Master Data Management Overview

Warning: Master Data Management Will Make or Break Your Business Intelligence Implementation

by Tom Oldham

Typical Scenario: The company is implementing a new business intelligence platform. You have been invited to attend a meeting to collect input from key stakeholders on what reporting and data is required. Is this a meeting that can occur without you? Why would a department VP care?

Answer: If you want to retrieve data easily and slice and dice it in a particular way to analyze your department’s KPI’s or make decisions, then to save yourself years of data analysis misery, attend the meeting and volunteer your department’s data expert to the ongoing team.

Master Data Management Overview

Companies struggle day in day out with cleaning data before they can analyze it or send it out as “actual”. There are many analyst hours spent on cleaning up or manipulating or structuring the raw data repeatedly. This is a symptom of weak master data management. Master data management is the rule book for how data is organized within the business system. For example, if you want the individual customer sites to roll up to a parent company, then you need an identifier defined for “parent company” when defining the customer profiles along with bill to and ship to addresses.

Develop a Master Data Strategy
It all starts with having a strategy for Master Data. This step identifies how the company wants to analyze the business as a whole and by each function. It includes all departments, levels of the business and stakeholders. The team needs to do the detailed work of mapping out every single record and coding system in the ERP. Then review how well the current data structure aligns with how the company wants to analyze the business. Is there important offline data? If there is, then identify a way to code it and load it. If there is no data matching the vision then challenge the team to create it. It will be worth it.

Define the Data Structure
Once the Master Data Strategy is defined, then it is time to refine and implement the data structures. Start with the most critical “pain points” identified in the process map. This may be as simple as adding a parent/child relationship to the existing customer record or as complicated as implementing a new smart model code structure. Also, put some standardized “intelligent groupings” around the business to ease the work required to understand trends and expose areas that need attention.

Many times, data is not in the system because it is too time consuming to load or maintain. This is when the IT department needs to develop automation tools to support configuration management. Also, when you hear a team member say it can’t be done, check with the ERP vendor or your BI vendor. ERP systems are much more sophisticated now and can handle complex reporting structures, supply chain strategies and pricing policies. There should be very little “offline” data and manual manipulation needed for daily, weekly, monthly, quarterly or annual reporting.

Implement a Governance Model
The final step for developing robust Master Data Management is establishing an ongoing process to ensure the maintenance and adherence to the structures. For this data governance model to work a dedicated resource(s) is needed. This resource develops and maintains the forms which control the addition of Customer, Supplier, Product, etc. data. They make sure the workflows are documented and followed. This resource will need support from IT to have automation tools to expedite the upload and download data process.

Seek out Experts
At Cyberscience, we understand the complexity of ERP data structures and have been helping customers define their master data strategies for years. With 25+ years of experience delving deeply in the world of ERP systems to retrieve data critical for managers to run their business, we know how to support your organization in setting up the best practices of Master Data Management for a successful Business Intelligence Implementation. Give us a call and start your journey.

Author: Tom Oldham, CMA, CFM, is a Product Manager at Cyberscience focused on Manufacturing Solutions. This is the first of a 5 part series of blog posts about Master Data Management. In additional posts, Tom goes into detail about developing a strategy, defining data structures and implementing a governance model.