Article ID: | iaor20032200 |
Country: | United States |
Volume: | 48 |
Issue: | 9 |
Start Page Number: | 1175 |
End Page Number: | 1195 |
Publication Date: | Sep 2002 |
Journal: | Management Science |
Authors: | Golder Peter N., Bohlmann Jonathan D., Mitra Debanjan |
Keywords: | game theory, marketing |
Several studies have demonstrated an order-of-entry effect on market share, suggesting that pioneers outperform later entrants. However, other research has pointed out the limitations of these studies and found evidence that many pioneers fail or have low market share. Given this background, the purpose of this research is to understand the conditions under which pioneers are more likely and also less likely to have an advantage. We propose a game-theoretic model that includes important sources of pioneer advantages as well as disadvantages. Specifically, we incorporate a pioneer advantage due to preemption in markets with heterogeneous tastes. In addition, we incorporate a potential pioneer disadvantage due to technology vintage effects, where later entrants utilizing improved technology can have lower costs and higher quality. The model allows us to evaluate the extent of vintage effects necessary to overcome a pioneer's advantage. Key relationships are found between the magnitude of the pioneer advantage or disadvantage and consumer valuations of product attributes (e.g., variety and quality). We empirically validate the model with vintage effect data in 36 product categories, and measures of consumer valuations of product variety and quality for 12 of these 36 categories. The results show that pioneers do better in product categories where variety is more important and worse in categories where product quality is more important. Pioneers in categories with high vintage effects are shown to have lower market shares and higher failure rates. Similar results appear when analyzing persistence of market leadership over time, further validating our model's major implications. We also present two case studies that illustrate key elements of the model.