A rapid what-if network planner tool for e-commerce fulfillment

Abstract

As online demand is growing, retailers are increasingly leveraging their store networks along with their e-commerce fulfillment centers (DC's) to fulfill online orders. Retailers use order fulfillment solutions, which are typically driven by a complex set of user-defined rules or a sophisticated predictive multi-objective optimization tool (recently launched IBM Order Optimizer) or a combination of both. However, as the fulfillment network expands from a handful of DC's to hundreds or thousands of stores, retailers are left with the hard question of how to plan fulfillment operations for the next promotional event or the peak season. Retailers are mostly left with the guesswork of either tweaking complex fulfillment rules based on the recent experiences or adjusting optimization parameters with a little view into the outcome. To address this challenge, we have developed a network planner tool, which is a rapid discrete event simulator engine that helps retailers answer a variety of "what-if" questions. The planner includes a framework to quickly ingest a retailer's historical data such as inventory, POS sales, e-com orders, network, etc. Depending on the user configuration, the planner uses a rule based or optimization based approach to perform order by order fulfillment simulation and generates KPIs for comparison.