different configuration of the forecast.
       • Smoothing factor used

6. More Forecasting Stuff.

   • Dependent demand and forecasts.
   • Managing demand variability.
   • Managing demand history.
   • Managing demand variability.
   • Forecast consumption.
   • Treating negative trends differently from positive trends.
   • Running multiple forecasts.