Faculty and Research

We review the discussion at a workshop whose goal was to achieve a better integration among behavioral, economic, and statistical approaches to choice modeling. The workshop explored how current approaches to the specification, estimation, and application of choice models might be improved to better capture the diversity of processes that are postulated to explain how consumers make choices. Some specific challenges include how to capture and parsimoniously describe heterogeneous mixes of heuristic choice rules, methods for building realistic models of choice, and non-traditional methods for estimating models. An agenda for important future work in these areas is also proposed.

Marketing investments are designed to change consumer behavior in ways that help goods compete in the marketplace. Previous research has
focused on using elasticity decompositions to measure how these investments affect either consumer decision making or competing goods. On the one hand, decision-based decompositions attribute the growth in own-good demand to changes in the consumers’ decision-making processes. On the other hand, unit-based and share-based decompositions attribute the same growth to either rivalrous or nonrivalrous sources.

The objective of this article is to provide a clear and accurate method that simultaneously attributes the growth in own-good demand to
changes in (1) consumers’ decisions, (2) competitive demand, and (3) competitive market share. I accomplishes this by settling some
confusion about what the decision- and share-based decompositions mean, by discussing how each of the decompositions are related to the
others, and by discussing the research questions that each of the decompositions can answer. From a managerial perspective, I discusses how this method can be used to simplify and clarify the data needed for decision making.

We introduce the idea of a massively categorical variable, a variable such as zip code that takes on too many values to be treated in the standard manner, and show how to use it directly as explanatory variables in an econometric model.

In an application of this concept, we explore several issues confronted in direct marketing. To begin with, the data offered by many providers, such as Experian and Claritas, are masked through aggregation to protect consumer privacy. Although this practice creates some difficulty when trying to construct models of individual-level choice behavior, we show how to take full advantage of such data through a hierarchical Bayesian variance components (HBVC) model. The flexibility of our approach allows us to combine several sources of information, some of which may not be aggregated, in a coherent manner, and we show that the conventional modeling practice understates the uncertainty with regard to its parameter values. 

To give economic meaning to our results, we develop targeting strategies under an array of financial conditions and show how to determine an organization’s willingness-to-pay for additional data.

Companies spend billions of dollars on marketing every year because it is essential to organic growth. Given these large investments, marketing managers have the responsibility to optimally allocate resources and to demonstrate that their investments generate suitable returns for the firm.

In this chapter we highlight a two-stage process for making and justifying marketing allocation decisions. In stage one, a model of demand is estimated. This model empirically assesses the impact of marketing actions on consumer demand for a company’s product. In stage two, estimates from the demand model are used as input in an optimization model that attempts to maximize profits. This stage takes into account costs as well as firm’s objectives and constraints (e.g., minimum market share requirement).

Over the last several decades, various methods that follow these two stages, either implicitly or explicitly, have been developed. We categorize these techniques in a three-by-three matrix, which suggests three different methods for stage-one demand estimation (decision calculus, experiments, and econometric methods) and three different methods for stage-two economic impact analysis (descriptive, what-if, and formal optimization approach). We discuss pros and cons of each method and provide illustrations through applications and case studies.

Previous research suggests that the random coefficients logit is a highly flexible model that overcomes the problems of the homogeneous logit by allowing for differences in tastes across individuals. The purpose of this paper is to show that this is not true. We prove that the random coefficients logit imposes restrictions on individual choice behavior that limit the types of substitution patterns that can be found through empirical analysis, and we raise fundamental questions about when the model can be used to recover individuals' preferences from their observed choices.

Part of the misunderstanding about the random coefficients logit can be attributed to the lack of cross-level inference in previous research. To overcome this deficiency, we design several Monte Carlo experiments to show what the model predicts at both the individual and the population levels. These experiments show that the random coefficients logit leads a researcher to very different conclusions about individuals' tastes depending on how alternatives are presented in the choice set. In turn, these biased parameter estimates affect counterfactual predictions. In one experiment, the market share predictions for a given alternative in a given choice set range between 17% and 83% depending on how the alternatives are displayed both in the data used for estimation and in the counterfactual scenario under consideration. This occurs even though the market shares observed in the data are always about 50% regardless of the display.

Prior research hypothesizes managers use ‘real actions,' including the reduction of discretionary expenditures, to manage earnings to meet or beat key benchmarks.  This paper examines this hypothesis by testing how different types of marketing expenditures are used to boost earnings for a durable commodity consumer product which can be easily stockpiled by end-consumers as well as who, within the firm, is responsible for these actions.

Combining supermarket scanner data with firm-level financial data, we find evidence that differs from prior literature.  Instead of reducing expenditures to boost earnings, soup manufacturers roughly double the frequency and adversely change the mix of marketing promotions (price discounts, feature advertisements and aisle displays) at the fiscal quarter-end when they have greater incentive to boost earnings. 

Our results confirm managers' stated willingness to sacrifice long-term value in order to smooth earnings (Graham, Harvey and Rajgopal, 2005) and their stated preference to use real actions to boost earnings to meet different types of earnings benchmarks.  We estimate that marketing actions can be used to boost quarterly net income by up to 5% depending on the depth and duration of promotion.  However, there is a price to pay, with the cost in the following period being approximately 7.5% of quarterly net income.

Finally, a unique aspect of the research setting allows tests of who is responsible for the earnings management. While firms appear unable to increase the frequency of aisle display promotions in the short run, they can reallocate these promotions within their portfolio of brands.  Results show firms shifting display promotions away from smaller revenue brands toward larger ones following periods of poor financial performance.  This indicates the behavior is determined by parties above brand managers in the firm.

These findings are consistent with firms engaging in real earnings management and suggest the effects on subsequent reporting periods and competitor behavior are greater than previously documented.

This paper empirically explores the contingencies that drive cannibalizing and complementary effects across channels to provide sales forecasting, promotion planning, and customer relationship management guidance to multichannel managers. We investigate three contingencies in a sales analysis of a leading U.S. retailer who adds a new retail store channel to existing catalog and online channels. We show that the emergence and strength of cannibalizing and complementary effects varies over time, across type of channel, and by type of customer, and provide insight into when and where managers can expect these effects to dominate and how to counter cannibalization and promote complementarity across channels.

We find that opening retail stores cannibalizes sales in the catalog and online channels in the short term, but produces complementary effects in both channels in the long term; cannibalization is magnified in the catalog channel, while complementarity is magnified in the online channel. Customer analysis suggests that opening retail stores paves the way for higher rates of customer acquisition and higher rates of repeat purchasing among existing customers in the direct channels in the long term.