THE LEARNING SUPPLY CHAIN

Recently, Mike Sparks, the director of supply chain systems at Urban Outfitters, a global apparel retailer, looked for a way to get more out of his distribution centers (DCS). The growing demands of digital commerce and direct-to-consumer shipments have created a significant amount of complexity and volume for distribution and shipping operations, that just a few years ago, were predominantly focused on wholesale and store replenishment. As a result, one of Urban Outfitters’ flagship DCs was projected to reach capacity in the near future, and they expected to need another facility soon.

In an attempt to change that trajectory, Sparks and his team reached out to Manhattan Associates® to see if leveraging machine learning could help them to better match their demand to their inventory, resources, and automation, in order to manage more tasks and extend the use of their DC. Within just a few months, it was clear that the potential of an intelligent optimization technology called “Order Streaming” was significant – it was already generating large reductions in click-to-ship times and remarkable increases in picking throughput. By early 2019, Urban Outfitters was projecting the ability to extend the use of its DC by years. But how was Order Streaming getting these results, and why was it so much more effective than traditional approaches taken by most warehouse management systems?