Here are a few of the 76 pages available to our subscribers

Home Site

    Page 34
16.6 Table of Contents Bottom of Page Site Map


Temporal Abstractions Model Customer Behavior in
Business Ecosystems:

Insightful Data Mining


By Bob Nisbet

       This article proposes a link between the philosophical studies of the nature of being and CRM studies of customer response. This link is obvious when viewing customer response through the metaphor of business as an organism, rather than a machine. This approach guides the design of powerful customer behavior models based on the abstractions obtained from historical time-series data. These time based or temporal abstractions, keyed to the response date for each customer (rather than the calendar date), reflect what customers did during the several months before their response was captured for this model. This voluntary attrition model, based on households with automobile insurance policies, indicates approximately 60 percent of the models with values over the random average is attributable to non-historical (static) variables while the temporal abstraction variables supply the other 40%.

Knowing How Our Customers Behaved Before They Acted
       To be competitive in today’s markets, a company must capture and exploit information from historical detail records describing what customer did in the past. This information helps define patterns in customer behavior leading up to their decision not to patronize the company. For a given customer, the decision to leave the company did not happen in a vacuum. Many factors contributed to this decision, such as dissatisfaction with service, enhanced perception of competitive goods and services, and changes in business or personal needs. While customer care programs track some factors, such as customer satisfaction, it is not possible to capture and store in the corporate databases most factors that directly contribute to attrition. The only method for reflecting these attrition variables is relating them to customer behavior patterns that are traceable from data in the data warehouse. This article examines how

historical information on customers who have left the company can predict patterns indicating which current customers have a high probability of leaving in the near future.

Transforming Corporations into Business Ecosystems: The Path to Customer Fulfillment
       Since the Industrial Revolution, Western Society tends to view the world as a machine, composed of components that functioned like cogs, wheels and springs. Although Newton formalized this approach in Science, it only worked within the range of Newton’s instruments. Later discoveries by Einstein (Relativity) and Quantum Physicists caused the Newtonian concept of the world change drastically.
       The business community also embraced this metaphor during the Industrial Revolution and viewed Henry Ford’s automobile assembly line as the paragon of efficiency. As long as the product was relatively simple in organization, this metaphor works. An efficient business became defined in terms of:

       a “well-oiled machine”
       “having momentum”
       “gaining steam”
       “firing on all eight cylinders”

       The primary business unit became the corporation with the prevailing attitudes; “It is Us Against Them” and “Only the Strong Competitors Survived.” For these corporations, with production being the primary business activity, they believed they had to maximize production to maximize revenue. Generations of Operations Research practitioners sought to optimize processes to maximize business revenue.
       Fast computers, flexible communications and (recently) the Internet, fostered a new business paradigm — The Business Ecosystem

To Page 28

16.6 Table of Contents
Top of Page

To Page 42


PC AI Magazine - PO Box 30130 Phoenix, AZ 85046 - Voice: 602.971.1869 Fax: 602.971.2321
e-mail: - Comments?