Andre Bonfrer

Investigating the Effects of Store-Brand Introduction on Retailer Demand and Pricing Behavior

Researchers have recently been interested in studying the drivers of store-brand success as well as factors that motivate retailers to introduce store brands. In this paper, we study the effects of the introduction of a store-brand into a particular product category. Specifically, we are interested in the effect of store-brand introduction on the demand as well as on the supply side. On the demand side, we investigate the changes in preferences for the national brands and price elasticities in the category. On the supply side, we study the effects of the new entrant on the interactions between the national brand manufacturers and the retailer introducing the store brand, including how these interactions influence the retailers pricing behavior. In doing so, we are also able to test whether the observed data are consistent with some of the commonly used assumptions regarding retailer pricing behavior. For the demand specification we use a random coefficients logit model that allows for consumer heterogeneity. The model parameters are estimated using aggregate data while explicitly accounting for endogeneity in retail prices.Our empirical results obtained from the oats product category based on store-level data from a multistore retail chain indicate that the store-brand introduction generates notable changes within the category. The store-brand introduction coincides with an increase in the retailers margins for the national brand. We find that the preferences for the national brand are relatively unaffected by the introduction of the store-brand. While consumers are, in general, more price sensitive (in terms of elasticities) than they were prior to store-brand introduction, a statistical test of the differences inmean price elasticities across stores and between the two regimes fails to reject the hypothesis of no change in these elasticities. Elasticities in specific stores however, do increase after the store brand is introduced. We also find that there is considerable heterogeneity in the preferences for the store-brand. On the supply side, we test several forms of manufacturer-retailer interactions to identify retailer pricing behavior most consistent with the data. Our results indicate that the data reject several, commonly imposed, forms of interactions. In examining the nature of manufacturer interactions with the retailer, we find that the manufacturer of the national brand appears to take a softer stance in its interactions with the retailer subsequent to store-brand entry. This finding is consistent with academic research and with articles in the popular press which suggest that the store brand enhances the retailers bargaining ability vis-A -vis the manufacturers of the national brands. We also provide results from a second product category frozen pasta) that are largely consistent with those found in the oats category.

The effect of competitive advertising interference on sales for packaged goods

Competitive advertising interference can occur when viewers of advertising for a focal brand are also exposed to advertising messages for competing brands within a short period (e.g., one week for television advertising). Although competitive advertising interference has been shown to reduce advertising recall and recognition and brand evaluation measures, no studies have examined its impact on brand sales. In this research, the authors use a market response model of sales for two grocery categories for a large grocery chain in the Chicago area to study the extent to which competitive advertising interference influences sales. The model enables the authors to capture the “pure” own-brand advertising elasticities that would arise if there were no competitive interference. The results show that competitive interference effects on sales are strong. When one or more competing brands advertise in the same week as the focal brand, the advertising elasticity diminishes for the focal brand. The decrease depends on the number of competing brands advertising in a particular week and their total advertising volume. The authors find that having one more competitor advertise is often more harmful to a focal brands advertising effectiveness than if the current number of advertising brands increase their total advertising volume.

The Impact of Brand Quality on Shareholder Wealth

This study examines the impact of brand quality on three components of shareholder wealth: stock returns, systematic risk, and idiosyncratic risk. The study finds that brand quality enhances shareholder wealth insofar as unanticipated changes in brand quality are positively associated with stock returns and negatively related to changes in idiosyncratic risk. However, unanticipated changes in brand quality can also erode shareholder wealth because they have a positive association with changes in systematic risk. The study introduces a contingency theory view to the marketing–finance interface by analyzing the moderating role of two factors that are widely followed by investors. The results show an unanticipated increase (decrease) in current-period earnings enhances (depletes) the positive impact of unanticipated changes in brand quality on stock returns and mitigates (enhances) their deleterious effects on changes in systematic risk. Similarly, brand quality is more valuable for firms facing increasing competition (i.e., unanticipated decreases in industry concentration). The results are robust to endogeneity concerns and across alternative models. The authors conclude by discussing the nuanced implications of their findings for shareholder wealth, reporting brand quality to investors, and its use in employee evaluation.

