Although the concept of Master Data Management (MDM) is nothing new, it is often overlooked in many organizations. However, if implemented properly, an MDM solution provides many benefits and is a cornerstone of an effective enterprise data strategy. The aim of this post is to provide a high level overview and discussion of the benefits of MDM, while future posts will discuss tools and technologies used to implement MDM.
What Master Data Management Does
At the high level, an MDM solution provides a central repository of data critical to the organization along with the policies, processes, and tools to manage this data. The data itself can be anything the organization feels is important enough to master, although typical examples of master data would include things like customer or product data. Actual transactions such as sales orders, medical claims, etc. would NOT typically be included in an MDM solution. A typical MDM solution will combine data from multiple data entry systems. For example, an MDM repository may combine customer data from an ERP system as well as a CRM system. An MDM can, when needed and appropriate, also be used to assist in synchronizing data between these systems.
The biggest benefit of an MDM solution is that it provides a centralized, consolidated repository that serves as the authoritative source of master data. Data in the MDM repository is consolidated from multiple sources, cleansed, and de-duplicated. Business rules and logic to transform data is applied at the MDM, eliminating the need to duplicate logic in every application (and thus eliminating the possibility of logic being applied inconsistently). The result is a data repository that is guaranteed to be clean, complete, and correct. This in turn guarantees that data in applications and processes developed against the MDM will be likewise be consistent, complete, and correct.
An MDM simplifies development by eliminating the need to connect to multiple data sources, as well as eliminating the need to clean and transform data. Doing this also reduces strain and contention on the source system(s) that would be created if applications and processes were developed directly against these source systems. Additionally, developing against an MDM repository decouples downstream applications from the source systems, which can dramatically reduce future development efforts when new system(s) are adopted and implemented.
For example, suppose you have a web portal and a data warehouse that connect directly to an ERP system. If a new ERP system is implemented, these downstream systems need to be rewritten against the new system. With MDM, the only point of integration that ever needs to change is between the source system and the MDM. All downstream applications and process remain unchanged, which can result in huge savings.
Cleaning and Combining Data
A common question when talking about MDM, especially during business intelligence engagements, is “Can the data warehouse handle the combining/cleaning of data?” The answer is that it can; however, a data warehouse is designed and optimized specifically to be consumed by reporting and analytic processes in a (typically) read-only fashion. An MDM repository can provide data to a data warehouse, but also interact with other applications. If the data warehouse is not the only system that needs a source consolidated/cleansed data, then an MDM solution should be considered as part of a more comprehensive data strategy.
When to Keep a Separate Copy of the Data
A closely related question that sometimes comes up is whether it is necessary to maintain a separate copy of the data. Remember that an MDM repository is a system that sits upstream from other systems and processes and serving a separate purpose. Aside from a data warehouse, maintaining a copy of the data separate from the source system(s) yields the benefits outlined in this post.
Many organizations choose to simply use their ERP system as their master data repository. This can be an acceptable solution if consolidating data from other systems is not a concern and the ERP system contains all of the desired data points and business rules. However, keep in mind that if these needs (or the ERP system itself) change, implementing MDM may become necessary and may be much more costly if there are significant downstream applications and processes to rework. Keeping this point in mind early on can save a great deal of heartache.
It is true that implementing MDM can require significant effort. However, as part of a larger data strategy MDM can bring substantial benefit and long term ROI. In this article, I’ve attempted to provide a high level explanation of the benefits of having an MDM solution. In future posts we will discuss some of the organizational considerations of MDM, as well as some of the tools and technology used to implement MDM.