Master data management is a business discipline - Consology

At the summit, where hundreds of MDM professionals gathered to share ideas, it was clear that world-leading companies have embraced MDM as a business necessity.

Organisations that are charting a journey into MDM should start by mapping organisational strategy and data governance before they consider the tools and technologies they will use. One of the hard lessons that many companies learn in their MDM projects is that MDM is not just a piece of software, but a business discipline that impacts processes, data, customers and governance. The biggest decisions and challenges in MDM projects are around culture, change management and data definitions rather than toolsets.

Embedded into MDM solution

MDM is rapidly moving beyond data stewardship towards convergence of task management, workflow, policy management and enforcement. For that reason, a complete MDM solution also requires rules and reference data to be applied across the customer, vendor and product data domains. The relevant data governance tasks, workflows, and policies should be embedded into the MDM solution to enable continuous data quality improvement.

"We can expect to see cloud-enabled MDM offer SMEs a means to engage in MDM without committing to long-term projects and major expenses. A large portion of their customer data is already in the cloud and mobile devices are redefining how this data is created and consumed, so the move makes sense for them," Lottering says.

"Data volumes are growing exponentially, and companies are likely to use their own data paired with public social media data and information from subscription databases to get a better understanding of their customers. We are seeing early adopters of this technology making use of existing MDM solutions as a registry hub that draws on social information."

A niche market

But emerging technologies are moving traditional MDM functionality, like matching and survivorship into public big data clusters. This is still a niche market with limited support from data quality tools vendors but expect this to change significantly in the next year or two.

Data virtualisation technologies that offer a unified, abstracted, and encapsulated view for querying and manipulating data stored in a heterogeneous set of data stores are starting to come to the fore. In other words, end-users will have access to one big single pool of data but won't know or care which underlying databases it comes from.