Knowledge that Transforms

Our objective in this paper is to measure the impact (valence, volume, and variance) of national online user reviews on designated market area (DMA)-level local geographic box office performance of movies. We account for three complications with analyses that use national-level aggregate box office data: (i) aggregation across heterogeneous markets (spatial aggregation), (ii) serial correlation as a result of sequential release of movies (endogenous rollout), and (iii) serial correlation as a result of other unobserved components that could affect inferences regarding the impact of user reviews. We use daily box office ticket sales data for 148 movies released in the United States during a 16-month period (out of the 874 movies released) along with user review data from the Yahoo! Movies website. The analysis also controls for other possible box office drivers. Our identification strategy rests on our ability to identify plausible instruments for user ratings by exploiting the sequential release of movies across markets—because user reviews can only come from markets where the movie has previously been released, exogenous variables from previous markets would be appropriate instruments in subsequent markets. In contrast with previous studies that have found that the main driver of box office performance is the volume of reviews, we find that it is the valence that seems to matter and not the volume. Furthermore, ignoring the endogenous rollout decision does not seem to have a big impact on the results from our DMA-level analysis. When we carry out our analysis with aggregated national data, we obtain the same results as those from previous studies, i.e., that volume matters but not the valence. Using various market-level controls in the national data model, we attempt to identify the source of this difference. By conducting our empirical analysis at the DMA level and accounting for prerelease advertising, we can classify DMAs based on their responsiveness to firm-initiated marketing effort (advertising) and consumer-generated marketing (online word of mouth). A unique feature of our study is that it allows marketing managers to assess a DMA's responsiveness along these two dimensions. The substantive insights can help studios and distributors evaluate their future product rollout strategies. Although our empirical analysis is conducted using motion picture industry data, our approach to addressing the endogeneity of reviews is generalizable to other industry settings where products are sequentially rolled out.

We study how opinion leadership and social contagion within social networks affect the adoption of a new product. In contrast to earlier studies, we find evidence of contagion operating over network ties, even after controlling for marketing effort and arbitrary systemwide changes. More importantly, we also find that the amount of contagion is moderated by both the recipients' perception of their opinion leadership and the sources' volume of product usage. The other key finding is that sociometric and self-reported measures of leadership are weakly correlated and associated with different kinds of adoption-related behaviors, which suggests that they probably capture different constructs. We discuss the implications of these novel findings for diffusion theory and research and for marketing practice.

This paper shows that when the alternatives offered to consumers span the preference space (as it would be chosen by a firm), search or evaluation costs may lead consumers not to search and not to choose if too many or too few alternatives are offered. If too many alternatives are offered, then the consumer may have to engage in many searches or evaluations to find a satisfactory fit. This may be too costly and result in the consumer avoiding making a choice altogether. If too few alternatives are offered, then the consumer may not search or choose, fearing that an acceptable choice is unlikely. These two forces result in the existence of a finite optimal number of alternatives that maximizes the probability of choice.

The price-aggressive discount format, popularized by chains such as Aldi and Lidl, is very successful in most Western economies. Its success is a major source of concern for traditional supermarkets. Discounters not only have a direct effect on supermarkets' market shares, but they also exert considerable pressure to improve operational efficiency and/or to decrease prices. We use an empirical entry model to study the degree of intra- and interformat competition between discounters and supermarkets. Information on the competitive impact of new entrants is derived from the observed entry decisions of supermarkets and discounters in a large cross section of local markets, after controlling for a number of local market characteristics. In our modeling framework, we endogenize the number of retailers and allow for asymmetric intra- and interformat competitive effects in a flexible way. We apply our modeling approach to the German grocery industry, where the discount format has stabilized after two decades of continued growth. We find evidence of intense competition within both the supermarket and discounter format, although competition between supermarkets is found to be more severe. Most importantly, discounters only start to affect the profitability of conventional supermarkets from the third entrant onwards. This may explain why many retailers rush to add a discount chain to their portfolio: early entrants may benefit from the growth of the discount-prone segment without cannibalizing the profits of their more conventional supermarket stores.

