Building smarter supply chain through business analytics (A roadmap for IBM global supply)

Abstract

IBM Global Supply manages the overall procurement requirements of IBM and some of it clients. Their annual spend was close to USD 42 Billion in the year 2009. With such a huge magnitude of expenditure, this study looks at the various possibilities available for IBM to use business analytics can to gain sustainable competitive advantage in the operations. Five key areas under the purview of IBM Global supply were analyzed as part of this study to identify problems which could be addressed through business analytics. Model building and factor analysis was excluded from the scope of this project as it would require an extended period of study. As the first part of this study, business analytic implementations across the industry and sectors were identified through secondary research. Based on suitability to IBM Global supply, few implementations were selected for further analysis. After rigorous study a framework was developed on the basis of a Harvard study which could be used to arrive at problem areas to be targeted for deployment of business analytics. The framework was applied on IBM Global Supply through responses obtained during our interaction with them. The two areas identified for possible application of business analytics were: spike in demand and order fulfillment during the end of financial quarters leading to under-utilization of resources at other times and the potential to reduce inventory costs which also involved correctly predicting price movement for critical components leading to huge cash expense in the spot market. As part of industry trends, discussions were held with executives at HP Invent Bangalore office. Each of the problem areas was further broken to arrive at key reasons/factors which could hold the key to resolving or at least reducing the impact of the problems. First problem area that was targeted was to develop a lean manufacturing process with inputs from forecasting models determining the pricing of critical components. An example of DRAM memory was used to explain, how predictive analysis for forecasting future prices of DRAM can reduce the expense as well as planning a lean manufacturing process with minimal inventory in terms of savings between holding cost and spot purchase cost.