Student projects focus on supply management segmentation, risk analytics, and predictive cost models.

My MBA 541 students presented results from their projects this semester.

One team on a project for Bayer Crop Science focused on supplier segementation methodologies – reviewing current Bayer methodologies, and how they looked at “base”, “key”, and “strategic partners” for their Crop Science business.  The team looked at industry best practices and what the methodologies were, and also applied a segmentation framework to the cotton /seed category of suppliers.  They focused on beginning with category spend, and driving down to specific actions in terms of KPI’s, prioritization of resources, and optimization of the relationship through collaborative approaches.

Another team working with Hanes Brands also looked at different approaches around preferred supplier programs, a common theme.  HanesBrands supports many different brands including Champion, Playtex, Bali, and others.    The goal was to “re-balance” their supplier scorecards, and seek to identify how to identify strategic suppliers.  The team developed a survey launched with ASQ, ISM, and APICS, and looked at what the different approaches were.  The team was able to get 114 responses to their survey.  One of the issues they came up with was the relative weighting of different categories, which included the highest rated being quality, cost, risk, on-time delivery, and information sharing.  Companies are clearly using many more criteria other than cost and delivery to evaluate suppliers, indicating they are willing to pay a bit more if they are able to get better overall value from suppliers.

A John Deere team discussed how they attempted to build a predictive cost model.  The team sought  to predict cost per unit – using forecasted purchase orders, and trying to estimate average cost per unit based on weight, forecast quantity, finish, diameter, width, and angle.  They were able to find a positive correlation relating to the cost per unit.  This was a challenge, especially when the existing product specifications they looked at were in Japanese or German!  They then looked at statistical outliers, and tried to see if there were any that didn’t make sense, or why the costs were higher or lower.  They were able to derive an R-squared of 0.80, which is excellent given the number of data points they started with.  The team also concluded with recommendations on future data quality recommendations, and the importance of coding parts effectively for future predictive modeling.  Examples include material codes (steel, cast iron, plastic) as well as other part characteristics.

Another team conducted a project looking at vendor analytics for a large financial services organization, and sought to build a set of key risk components.  A framework was developed using both quantitative and qualitative analytic methods.  They had to begin to understand the components that would affect vendor disruption.  The creation of a relationship diagram identified elements of the vendor contagion/propagation along the supply chain, the direct and indirect impact, and the cycling damage that could occur.  A process model linking the scenario to the vendor, and teh delay that can occur, as well as the impact on the purchaser – provided the basis for an analytical model.

The last team focused on a supplier financial risk analysis for Hanes Brands.  The approach was unique, in that they sought to pull together approaches that transcended traditional vendor analytical projects.  They advocated that companies need to “be your own Credit Reporting Agency”, due to the lack of insight and visibility as to how these scores are derived.  Their recommended tool and approach combined several different supplier financial risk scores:

• Financial ratio analysis – Calculation of the 14 essential ratios to understand the financial health of a particular supplier. This category has a weight of 30% towards the final score.
• Altman Z-Score – A popular score used across the financial industry to predict bankruptcy of a company in question. This category has a weight of 25% towards the final score.
• Credit Reporting Agency (CRA) – The score taken from various CRAs will be standardised to a scale of 0-100 and added to the final score. This category has a weight of 25% towards the final score.
• Geo-Political Risk – The risk score is calculated by taking into account the following measures: country economic risk, ease of doing business index, failed state index and corruptions perception index. This category has a weight of 20% towards the final score.

This was a great set of applied research methods and approaches!  Well done, Class of 2015 MBA Supply Chain students!