Q&A: How to Achieve Master Data Management Success
MDM and governance go hand in hand, but how do you know if your MDM approach is working, especially in light of Big Data and social media?
- By James E. Powell
Master data management (MDM) and governance go hand in hand, but how do you know if your MDM approach is working, especially in light of Big Data and social media? What is the best way to implement your MDM program, and how does MDM ensure data quality? For answers to these and other MDM-related questions, we turned to David Corrigan, IBM Director of Strategy for the company’s InfoSphere Portfolio.
BI This Week: What's the nature of the relationship between governance and master data management? How should companies relate the two initiatives?
David Corrigan: You cannot do master data management without governance. MDM unites multiple users and data sources; governance creates an agreement on the rules of interaction among systems. Governance enables MDM’s success by providing business context and frameworks and ensures that MDM is not treated as a simple IT project. It brings users together to discuss business rules for data usage. MDM, in turn, makes governance more relevant because those governance policies become tangible. You unite the theory of governance with the application of MDM.
What are the top business processes and applications that need master data management?
The top processes and applications that need MDM are new account openings, customer service and retention, Web self-service and data warehousing. Additionally, MDM is helpful for new product introductions and product bundling. What is common among these situations is the generation of multiple records about the same customer, product, or service. All these examples either can create more value out of a customer relationship or product offering or lose opportunities because of duplicate records.
How do enterprises typically go about implementing MDM? What problems surround that approach, and how do you recommend enterprises implement the technology instead?
There is no standard pattern. However, companies frequently treat Master Data Management as an IT project when it is a business strategy with IT deliverables. Companies often start MDM with a “bottom- up” approach by identifying pain points and centralizing data from various sources. In order to succeed and drive MDM adoption, those companies should think of consumption-centric approach -- to define who needs master data, then what data they need, and only then define the sources of that master data. You also have to link governance and MDM as they provide the policies and tools. Governance creates the strategy and best practices for handling data, MDM provides the resources required to work with this information.
Some standard questions organizations should answer when implementing MDM include:
- Which business applications and processes are hampered by a lack of master data, and what data do they need?
- What is the master data for your organization and what does it mean to you?
- Who are the key stakeholders to your MDM initiatives?
- Why should an organization focus on MDM now?
- Where is master data physically dispersed throughout the organization?
- When will we able to show results?
- How do you start the MDM initiative and how will the governance program leverage the initiative for success?
Can you implement MDM in stages? If so, what are the pros and cons, and what is the impact of this approach?
Proceeding with master data management in stages is the only way to implement it. The impact of this approach is that planning beyond Phase 1 is critical to the success of your MDM initiative. MDM is a journey not a destination. The benefits to this approach include having a quick win and the ability to generate momentum and excitement among the team. There are key considerations around the people and processes. Think about who is ready for the change that MDM will bring and if these individuals have the ability to adopt it. Also, link phase 1 and phase 2 in a cohesive manner to generate momentum.
Finally, make sure the timelines will not be too short, but at the same time deliver business value. Once you see value from Phase 1, the organization will want to deploy MDM across the business
How do you know if your MDM approach is working? How do you know when it’s time for a reassessment?
One question I recommend asking is if you have completed phase 1 but haven’t advanced to phase 2, why not? MDM should have a series of stages. Ask yourself why MDM is a point and not an enterprise solution. Is it the people, the processes, or an underlying technology problem that doesn’t’ facilitate true enterprise deployments? A substantive governance strategy can often kick-start it again.
One true sign of MDM not working is when you always have to build functionality in each phase vs. turning on MDM features. This looks innocent enough in phase 1 and 2, but ask your business team what they think of the enhancement schedule of their legacy systems and how responsive IT is. That could be your future if you build out everything in MDM or a big risk if you don't architect it for change.
What unique challenges does Big Data present for MDM and the information governance process?
Conceptually, it is very similar to the relationship between master data management and data warehouses (DW) -- both the DW and Big Data platform are designed to find insights. Big Data is simply for a larger variety, velocity, and volume of data than before. This explosion of information is often focused on the consumer or the brand, and what is the market saying about the organization. To answer that question, you need a single view of your customers and products. Which customers do you care to follow in social media? Often I speak with our clients about what to do with the insights gained from Big Data and how to operationalize it. Specifically, how do you put it back in MDM to affect business processes to address data quality and reconcile differences in records?
How has the rise of social media affected information governance strategy?
Organizations are eager to tap into social media for new customer insights and insights into the market perception of products. It’s like an information gold rush. However, governance needs to be a part of this; it isn't the Wild West. Many companies are hesitant to start because they don't know how customers will perceive them using social media to learn more about them. One approach we recommend to our clients is asking your customers about their preferences. Organizations need to implement a cohesive social media and governance policy on how you will use that data, and how a customer will benefit. Master data management is the place to store that consent, if you receive it.
This whole trend does introduce new challenges. There are a variety of new data sources that brings new requirements for governance technology. The overall governance process remains the same. The framework for business and IT working together becomes larger because you are working with Big Data. It is making that information useful and empowering your employees with straightforward governance policies they can employ.
What final thoughts or best practices can you offer for implementing and using MDM and for ensuring data quality?
Understand your business problems and create a road map where you want to go to address those issues. The master data management journey involves striking a balance between getting phase 1 completed in a reasonable time without impacting phases 2 or 3. You have to set expectations appropriately. If you shorten phase 1 to meet a time line, but may need more time to finish it up in a later phase, address this with your stakeholders.
One final thought: make sure you solve the core of the problem, the application or business process that initially made the data bad. Even though you may clean up a set of instances of bad data, you must solve the problem at its source. This is why MDM is strongly linked to business processes and apps.
What role does IBM play in MDM and data governance?
IBM helped define and develop the market almost a decade ago and since then we have focused on helping people -- people and processes and technology come together. In the area of people, IBM leads the InfoGov Community which has 2,000 thought leaders and practitioners who share best practices on information governance. In the area of people and processes, IBMers author books on the intersection of MDM and information governance as well as governance processes. In the area of technology, IBM leads in MDM and governance with well over 600 customers spanning banking, insurance, government, and health-care institutions with our InfoSphere Master Data Management product.