During the past few years, many have observed traditional Master Data Management (MDM) evolve over time, especially when discussed in relation to analytics. Historically, MDM projects have focused on creating a single view of the truth for consumption by downstream business processes. By managing and de-duplicating related entities from multiple sources across medium and large enterprises, one can attain significant cost savings and competitive advantages in any one of several ways:
As a result, MDM continues to be a core technology foundation for multiple initiatives. However, in the age of big data and analytics, organizations are looking to gain a greater competitive advantage by utilizing their master data investment to provide added trust and accuracy to their analytical systems.
Analytics are only as good as the data you feed into it. If your analytical systems are fed by disparate, poor-quality, duplicated data, then the analytics and business decisions that are made from those analytics could lead to incorrect and ultimately damaging decisions.
Organizations are looking to leverage their MDM projects to extend beyond just operational usage into analytical usage, which IBM refers to as entity analytics. Being able to utilize a single view of key business entity data and knowing it is a high-quality, trusted and highly accurate source available within the enterprise can significantly improve the understanding of the other data you’re utilizing for analytics. This entity data could include account, customer, employee, partner, product or supplier information.
Master data used within an analytical context can connect the dots between previously unknown relationships among entities such as who lives in the same household, works for the same company or bought the same product. It can unlock and make sense of otherwise dark data assets by embellishing the data assets with attributes from the master entities, to provide context that makes the data assets understandable and relevant.
As a result of the recognition of the power that master data can bring to analytics, a number of innovations and evolutions can support the move to entity analytics.
Graph database technologies that support effective storage, discovery and manipulation of relationships between entities are seeing a marked rise in interest. Machine learning advancements enable data to be automatically discovered, classified and managed. And the growth of the application programming interface (API) economy can open up data to application developers for access in new and innovative ways.
A new breed of technology-savvy business users expect to be able to get access to data with a set of self-service tools and get on with their jobs without the need for IT. They have a job to do working directly on the data, and they don’t have the time or will to wait on anyone or anything else. In addition, traditional MDM has been very specific in handling core business entities and attributes that were considered master data. In the world of entity analytics, this definition is blurring. If a business user wants to augment MDM data with operational or transactional data, then that user should be able to do so.
The volume of data now available to business users who want to run analytics on it continues to grow. Enterprise-wide data lakes now empower users to go fishing for data that previously would not have been available to them. Cloud data providers and open data sets are more available then ever before and offer significant value when being used in analytics. Add to this level of access the ever-increasing volume of social media data. And when you consider including unstructured data—documents, images and so on—you can see how a huge amount of new data is now available that can benefit from being related to the core master data when used for entity analytics.