Journal of Interactive Marketing

This article examines the role of the code of ethics in the context of other approaches to institutionalized improvement in business ethics, and reports the results of an empirical study among on-the-phone telemarketing personnel regarding their support of various ethical tenets embodied within the industry's professional code. The results indicate that practicing telecommunicators do not indiscriminately support or reject the various tenets of the 12-tenet code. Clarity and honesty in presenting telemarketing offers received the strongest endorsement. Conversely, however, there was significant disagreement with other tenets that called for no telemarketing deceit. The tenet prohibiting delivery of recorded messages without first obtaining permission with a live operator received the least support, implications of the findings suggest that a concerted industry educational effort is needed on telemarketing's code of ethics. There is also the possibility that the code itself may lack the specificity necessary to make it work up to its greatest potential as an institutionalized tool of telemarketing industry self-regulation.

Both the logistic and decaying exponential equations have been used to describe the accumulated growth curve of mail survey returns. While empirical fits have been presented for both curves, a conceptual rationale for either curve is lacking in the mail survey research literature. This paper thus presents a conceptual reason, in addition to empirical fits, for using the decaying exponential equation rather than the logistic equation in describing the accumulated mail survey response curve. It also shows that past comparisons of the logistic and decaying exponential equations have been inappropriate because of the use of different asymptotes. Based on an examination of two decaying exponential equation forms, a strategy for future research is suggested.

We report the results of a well-controlled experiment which tested the readability of direct-mail sales letters. Surprisingly, the mailings of a subscription offer for technical newsletters, sent to both professional and layperson segments, produced no significant differences in response rates as a function of readability levels. This study suggests that readability formulas, as tools for tailoring messages for different segments, may not have broad applicability in the field.

This article describes CART (Classification and Regression Trees)—an interesting and often powerful nonparametric alternative to parametric methods in modeling. The algorithm and accompanying software is described and contrasted to CHAID (Chi-Square Automatic Interaction Detection). Finally, a case study showing CART's application in response modeling is discussed.

Conjoint analysis is described as an important method for designing direct marketing offers prior to a market test. The method is demonstrated with an application in designing direct marketing offers which focus on the price of the product, the channels through which the product may be purchased, and the exchange/refund policy. The results of conjoint analysis provide the utilities of each variable at each level. This information is then used to construct alternative direct marketing offers. The simulated willingness to purchase is also discussed.