Scalable Inference of Customer Similarities from Interactions Data Using Dirichlet Processes

Under the sociological theory of homophily, people who are similar to one another are more likely to interact with one another. Marketers often have access to data on interactions among customers from which, with homophily as a guiding principle, inferences could be made about the underlying similarities. However, larger networks face a quadratic explosion in the number of potential interactions that need to be modeled. This scalability problem renders probability models of social interactions computationally infeasible for all but the smallest networks. In this paper, we develop a probabilistic framework for modeling customer interactions that is both grounded in the theory of homophily and is flexible enough to account for random variation in who interacts with whom. In particular, we present a novel Bayesian nonparametric approach, using Dirichlet processes, to moderate the scalability problems that marketing researchers encounter when working with networked data. We find that this framework is a powerful way to draw insights into latent similarities of customers, and we discuss how marketers can apply these insights to segmentation and targeting activities.

Real-Time Evaluation of E-mail Campaign Performance

We develop a testing methodology that can be used to predict the performance of e-mail marketing campaigns in real time. We propose a split-hazard model that makes use of a time transformation (a concept we call virtual time) to allow for the estimation of straightforward parametric hazard functions and generate early predictions of an individual campaigns performance (as measured by open and click propensities). We apply this pretesting methodology to 25 e-mail campaigns and find that the method is able to produce in an hour and fifteen minutes estimates that are more accurate and more reliable than those that the traditional method (doubling time) produces after 14 hours. Other benefits of our method are that we make testing independent of the time of day and we produce meaningful confidence intervals. Thus, our methodology can be used not only for testing purposes, but also for live monitoring. The testing procedure is coupled with a formal decision theoretic framework to generate a sequential testing procedure useful for the real time evaluation of campaigns.

Recovering stockkeeping-unit-level preferences and response sensitivities from market share models estimated on item aggregates

The authors propose a new approach to obtaining stockkeeping-unit (SKU)-level preferences and response sensitivities. The authors distinguish an attribute-level model, in which the unit of analysis is the market share for an alternative created by aggregation (e.g., Colgate toothpaste), from a truly disaggregate SKU-level model, and they establish an analytical relationship between parameters that they obtain from the two models. The authors show that SKU-level parameters can be recovered by calculation from estimated attribute-level parameters, circumventing the need for direct estimation of the more complex SKU-level model. They calibrate the store data market share model using 98 weeks of data for ten brands and 168 SKUs of toothpaste. Rather than estimate 168 preference parameters (when there is an “outside” alternative in addition to the 168 “inside” ones), it is only necessary to estimate ten brand-preference parameters from which the 168 parameters can be computed, as long as share and marketing-mix data are available at the SKU level. Covariate effects, such as marketing-mix response parameters, can be recovered in a similar fashion. Holdout tests demonstrate superior predictive performance, and the authors discuss implications for the derivation of elasticities for new SKU introductions.

Recovering SKU-level Preferences and Response Sensitivities from Market Share Models Estimated on Item Aggregates

We propose an alternative approach to obtaining SKU-level preferences and response sensitivities. An attribute-level model in which the unit of analysis is the market share for an alternative created by aggregation e.g., Colgate toothpaste) is distinguished from a truly disaggregate SKU-level model and develop an analytical relationship between parameters obtained from these two models is established. We show that the researcher can recover SKU-level parameters via calculation from estimated attribute-level parameters, circumventing the need for direct estimation of the more complex true SKU-level model. Our market share model is calibrated using 98 weeks of data for 10 brands and 168 SKUs in the toothpaste category. We show that instead of estimating 168 preference parameters (when we have an outside alternative in addition to the 168 inside ones), one need only estimate 10 brand preference parameters from which the 168 parameters can be computed as long as share and marketing mix data are available at the SKU level. Marketing mix response parameters can be recovered in a similar fashion. Estimation on holdout samples demonstrates superior predictive performance compared with other available methods. Implications for the derivation of SKU-level elasticities and forecasts of market share and response sensitivity for new items introduced to the category are discussed.

A series of two-urn biased sampling problems

ABSTRACT Much of the literature on matching problems in statistics has focused on single items chosen from independent, but fully overlapping sets. This paper considers a more general problem of sampling without replacement from partially overlapping sets and presents new theory on probability problems involving two urns and the selection of balls from these urns according to a biased without-replacement sampling scheme. The strength of the sampling bias is first considered as known, and then as unknown, with a discussion of how that strength may be assessed using observed data. Each scenario is illustrated with a numerical example, and the theory is applied to a gender bias problem in committee selection, and to a problem where competing retailers select products to place on promotion.