Master data management (MDM) has had a rocky history in the insurance industry. Back in the 2000s, MDM was a hot topic. Insurers eagerly invested in MDM tools and projects, believing MDM would solve their data quality issues and provide a “single source of truth” for critical data domains like producers, customers, products and risk locations.
The results were often underwhelming. MDM projects tended to be large, expensive IT initiatives that dragged on for years. The promised data quality improvements and business benefits rarely materialized to the degree expected. Frustrated with poor ROI, many insurers scaled back or abandoned their MDM programs. MDM became a dirty word in insurance circles.
Cut to today, and interest in MDM is reemerging. So what’s changed? Here are a few factors driving the cautious MDM revival:
- New technologies like cloud computing and data virtualization are making MDM easier and cheaper to implement. The heavy lifting required is decreasing.
- There’s greater understanding that MDM requires solid data governance – the people, process, and policy aspects – not just technology. Data governance lays the foundation for MDM tools to work effectively.
- Use cases related to AI-powered chatbots, analytics, customer engagement, and digital transformation have made clean, consistent data more important than ever. MDM is recognized as an enabler for these initiatives.
- Data quality expectations are higher all around. Both customers and regulators are demanding more data transparency and accuracy from insurers.
While past failures have made the industry appropriately skeptical, it seems MDM may finally be ready for prime time. The key lessons are to start small, focus on tangible business value, and embed MDM within an overarching data governance strategy. Avoid the “big bang” approaches of the past. Incremental success will help MDM shed its reputation as an expensive white elephant.
Approached correctly, MDM may finally deliver on its promise to provide trusted and authoritative customer, policy, and claims data for insurers. But it will require restraint and realistic expectations to make MDM work this time around.
