10 principles of forecasting

1. Forecasting is mostly wrong. Keep track of forecasting errors and methods. Errors are likely to occur, so make adequate provisions. Forecasting is more accurate for product groups than individual items.
2. Forecasting focuses on a time series. Determine the time series and evaluate the suitability of the time series against the objective of the forecasting.
3. Qualitative and Quantitative information. Evaluate both types of information and select the appropriate method(s). Present forecasting in different scenarios.
4. Weighted data. More reliable and relevant data should be weighted more heavily.
5. Forecasting should be independent. This will limit groupthink and promote independent thinking.
6. Unbiased individual input. Ask unbiased individuals to evaluate the forecasting process and output.
7. Keep it simple. Try to keep methods simple. Test employee understanding of the methods.
8. Record demand influencers. Record all events that might influence demand, such as floods and sales promotions. Describe all data assumptions.
9. Remove uncharacteristic events. Remove events from the baseline data, such as strikes and sales events, that are very unlikely to occur frequently.
10. Ensure data consistency. When recording time series, ensure consistency in data collection methods. Make sure information is reliable and check sources.