Abstract
In this paper, we develop and estimate a series of Bayesian generalized gamma family and mixture family models of product modification timing that take into account both firm- and industry-specific effects that are time varying. Our models are capable of capturing non-standard modification behavior such as multi-modality and non-monotonicity in the hazard rates motivated by real data. Additionally, we explore the existence of latent groups with respect to product modification timing strategies. The models are estimated using Markov chain Monte Carlo (MCMC) methods using a panel data from the automotive industry that covers 50 years and contains 6598 car model-year observations for 683 car models. The results reveal the non-monotonic modification behavior over time and the existence of three latent groups. Larger product portfolios and higher industry product proliferation lengthen the modification time. Brand and competitive modification dynamism increases the frequency of major modifications.
Original language | English (US) |
---|---|
Pages (from-to) | 85-97 |
Number of pages | 13 |
Journal | Marketing Letters |
Volume | 28 |
Issue number | 1 |
DOIs | |
State | Published - Mar 1 2017 |
All Science Journal Classification (ASJC) codes
- Business and International Management
- Economics and Econometrics
- Marketing
Keywords
- Bayesian inference
- Finite mixtures
- Generalized gamma models
- MCMC estimation
- New product development
- Product modification