ERP and Supply Chain: Master Data Decisions

It is often surprising for a supply chain team to see how much master data they create and maintain when it is all collected and organized in one place. Because ERP generally provides increased functionality, there is also usually additional master data required to use that functionality. Without attempting to examine every master data field in ERP, we will discuss three broad areas, which will hopefully help you start thinking about supply chain master data issues.

1. Data ownership when there is overlapping responsibility

An easy example of this category is bill of materials. Many functional areas depend on the information in the bill of materials: finance, for product costing; supply chain for material demands; engineering for spec sheets or blueprints; development for new product adoption. Which of these functional areas should have ultimate authority over the numbers and relationships in the bills of material? Typically, in legacy, every functional area created their own version of a BOM, because every functional area had a different agenda which marginally affected the data. Costing wanted numbers that reflected the lowest possible product cost; supply chain wanted higher numbers to ensure they never ran short of anything. The best candidate for ownership of this type of data is the one with the least agenda, such as development. The paradox is, because they have no agenda, they have no real passion for accurately maintaining the data.

2. Yields and tolerances

This tends to be a greater issue in process manufacturing than discrete, but it needs reasonable consideration in both types of industries. In this context, “yield” is referring to the calculated expectation of how much first quality product will be produced on average from a fixed amount of components. “Tolerances”, in this context, refer to how much over or short you can be in filling an order, and still be of value to the customer. These two data pieces work in tandem as a hedge against manufacturing variation to determine how consistently you can satisfy customer expectations.

3. Computing rules

These are shorthand codes that tell the MRP portion of ERP how to behave. Each rule is generally understandable on a stand-alone basis, but as the rules begin influencing each other, the results –while always logical – can be complex, unexpected, and unwanted. These computing rules involve everything from how to treat safety stock inventory to whether a material is purchased or manufactured to whether a material is make-to-order or make-to-stock. To get good with these rules generally involves experimentation, rather than intuiting the setup based on the written explanations.

To the maximum extent possible, assign smart people to figure out how to set up supply chain master data as soon as legitimate testing can occur. These people don’t have to own master data forever; they just need to discover and document what the right settings are. In supply chain, master data has almost as big an impact on ERP performance as configuration does.