Presenting one centralized view of the customer is key to getting, keeping and growing customer relationships, but it's easier said than done. Customers provide measurable value to an organization, and the best way to track that value is to monitor individual customer behavior. However, many companies have multiple databases packed with customer information, and merging those records is a long, arduous and expensive task. What approaches can a firm take to successfully leverage customer data to create profitable relationships and bottom-line results with its best customers?

A critical first step in any customer data integration project is to identify customers by assigning each person or entity a unique ID. This ID should be the same in all records on that person or entity across all systems and disparate files. To define the ID, first create an inventory of all ID components. The definitions of these key components are basic, but neglecting one of them is all too common:

  • Business@location: a uniquely named organization at one physical address. This defines the type of organization, as well as geographical attributes of a customer.
  • Business Process: the types of business activities an organization performs.
  • Customer: a business entity or person who utilizes a paid service of the firm.
  • Person: a human customer.
  • Role: the position a person holds in a business. A person can have multiple roles.
  • System: the computer facility used by the enterprise or its agents to aid or implement a business process.
  • User: any person who accesses or uses a system.
Making the machine work
Once ID components are defined, it's time to focus on the structure of the ID system. We recommend the "hub and spoke" model, which represents customer identification as a service. This idea goes against conventional wisdom at many firms, though it is the most cost effective and efficient for companies with legacy environments. And, this service model runs only when needed, so it doesn't burden other customer-facing systems with constant demands for data.

The hub and spoke system places a central server in a network, connected to other databases from which to access information. With pre-defined ID components, users can set business rules to perform activities, such as ID generation, look-ups, matches, near matches, ID distribution and audit-trail maintenance. It also maintains a "credibility" rating on the matches it performs, so that future updates and matches can be evaluated as potential replacements for existing pairings. Separate customer profile data can also be placed into the server, though maintaining extensive profile data can elevate costs.

Treating the ID as a service creates maximum flexibility, because new data components -- or "spokes" -- can be added as needed. The ID service concept also minimizes cost and untangles the Gordian knot of having the ID bundled into many applications. Further, implementation costs of a unique ID system are generally much lower than implementations that provide a wide variety of other functions, such as ERP and CRM systems.

Keeping the motor humming
The recommendations above are just a glimpse into what is a detailed and strategically driven process. In a legacy environment, the hub and spoke model may be the approach most likely to help achieve success for reasons of cost, organizational challenges, value derived, ability to manage technological risks and business processes.

Unique ID solutions provide the first step in the customer-focused process, setting the stage for organizations looking to seize competitive advantage by identifying and acting upon customers' needs and value. Those firms able to match strategic direction with deep analytics, and use that know-how to create profitable relationships with key customers, will compete and win in the evolving customer-centric economy.

INSIDE 1to1, October 2002
© Peppers and Rogers Group