Many businesses track repeat transactions on a discrete-time basis. These include (1) companies for whom transactions can only occur at fixed regular intervals, (2) firms that frequently associate transactions with specific events (e.g., a charity that records whether supporters respond to a particular appeal), and (3) organizations that choose to utilize discrete reporting periods even though the transactions can occur at any time. Furthermore, many of these businesses operate in a noncontractual setting, so they have a difficult time differentiating between those customers who have ended their relationship with the firm versus those who are in the midst of a long hiatus between transactions. We develop a model to predict future purchasing patterns for a customer base that can be described by these structural characteristics. Our beta-geometric/beta-Bernoulli (BG/BB) model captures both of the underlying behavioral processes (i.e., customers' purchasing while “alive” and time until each customer permanently “dies”). The model is easy to implement in a standard spreadsheet environment and yields relatively simple closed-form expressions for the expected number of future transactions conditional on past observed behavior (and other quantities of managerial interest). We apply this discrete-time analog of the well-known Pareto/NBD model to a data set on donations made by the supporters of a nonprofit organization located in the midwestern United States. Our analysis demonstrates the excellent ability of the BG/BB model to describe and predict the future behavior of a customer base.

Can negative information about a product increase sales, and if so, when? Although popular wisdom suggests that “any publicity is good publicity,” prior research has demonstrated only downsides to negative press. Negative reviews or word of mouth, for example, have been found to hurt product evaluation and sales. Using a combination of econometric analysis and experimental methods, we unify these perspectives to delineate contexts under which negative publicity about a product will have positive versus negative effects. Specifically, we argue that negative publicity can increase purchase likelihood and sales by increasing product awareness. Consequently, negative publicity should have differential effects on established versus unknown products. Three studies support this perspective. Whereas a negative review in the New York Times hurt sales of books by well-known authors, for example, it increased sales of books that had lower prior awareness. The studies further underscore the importance of a gap between publicity and purchase occasion and the mediating role of increased awareness in these effects.

We develop a conceptual framework about the impact that branding activity (the audiovisual representation of brands) and consumers' focused versus dispersed attention have on consumer moment-to-moment avoidance decisions during television advertising. We formalize this framework in a dynamic probit model and estimate it with Markov chain Monte Carlo methods. Data on avoidance through zapping, along with eye tracking on 31 commercials for nearly 2,000 participants, are used to calibrate the model. New, simple metrics of attention dispersion are shown to strongly predict avoidance. Independent of this, central on-screen brand positions, but not brand size, further promote commercial avoidance. Based on the model estimation, we optimize the branding activity that is under marketing control for ads in the sample to reduce commercial avoidance. This reveals that brand pulsing—while keeping total brand exposure constant—decreases commercial avoidance significantly. Both numerical simulations and a controlled experiment using regular and edited commercials, respectively, provide evidence of the benefits of brand pulsing to ward off commercial avoidance. Implications for advertising management and theory are addressed.

We examine a previously unstudied category of exchange items in which the true value is unknown to both the buyer and seller at the time of exchange but becomes known to both at a future time after the exchange. Real-world examples of such exchange items as in our study include forward contracts and fixed-fee turnkey contracts. We demonstrate that the discrepancy between the seller's willingness to accept (WTA) and buyer's willingness to pay (WTP) increases with (1) the level of uncertainty about the exchange item's value and (2) the exchange parties' level of risk aversion. In a series of studies, we manipulate and measure the level of uncertainty of the exchange item, measure the level of risk aversion of the exchange parties, and study the respective effects on decreasing the WTP while increasing the WTA.

Choice decisions in the marketplace are often made by a collection of individuals or a group. Examples include purchase decisions involving families and organizations. A particularly unique aspect of a joint choice is that the group's preference is very likely to diverge from preferences of the individuals that constitute the group. For a marketing researcher, the biggest hurdle in measuring group preference is that it is often infeasible or cost prohibitive to collect data at the group level. Our objective in this research is to propose a novel methodology to estimate joint preference without the need to collect joint data from the group members. Our methodology makes use of both stated and inferred preference measures, and merges experimental design, statistical modeling, and utility aggregation theories to capture the psychological processes of preference revision and concession that lead to the joint preference. Results based on a study involving a cell phone purchase for 214 parent-teen dyads demonstrate predictive validity of our proposed method.