Subrata K. Sen

Market Segmentation Using Models of Multidimensional Purchasing Behavior

MARKET segmentation has become a widely discussed concept among both managers and researchers. Conceptually, a strategy of market segmentation involves the development and implementation of different marketing programs for different components of the overall market. Products and advertising copy can be specially designed to appeal to the particular preferences of each segment, while promotion (e.g., couponing and sampling) and media selection can be geared to each segments demographic characteristics and media habits. However, to translate the concept of market segmentation into a profitable managerial strategy, it is necessary to define segments that exhibit differing responses to changes in the firms marketing mix variables. It is also desirable to be able to identify each segment in terms of customer characteristics. A review of existing market segmentation approaches indicates that no single approach fulfills the above segmentation criteria adequately. Therefore, this article presents a different approach, one that could lead to a more profitable segmentation strategy. After a review of the literature, the new segmentation approach is described in detail. The feasibility of the approach is illustrated by an analysis of a set of panel data on aluminum foil purchases. The final section of the article presents a discussion of how the proposed segmentation approach can help companies develop effective marketing strategies. Literature Review

Measurement and Estimation of Conjoint Utility Functions

Conjoint analysis studies usually calculate utilities for several discrete attribute levels and then use linear interpolation to predict utilities for other attribute levels. Instead of linear interpolation, utility functions can be used. This paper compares the predictive abilities of the utility function approach and the procedure of linear interpolation.

Electoral Participation in the French Fifth Republic

Variations in second ballot abstention and blank and invalid ballot rates (over the cross-section of French election districts) are examined for all four legislative elections of the French Fifth Republic. Analysis was conducted primarily through a heuristic decision-making model and a spatial model developed from the theories of Riker, McKelvey, and Ordeshook, and Davis, Hinich, and Ordeshook. Abstentions appear to be primarily influenced by long-term factors and the competitiveness of the contest. Blank ballots appear to be primarily dependent upon short-term factors, especially nonvoting from the alienation that results when a candidate present on the first ballot is not present on the second. The alienation model and the heuristic model, though partly collinear, make independent contributions to the explanation of the blank ballot variance.

Purchasing Strategies across Product Categories

This article describes the degree to which consumers use identical or similar brand and store choice strategies across product categories. The analysis is based on data on two pairs of frequently-purchased products and the results indicate that consumers frequently use identical or similar purchasing strategies across product categories. The principal implication of these results is that buying behavior may be governed by general household characteristics such as demographics to a greater extent than past research in this area had indicated.

Should NFL blackouts be banned

In this paper, the authors examine the economic and policy implications of the National Football League (NFL) ‘blackout rule,’ a league rule that prohibits local television broadcast of games that are not sold out at least 72 hours prior to game time. The foundation for understanding and assessing the impact of the blackout rule is an analysis of attendance using data on games during the 1996–1997 National Football League season. Expanding on previous research, three separate components of attendance (season ticket sales, game day ticket sales, and game day noshows) are examined in detail. Accounting for the endogeneity of key variables, Tobit and Probit analyses are used to estimate and predict individual game attendance. These empirical estimates are then used as a vehicle to assess the implications of game day blackouts and the potential for public policy intervention. More specifically, the authors begin by estimating the impact of the blackout on game day attendance. Using these estimates, they assess the implications of imposing a local blackout for individual team revenues. The gain in on-site stadium revenue due to the blackout (e.g., through additional ticket and concession sales) are then viewed in the broader context of the societal loss due to the game not being broadcast in the local area. The empirical results suggest that the gain in team revenue is small in comparison to the loss of viewership rights. This suggests that public policy intervention may be possible that would result in a Pareto superior market outcome. The paper concludes by exploring possible intervention strategies.

Measuring Quality Perceptions

The measurement and management of product quality is an important factor in maintaining a firms competitive edge. In this paper we (i) describe a procedure to obtain quality perceptions from users of office copiers, (ii) model the relationship between perceived quality and the underlying engineering attributes, and (iii) compute price-quality elasticities and suggest segmentation procedures.

Quantifying the competitive impact of a new entrant

Abstract Quantifying the impact on sales and profits of existing firms due to a competitive entry is an important practical problem from both a managerial and a policy perspective. In this article we apply the well-known gravitational model to measure the sale impact of the planned entry of a new shopping center. The use of the model in this context is illustrated through two actual applications. Data were collected from a telephone survey of a random sample of local households. Respondents were asked a variety of questions about their shopping behavior, including their eatings of the arreactiveness of the new and existing malls. In addition, we collected the usual data on mall size, travel time, population size, etc. The gravity model was formulated as a logit model and estimated by the well-known maximum likelihood procedure. It was found that mall choice could be explained by a simple model that used the following three variables: mall size, travel time, and the attractiveness of the mall. The estimated model was then used to quantify the competitive impaxct of a new shopping center under alternative scenatios. The article concludes with some recommendations on the practical use of such models.

Algorithms: Algorithm 333: MINIT algorithm for linear programming

p r o c e d u r e Chebyshev (A, d, h, m, n, refset, epz, insufficientrank, zerolambda ) ; va lue m, n; i n t e g e r m, n; r ea l a r r a y A, d, h; i n t e g e r a r r a y refset; rea l epz; l abe l insu~cienlrank, zerolambda ; c o m m e n t Chebyshev computes a solution in the Chebyshev sense to an overdetermined system of linear equations, Ax = d. Details and notat ion are given in a paper by Bartels and Golub [Comm. A C M 11 (June 1968), 403-408]. The parameters to procedure Chebyshev are:

Candidate selection and voting behavior in France

SummaryAs a measure of the component of electoral change resulting from candidate selection variables, the foregoing regression analysis suggested that about 3% of all voters in the districts considered change their votes as a consequence of candidate selection policies. In competitive elections, these 3% are a highly significant proportion. Since this change largely favors the Gaullists, one can suggest that the French Left (Communist and Non-Communist) has not manipulated candidate selections as successfully as the Gaullists.Although the patterns of variation are different for Gaullist candidates than for Federation or Communist candidates, we have shown that the performances of new candidates differ systematically from those of old, repeat candidates. Thus, we hope this research suggests the interest of looking at candidate selection strategies and the communications psychology of multiple exposures as relevant variables in the study of voting behavior.The early study of voting behavior was largely influenced by looking at thesocial determinants of voting behavior through such variables as social class, religious practice, race, etc. The use of psychological theory in the study of voting behavior has, in fact, largely been confined to the analysis of the cross-pressures evidenced through these social variables. In other works, we have indicated how elite behavior influences voting behavior through coalition decisions. To this we can now partially suggest how elite behavior influences voting behavior through candidate selection decision. With as much empirical knowledge aboutelite influences as we now have aboutsocial influences, the variations in behavior that we must ascribe to the peculiarities of districts and their candidates may well be very minor.