Building CRM Databases
One of the hottest areas of growth I've noticed lately is in the area of customer relationship management (CRM) applications. Increasingly, organizations both large and small are building CRM applications. In some cases, these databases are being built using sales force automation (SFA) packages, while in others they are based on data warehouse technology. Sophisticated organizations are blending both approaches: using SFA packages to manage the sales cycle, and then moving the data into decision-support databases for pre- and post-sales marketing and long-term customer management. They are also consolidating this sales-related data with other data, such as customer service, billing and receivables, and external demographic or market research data, to build a richer customer profile.
If you build a CRM database, it can yield major benefits for your company. There are a number of important issues that you will need to address when building a CRM database.
The first is data quality. Validation and verification is essential for a CRM database; otherwise, you take the risk of creating multiple records for the same individual or identifying two people as the same individual. In either case, you can lose money and alienate your customer by, for example, sending multiple copies of a mailing to a single individual. This why you need to pay close attention to the problems associated with customer identifiers that are created by uncoordinated applications. For example, a health-care institution has to be sure that each patient is identified by a unique medical record number that is used by all applications throughout the institution, such as admitting, billing, laboratory, radiology, pharmacy, and so on.
Names and addresses require particular attention. Names, which are unique to an individual, should be parsed and normalized before they are entered into a CRM database. Parsing is the process of decomposing the name into its elements, such as first name, last name, surname, prefix (Dr., Rev., Mr.), suffix (M.D., Ph.D., Jr., III, Esq.) and relationships (Trustee for …, and so on). Normalizing refers to reassembling the name into a consistent format. Addresses are somewhat easier to normalize and validate than names, since they have a more consistent structure and can be verified using specialized databases. It's relatively easy, for example, to validate a ZIP+4 field against a table, available from the postal service, of all ZIP codes in the U.S.
Another issue to consider is the rich, relatively unstructured set of data you'll want to collect on customers. Items such as support calls, demographic data and customer inquiries are all grist for the CRM database, but these tend to be unstructured, text-based data elements that can be difficult to query and analyze in a decision-support environment. You'll need to carefully consider the kind of data your users will require, and how to structure the data in a way that will be meaningful and useful to them.
Once you've got the data in a database, the next thing you need to consider is how to maximize its value. First, and foremost, is to put the data into the hands of the people who can best use it. Every customer touch point, such as customer service or telemarketing reps, should have easy, quick access to this data.
This kind of data is essential for determining customer lifetime value. This process helps to identify the good customers who should be wooed and cultivated from those who consume products and services but are unprofitable and should be gently steered in the direction of your competition. Profitable customers can be provided with additional benefits, services and special premiums or offers that will entice them to remain good customers. Nonprofitable customers can be managed differently, through additional fees or reduced services. These techniques can have the effect of either making these customers profitable or encouraging them to take their business elsewhere.
Finally, this kind of data is marketing campaign gold. One of the biggest costs of marketing is the cost of direct mail, or other targeted campaigns, that demand large-scale distribution of marketing offers and other collateral material. By cultivating and massaging the CRM database, you can identify customers and prospects who would be amenable to a specific offer, and can send out fewer offers that result in a higher response rate. Robert Craig is director, Data Warehousing and Business Intelligence Division, at Hurwitz Group Inc. (Framingham, Mass.). Contact him at firstname.lastname@example.org or via the Web at www.hurwitz.com.