At the first ever CDI-MDM summit being held here, conference chairman Aaron Zornes opened his keynote address by telling the crowd of vendor representatives and company technology professionals, "You are the vanguards." Zornes is the founder and chief research officer for the CDI Institute, a San Francisco-based analyst firm, which is predicting major maturation and growth in the CDI-MDM software and services sector -- to the tune of becoming a $1 billion market by 2008.
Over the next few years, more companies will seek out CDI and MDM, and growing demand will propel the young technologies -- and data professionals with relevant skills -- into the mainstream, Zornes said. Market momentum will be driven by the best practices and experiences of enterprise architects, data stewards and other technology experts, whose skills will be in high demand.
"Protect your people," Zornes said during his address. The burgeoning market is facing a skills shortage, he said, and both companies and systems integrators will be looking for experienced professionals for new CDI-MDM initiatives.
Compliance, differentiation and increasing profitability are major drivers of the market's growth, Zornes explained. Companies know that they have up-sell and cross-sell opportunities that require a "panoramic view" of the customer, he said. And while early adopters built their first data hubs in-house, both these companies and newcomers are looking seriously at commercial offerings.
CDI-MDM products have matured to a viable level, Zornes said. The realization of the pricey, long-term maintenance requirements of custom data hubs are also driving the move toward commercial products but, he cautioned, as the market grows, companies may face issues of vendor consolidation and diversification.
The state of the market
Just as SAP struggled to bring its MDM tool to market, IBM and Oracle will also encounter architectural and functional challenges in developing their products, Zornes said. At the same time, best-of-breed vendors will continue to develop for specific vertical industries and applications. Companies will face challenges as they seek to integrate multiple hubs in their data integration projects. Data model flexibility and performance issues are major concerns for many companies in the evaluation stages, and picking the right CDI-MDM vendor will not necessarily be an easy or obvious choice.
Enterprise application integration (EAI), extract, transform and load (ETL) and data integration vendors are aligning with CDI-MDM technology and will eventually become "fully assimilated," Zornes said. Standalone data quality vendors will all but disappear, and those remaining will focus on postal data compliance or transforming to offer their own CDI hubs.
"EII [enterprise information integration] is the exception," Zornes said in a later interview. "That technology is a complement [to CDI-MDM projects] that helps adjudicate the politics and performance issues." EII acts as middleware that can address the problem of managing multiple data hubs.
To manage the political issues of data ownership and data integration change, high-level "steering committees" should be created to develop CDI-MDM projects at companies, Zornes said. These steering committees can drive "good politics" and ensure that a true enterprise-wide strategy is developed. In an ironic twist, companies run the risk of politics and performance concerns leading to standalone CDI and MDM projects in different divisions or departments, which can ultimately create more data silos -- just the problem companies are trying to solve or avoid. To mitigate this risk, companies must determine flexible, extensible, enterprise-wide strategies, addressing critical components like data governance in an overall system design.
Ultimately, companies must not lose sight of the goal of growing and optimizing their business. CDI and MDM systems must change from being data hubs to being process and policy hubs, Zornes said. Business process management (BPM) and flexible workflow management must be part of the overall strategy.
"If all you're doing is building another data hub, you're repaving the cow path," Zornes said.