Design for implementation : redesigning a planning process in a multinational company

Abstract

As a consumer-packaged-goods company dealing with multiple suppliers, "Company X" (the company's real name has been disguised) has a clear need for an accurate raw material forecast. The current planning process has not managed to supply a forecast that will answer Company X's raw material procurement needs, and so a need for a new process has arisen. Process change initiatives are a task that every company frequently encounters, and creating the right setting for a successful change initiative is critical to determining whether the company will be able to drive the cost savings and remain competitive. This thesis outlines the process of redesigning the raw material planning process within the company's regional operations in several developing countries. The new process was designed in an attempt to balance creating a more accurate and more robust forecast with designing an implementable, wellaccepted new process that will establish the settings for a successful change initiative. In the process evaluation section, we examine the new procedure's potential accuracy improvement, as well as two unique management tools to evaluate the process' implementation potential and the trade-offs made in designing the process. We evaluate the impact of the finished goods forecast and production error on the overall material plan accuracy. One common course of action companies typically take in order to decrease material plan variability is to reduce the number of finished goods SKU by aggregating similar SKUs. SKU reduction increases the forecast mean of the individual SKUs and thereby decreases the percent error of the forecast. One interesting conclusion this thesis outlines is that when the production error is not negligible, SKU reduction will only help if the decreased finished goods forecast percent error is above a certain ratio (the square root of the forecast mean). This conclusion suggest that local optimizations of the sub-processes that generate the full material plan will not necessarily optimize the resulting material plan, and that material plan optimization should be done using a comprehensive approach, not as a set of sub-processes optimizations. While these results, and the methodology for the design and assessment of new process was revealed during my work within company X, the material plan accuracy results are relevant for any material plan process improvement, and the change management methodology discussed is relevant for any company with a need for a smooth and successful implementation of a new process.