Forecast Accuracy Challenges Faced by Intermittent Demand

Thrive’s Groupcast aggregate forecast module was designed to help wholesaler / retailers that face the challenge of having to stock SKU’s with intermittent demand. For example, automotive and marine parts wholesalers have to stock many items with low volume and intermittent demand due to the sheer multitude of models and makes and years of vehicles and boats.

Creating accurate forecasts for SKU’s with low volume, intermittent demand, or ‘lumpy’ demand is definitely a challenge for most wholesaler / retailers. Is there any seasonality? Are sales trending up or down? These questions are very difficult to answer when you only sell the item a couple times a year. There are scientific formulas used by Thrive’s forecast engine to help better forecast intermittent demand (egCroston’s). Nonetheless forecast accuracy is difficult to attain.

Groupcast Aggregate Forecast for Increased Forecast Accuracy

To increase forecast accuracy, Thrive has developed a proprietary process called Groupcast, which groups like items with intermittent demand together to create an aggregate forecast which gives our forecast engine more data with which to identify seasonality and trends. Then the aggregate forecast can be pushed back down to its individual like items to increase the forecast accuracy.

The Impact of Aggregate Forecast

We put this automated process in play at a $600 million automotive parts distributor because 70% of their stocked items had low volume intermittent demand. We increased their forecast accuracy by 15% at their velocity D, E, and F items (primarily low volume intermittent demand), reducing their lost sales by 72%. Some of these items were high dollar items as well (eg engines) so increasing the forecast accuracy with Groupcast’s aggregate forecast resulted in more sales with a significant dollar impact.