Smoothie® includes an "expert" automatic time-series forecasting engine that generates accurate forecasts for thousands of items in minutes. Each time-series is evaluated for trends, seasonality or sporadic behavior. Then, the engine automatically develops a model for each series with optimal parameters.
The statistics are advanced, but you don't have to be a statistician to use Smoothie. The automatic forecasts will be excellent, and you may optionally guide the engine by suggesting use of trended or seasonal models, making top-down changes, or even forecasting with simulated histories (without losing visibility to actual demand). The engine also allows for capturing or creating pre-defined seasonal patterns for items. This capability can be extremely useful when there is not enough history or when SKU-level activity is too random to demonstrate known seasonal patterns. Accuracy is further improved with powerful level synchronizations and exception filters that seamlessly highlight items or customer forecasts that may require attention.
The statistical forecasting engine in Smoothie was designed for the "noisy" tactical data that prevails in product and channel-based sales and operations planning. It is extremely robust and does particularly well with the difficult-to-forecast time-series that prevail in industry. Many customers and prospects have benchmarked Smoothie's accuracy
against other commercial statistical engines and Smoothie routinely
achieves equal or superior accuracy in such comparisons. Not only does Smoothie produce excellent forecasts, but it also uses the error estimates, lead times and desired service levels to help businesses optimize inventory levels and prepare for the unexpected.