Measurement of competitive issues in product entry decisions

Gonzalez, J (2008) Measurement of competitive issues in product entry decisions. Doctoral thesis, University of London: London Business School.

Abstract

This thesis aims to make two types of academic contributions. The first is the development of empirical models in situations of new product entry in the marketplace. The second type of contributions is to provide substantive results regarding the impact of launching new products in sectors such as the mobile telephony, pharmaceutical, and motion picture industries. In Chapter 1, I develop a brand level diffusion model to measure the substitution and category expansion effects of the launching of a ucw product into an existing category. This model accounts for unobservable factors that have carry-over effects and influence the diffusion of the new product introduced in the market. I use the proposed model to analyze the differences in adoption of mobile phone services in France, Germany and the UK. I find that some of these differences can be explained by the changes in the adoption costs and the different impact of the launching of prepaid plans in the three countries. More specifically, I find that France, the least developed of the three markets in terms of total percentage of adoption, is the country where prepaid plans were less effective in terms of category expansion and where there was more substitution of mobile services with contract plans. This finding suggests that prepaid plans in France were less appealing for light users of mobile telephony than in the other two countries. Therefore, I claim that the lower penetration of mobile telephony in France can be explained by the customer acquisition policy of French mobile network operators, that was more conservative than the policies followed iu Germany and the UK. In Chapter 2, I analyze the impact on competitors of the launching of generic versions of a drug after patent expiry. Previous research on generic entry has focmmd on the impact of a generic on the branded drug which lost patent protection (i.e. within molecule competition). In this study, I suggest that public officials need to look beyond within-molecule competition and study the full competitive landscape in the relevant therapeutic class. Such a competitive landscape will include also the actions of non-bioequivalent competitors (between molecule competition). In addition, I contend that public officials need to consider carefully the role of physicians and their prescribing behavior across all competing molecules, and study not only their reaction to prices but also to other marketing activity. To provide support for this contention, I studied the evolution of the market for selective serotonine reuptake inhibitors (SSRI) in the United Kingdom (UK) after generic versions of fluoxetine (brand name Prozac) were introduced. Using a data set on physician prescribing and competitive marketing activity, I found that the combined sales of the drug that was losing patent protection (branded + generic versions) decreased after patent expiration, despite the availability of generics at significant price discounts. This reduction occurred because a segment of physicians switched from the multi-source molecule to other drugs in the category after generic entry (between-molecule effect). These were physicians sensitive to the promotional activities of Prozac, which significantly reduced its marketing effort after generic entry. In addition, although a smaller segment of price sensitive physicians increased the use of fluoxetine, that was not sufficient to compensate for the previous reduction in fluoxetine prescribing behavior. These results suggest that generic adoption is influenced by the actions of all competitors, not just the original branded version. Therefore, to fully understand the impact of generics, it is important to account for the marketing activity of all pharmaceutical companies that are close substitutes and determine how physicians respond to the marketing actions and drug prices at the individual level. Finally, in Chapter 3, I measure the effect of the retail channel on brand performance using data on consumer purchases across competing retailers in the context of the movie exhibition market. I develop a model to measure the relative performance of competing theaters that takes into account the differences in the assortment of movies. This model accounts for the temporal evolution of box-office revenues and the exhibitor decision to show a movie for another week or to change it for a new release. I use this model to compare the relative performance of two chains of theaters in a Spanish city. I find that one of the theater chains performs, on average, consistently better than the other. As a result, a movie that is shown in this best performing chain is more likely to perform well, all else constant. However, brand strength plays a moderating role: differences in performance across the two chains are more important for less popular movies; strong movies present smaller differences across the two chains. From these results, I suggest that strong brands rely less than weaker brands on retailers for their performance. These results highlight further the importance of developing strong brands: not only will weaker brands be bought less often by commmers and carried less often by retailers, weaker brands will abo have less bargaining power with channel members because their performance will depend more on the "quality" of the channel members.