03.01 "Will It Ever Fly? Modeling the Takeoff of Really New Consumer Goods"
by Joseph A. Foster, Peter N. Golder, Gerard J. Tellis

Whirlpool Corporation, a major U.S. manufacturer of consumer durables, test marketed the Personal Valet Clothes Vitalizing System in 2001, and launched it in 2002. The Personal Valet is the first in a new category of appliances: a substitute for dry-cleaning services, providing an in-home service to customers, replacing a service currently performed outside the home. Hence, it generates significant cost savings and added convenience for consumers. Management needed a sound model for the timing of sales takeoff. The project manager applied our “time-to-takeoff model” (which correctly identifies takeoff over 90% of the time). Instead of projecting a linear pattern of sales in the first few years, the project manager used our model to project 10-year sales with takeoff in the 6th year. He also simulated various pricing and product modifications schemes to predict time and probability of takeoff. In the absence of the model, the accounting department, using linear growth, would have forecasted much higher sales than is typical for such a new consumer durable as shown by our model. Such forecasts would have led to actual sales disappointment with actual sales (which were low) and to premature termination of the project.

03.02 "Defending Marketing Share Against a New Entrant"
by John H. Roberts, Charles J. Nelson, Pamela D. Morrison

Telstra, the Australian telephone company, was facing the threat of competitive entry by a major rival, Optus, and sought help in developing a defensive marketing strategy. The problem was to assess the obtainable market share of the new entrant in the residential Australian long distance telephone call market , to gauge how quickly that share would be gained, and determine the factors that would influence its dynamics and ultimate market appeal. We developed probability flow models to provide a framework to generate forecasts and assess the determinants of share loss. Telstra used the models to set prices adaptively, direct service initiatives, design advertising copy, and dimension the network (including financial and manpower planning). Decisions based on the model included a move to compete on service (rather than price) including specific service components, a decision not to oppose an early move to preselection by Optus, the setting of actual price levels and formats, a targeting plan for telemarketing, and dimensioning decisions based on model forecasts. Telstra avoided head-to-head price plan comparisons, representing an increase in contribution of $US 22 mm per year and other benefits of the model application lead to incremental revenue of over $US 50 mm a year.

WINNER: 03.03 "Optimizing Rehanania's Direct Marketing Business Through Dynamic Multi-Level Modeling (DMLM) in a Multi-Catalog-Brand Environment" by Ralf Elsner, Arnd Huchzermeier, Manfred Krafft

Rhenania, a German direct mail-order company, turned its catalog mailing practices around within one year and consequently moved up in market position from number five to number two. A new dynamic multi-level modeling (DMLM) approach was developed to answer the most important direct market questions: (1) when, (2) how often, and (3) to whom in our customer base should we mail to? The uniqueness of this modeling approach is that (1) customers are being dynamically evaluated based on their past purchase history, and (2) it allows for a derivation of a threshold level of sales per customer for profit maximization of mailing campaigns over time and across multiple customer segments. The DMLM approach was so effective that Rhenania outperformed the market by far and in a relatively short period of time could acquire two of its major competitors (one of them bigger than Rhenania, i.e., a subdivision of the Springer Publishing Company).