Why Life Sciences Must Go Beyond Master Data Management (MDM)
Many of the team here at Reltio formed the nucleus of Siperian (acquired by Informatica in 2010), the leading on-premises MDM tool widely adopted by life sciences companies. Back in 2005, master data management (MDM) was just taking shape and companies used MDM primarily to improve Siebel CRM data quality before upgrading and migrating their on-premises systems. Back then Siperian was preferred by many to the "seamlessly integrated Siebel Universal Customer Master (UCM)" offering, proving that best-of-breed solutions can be superior to integrated offerings that are designed for a single primary purpose.
One of the biggest issues we faced with Siperian (now Informatica MDM) was defining a relational life sciences data model that could capture not only the basic attributes of healthcare professionals and organizations, but represent real-world HCP-to-HCP, HCO-to-HCO and HCP-HCO relationships. Also thrown in for good measure was an emerging need to master product data, product hierarchies, groups and baskets for pricing and competitive analysis, and to feed product information (PIM) systems.
At Siperian we admittedly struggled with basic hierarchy management and performance issues with merge and especially unmerge. While we preached multi-domain and coined the term Universal MDM, we were never completely successful with standalone product master data management, let alone bringing together both customer and product data into a single consolidated Siperian Hub.
Back then, the best databases we had to model and store life sciences entities and their relationships were the likes of Oracle, DB2 and SQL Server. Cloud and big data technologies such as graphs, columnar stores, HBase (on Hadoop) and Cassandra simply weren't available.
Fast forward to the present, the MDM landscape remains more or less unchanged despite a quantum leap in technology.
Informatica MDM is still an on-premises solution with many of the same challenges we faced while at Siperian
Cloud-based customer master only offerings focused on improving CRM data quality a limiting. Much like Siebel UCM did for Siebel CRM over 10 years ago
Customer and product masters are still supported through separate siloed hubs, even when built using the same tool. In fact, Gartner continues to publish separate customer and product magic quadrants as if to re-enforce this fact
Master data must still be delivered to data warehouses or operational data stores in order for business users to get a promised "complete view"
MDM systems and tools built on 1990s relational database technologies continue to hinder the ability to model real-world many-to-many-to-many relationships that graph technologies are designed for
For the most part, life sciences companies are no closer to getting basic affiliation management functionality, or their dream of an all encompassing key account management application as they are hindered by legacy MDM tools. Even a new wave of cloud-based MDM solutions do not make things any better. The good news is that life sciences companies can avoid a new kind of MDM ("Making Da-same Mistake").
The most popular consumer facing applications today such as LinkedIn and Facebook have shown that business facing data-driven applications can be cloud-based, handle multiple data domains, manage structured, unstructured, master, transactional, activity and social data. Companies should expect complete end-to-end "modern data management" instead of relying on recurring "next generation master data management" promises that remain unfulfilled.