Technology can solve a lot of integration problems, but there's still no silver bullet for enterprise-wide data
governance. That's why companies are overhauling data management on an "entity-by-entity" basis, according to a recent study on market trends from Forrester Research Inc. of Cambridge, Mass.
Rather than dive into huge master data management (MDM) initiatives, companies are first addressing customer data integration (CDI) and product information management separately, the study found. But, even for companies simply taking on CDI, there's a lot to consider -- varied technology approaches, high implementation costs, and a market that's changing quickly.
"The toughest part about master data management is the governance -- getting everyone to agree who owns the data, how that data is manipulated, who's in charge of cleaning up the information," said one of the study's authors, R. "Ray" Wang, senior analyst with Forrester.
"You lose a lot of the richness of a [operational] data hub," Wang said.
So, vendors are offering more flexibility and "richness" by developing harmonized or hybrid CDI systems, incorporating features of both registry and operational data hub styles. Registry vendors will start to look more like operational data hubs, Wang predicted, and operational data hub vendors will work to speed up ROI for their products. Ultimately, the real winners will be vendors that can navigate the underlying data model challenge, inherent to CDI success.
"Most of the customer data hubs have very rigid data models that you can't modify," Wang explained. "If you want to add another field or if you want to do something else, it becomes a customization."
The data model issue will remain a hot CDI topic for vendors and companies, Wang said. It will play out with early adopters choosing a more flexible approach, followers likely choosing market-tested models, and vendors solidifying their data model messages.
While companies grapple with CDI, many are also implementing business intelligence (BI) tools, and these technologies will meet in the middleware. The study predicts that BI and middleware vendors will collaborate with CDI vendors to provide tighter integration. The market has already experienced some vendor gyrations, with Hyperion's acquisition of Razza and investment in CDI vendor Purisma; IBM's purchases of SRD, Trigo, DWL, and Alphablox; and Oracle's acquisition of Siebel, among others. Market consolidation and vendor partnerships will continue, Wang said.
Additionally, given the importance of trustworthy data to both BI and CDI projects, the study predicts that "trusted data sources" such as Dun & Bradstreet, Acxiom, and Experian will play a bigger role in the market.
"You can spend $10 million doing a master data initiative, and then the data gets corrupted in three to six months and you've wasted all these efforts," Wang explained. "You're opening up the system to Web self-service, salespeople entering data, field service entering data, call centers entering data. The data gets corrupted pretty quickly."
Rather than argue over which department's data is right, companies will look to trusted data sources to append or be the source of record for customer information. These data vendors are offering more services to the CDI market today and, in the future, they may be more aggressive in the space or even acquire a CDI provider, Wang predicted.
Last, but definitely not least, a quickly growing part of the CDI market is professional services, Wang said. Data integration, data governance, and training are all major costs, whether companies use in-house or outsourced resources. The study found that, on average, service costs represent almost three times the technology licensing fees. That's one reason companies are starting with less resource-intensive, registry-style CDI approaches, which have comparatively lower implementation costs, Wang said.
With all of these market changes and challenges, companies should proceed with CDI strategically, Wang said. First, choose service-oriented architecture (SOA)-enabled technology, which offers the long-term flexibility to tie data models back to business processes. Next, take time to determine the correct data-model strategy, Wang recommended. Companies should ask potential vendors about the extensibility and support of their data-model approach. And finally, buyers should ensure that a vendor has enough resources trained and available to support them on their CDI journey.
"That's really the tough part," Wang said. "There's still not enough trained resources out there and that's hampering growth."