Capability differences: CMDM vs. Customer Analytical Repository
The adjacent figure illustrates the typical capabilities of CMDM solutions, compared with the additional capabilities that an analytical repository is intended to address.
Based on the CMDM vs Analytical Repository comparison figure, a few key observations that stand out are:
- For a 360-degree customer analytical repository to be effective, it needs several foundational capabilities that are inherently part of typical CMDM solutions. Hence, there is significant overlap in the capabilities of both technologies.
- With customer engagement increasingly going digital, a CMDM implementation cannot be successful without tracking the multi-dimensional relationships between a customer and the retailer.
CMDM solutions have matured significantly in the last decade, with respect to underlying data models, user interfaces of customer systems, and services to consume or update the master data.
360-degree customer analytical repositories are an emerging trend, with companies offering ready-to-deploy Big Data-enabled solutions that can be used to create a 360-degree view of the customer. The capability differences can be understood from the following two dimensions that relate to the way such customer data systems are built and used in the current context:
- Customer data creation and consumption
In medium-to-large size retail organizations, formal CMDM installations service multiple operational processes such as Loyalty, CRM, online user registrations, mobile registrations, Layaways, and so on, in real-time as well as offline mode. Transaction-style MDM implementations involve the orchestration of business processes between the operational systems and the CMDM solution to achieve a specific outcome such as new user registration.
On the other hand, 360-degree customer analytical repositories are evolving with respect to hardware, software, and the nature of workloads they handle. Hadoop-based systems have traditionally supported high volume analytical processing and are maturing with respect to handling high volume individual read and write transactions such as those originating from a new customer registration or change in the customer profile information. Proprietary Big Data vendor products and in-memory appliances have started addressing these capabilities well.
- Customer Data Management
- Identity resolution: Leading CMDM solutions now bring a high degree of maturity with regard to identity resolution, and they even cater to industry-specific needs. However, most lack the ability to identify individuals outside of enterprise applications, such as social media interactions—an area that 360-degree customer analytical solutions actively explore within the boundaries of customer privacy and consent.
- Governance and data quality: CMDM systems have been built to formally govern data including data quality tools, country-based address and name standardization and stewardship functions. However, this is not necessarily an out-of-the-box function delivered by leading 360-degree customer solutions.
- Analytical data capture: 360-degree customer analytical solution implementations have a sharper focus on the omni-channel customer, and are constantly aggregating cross-channel metrics on purchases, web and in-store activity, social activity, response to communications and individualized preferences. Overall, the focus is more on psychographic than on the demographic and sociographic aspects.
- Data consumption: Formal CMDM solutions provide a fine-grained set of data services that can be knitted together into coarse-grained functions, but these are generally restricted to reading and managing customer records. Analytical repositories, on the other hand, need to deliver a richer set of services that publish analytical metrics about the customer, perform analytical functions or create target lists, and allow end-users to slice and dice data to analyze what-if scenarios.
Realizing Unified View of the Customer: Is MDM a prerequisite?
The question “Can I implement a unified view of the customer (UVC) without an MDM?” can be interpreted in two ways:
- Can I implement a UVC without an MDM discipline?
It is clear by now that the fundamental capabilities required for customer data management are a prerequisite for implementing UVC. No UVC implementation can afford to ignore them, although the level of sophistication may vary.
- Can I implement a UVC without a formal MDM solution?
Basic customer data integration can be a starting point to achieve a universal view of the customer. An implementation without a formal MDM solution can be a starting point to achieve customer data consolidation without the sophisticated capabilities of a formal MDM solution. However, the retailer must have access to a basic set of data management tools for quality management, de-duping or identity resolution, metadata management, data integration, and to persist changes to customer data. Capabilities such as source data history and blacklists or watch lists will have to be custom-built into the data repository.
Over time, retailers can migrate the ‘master’ part of the repository into a formal MDM solution when the need and business case are established.
The growing demand for Omni-channel customer information and real-time customer analytics at ‘points of presence’ is introducing significant change to the architecture of customer data hubs. The need to author customer data anywhere, anytime and deliver real-time personalized communications across channels consistently requires customer data hubs to cater to a wide variety of workloads. This requires a shift from traditional customer data platforms to newer ‘Big Data customer repositories’ co-existing with high-response data platforms.
Creating a robust realization plan for a unified view of the customer involves a careful study of the business purpose, corresponding benefits, and workloads. The architecture data model design and technology considerations are closely linked with each other and are bound to evolve over a period of time. Inadequate planning upfront can result in several challenges, ranging from business disinterest due to the lack of perceived benefits, re-work, or even the inability of the solution to scale up to meet desired workloads.
Retailers who want to quickly leverage customer information for analytical purposes can start with the creation of an analytical repository by using basic customer data consolidation infrastructure such as data matching and quality management. However, the long-term roadmap needs to include the development of a more formal MDM discipline combined with analytical usage. This solution should have the capability to serve multiple user bases and consumption patterns in a scalable manner while preserving data integrity.
TCS Retail Forum Journal
Read other articles
- Unified View of Customer: It’s All About Customer Experience
- Curate Digital Customer Engagement with Extreme Personalization
- How Retailers Can Create the Right Social Circle
- Customer Loyalty Experience: You Can't Afford to Ignore It
- Realizing Unified View for Better Customer Engagement: Understanding What to Build
Download Journal | Journal Home | Contact a Consultant today