Designing an expert system to support competitiveness through global sourcing

Performing global sourcing – which is defined as the amount of purchases made by a plant outside its continent of origin – effectively is highly complicated because of several concurring variables (e.g. number of suppliers, supplied goods, and the location of the buyer and supplier, among other factors). However, the findings available in literature – which are often focused on few variables – can be difficult to be applied by managers to take decisions in this domain. Because of this limitation, we built an expert system (called global sourcing expert system [GSES]) to suggest what type of supply management (e.g. supplier base management, supplier selection and supplier integration) practices are most suitable given the level of global sourcing, strategy and other characteristics of the company. To build the theoretical grounding of the GSES, we performed a deep literature review to identify the highest number of variables that can be related to a global sourcing strategy. Next, we built the GSES exploiting data from the International Manufacturing Strategy Survey (IMSS) as a knowledge base. This data-set is composed of 725 companies surveyed in 2009 in the assembly manufacturing industries (ISIC code range 28–35) distributed across 19 countries. Finally, the system was validated by means of simulations and five real cases to verify that the expert system was actually usable by practitioners and provided robust insights. The paper also includes a detailed description of how the tool is built to allow future replications.