Customer Data Integration
Achieving a well-rounded view of the customer is the next big challenge in enterprise data management. Companies understand all too well the problems associated with a lack of customer intelligence. After all, organizations without high quality, integrated customer data not only face operational inefficiencies, but also miss out on opportunities to better service customers. But what is not well understood is how to go about selecting a solution that best meets a company's customer-identity needs.
In recent years customer data integration (CDI) has gained market traction and received industry recognition as one of the best methods to address the customer-identity madness. Gartner defines CDI as the combination of the technology, processes, and services needed to create and maintain an accurate, timely, and complete view of the customer across multiple channels, business lines, and, potentially, enterprises where there are multiple sources of customer data in multiple application systems and databases ("Create a Single Customer View with Customer Data Integration," 7 October 2004).
With such a broad definition CDI is easily confused with master data management (MDM) and even data quality. It may be helpful to look at CDI in the context of the various solutions aimed at managing customer data.
Registry Hubs, Data Synchronization, MDM
CDI is best achieved through one of three competing techniques: registry hubs, data synchronization hubs, or as part of a larger MDM solution. Registry hubs support a decentralized approach to consolidating and managing customer data across departments. Rather than storing customer data, the hub maintains a customer-identity cross-index and links to multiple source systems for data integration. Data appears in read-only form and may be referenced in real time, allowing for a dynamic consolidation process during data reads.
In data synchronization hubs, also called coexistence hubs, customer reference data is centrally stored and synchronized after cross-department transactions occur, a feature that supports a master-data approach, but which means data updates do not happen in real time. As a result, synchronization hubs may require cross-system modifications and may raise data ownership concerns.
MDM, on the other hand, requires a centralized data model and cohesive business processes that all departments within an organization must implement and adhere to. MDM hubs, using APIs and SOA interfaces, connect to other systems and send data out from real-time transactions. Therefore, all customer data systems within an enterprise use the same master or centrally controlled data. Adopting MDM is an extensive process that involves re-architecting current systems and data models, meaning organization-wide changes concerning data ownership. This approach is expensive and time consuming, with deployments often taking years.
Choosing the Right Approach
Consider your business objectives when selecting the CDI technique that best meets your customer data needs. Is your organization aiming to leverage customer intelligence to drive revenue, or is it looking to improve operational efficiency by automating business processes and reducing customer maintenance costs?
If customer knowledge is your goal, consider registry or data synchronization hubs. Most customer data is created across multiple departmental sources and not from a central command post, so data is typically fragmented to the point where the organization has no single view of the customer. Registry data synchronization hubs are beneficial in organizations where individuals need to access and build on customer information consolidated across disparate departments and systems.
Yet, important differences exist between these two approaches. Both types of hubs require a central data repository to integrate master data across heterogeneous systems. But data synchronization hubs require pushing information back out to the sources whenever there are modifications across different systems. This may necessitate modifications across affected systems and can raise data ownership issues. Registry hubs, arguably the most cost-effective, unintrusive, and timely approach, do not require operational system changes or process reengineering, giving the technique a lighter footprint with faster deployment.
To improve operational efficiency through automation of customer transactions, MDM is worth considering. The complexities and crucial timeliness of these transactions may be best handled by a centralized management approach to customer information, found with MDM. As "master" of customer data, MDM hubs control one customer master file and use publish/subscribe APIs to push this data out to other systems for real-time transactions. However, beware that MDM deployments can, and often do, involve extensive, time-consuming changes to a company's business processes and systems.
The three approaches to CDI offer different functional and architectural advantages. Before undertaking any CDI investigation, it is best to define your overall business objectives--whether they are driven by revenue or operational efficiencies--and then choose the technology that meets your needs.
About the Author
Paul Friedman is CTO and cofounder of Purisma, Inc. He is a software industry veteran with more than 20 years of experience in engineering and technical management roles. Please visit Purisma at www.purisma.com
Reclaim Your Sanity and Win the Business
No more customer-reference fire drills.