Analysis of predictive sorting and process standardization at Amazon.com

Abstract

In any process, complexity has the potential to decrease processing rates of the operation. Mass production has driven up operator efficiency by removing complexity to create a standardized and simplified process. However, removing complexity from a process reduces the operator's ability to customize their products and remain flexible to variations in demand. Businesses need to decide the appropriate level of process standardization that allows their operators to work as efficiently as possible while remaining flexible enough to satisfy customer demand. The goal of this research is to analyze the standardization of the least efficient process in an outbound order fulfillment system at Amazon.com. Various tests were performed during a six-month internship at Amazon's Lexington, KY facility to determine if this process could benefit from standardization. The testing showed that Amazon's processes must maintain a high level of flexibility to satisfy the variations in customer demand and the diverse mix of products that they ship. This research ends with a proposed process improvement that will improve the efficiency of the baseline scenario while providing flexibility to handle variation in order flow. During the internship, the per-order cycle time at the packing operation has been reduced by 9.7% and another improvement was tested which reduced cycle times by 10.9% when implemented. When the final improvement is implemented the total cycle time reduction will be 20.6%.