Relating Inertia and Experience in Technology Markets: An Analysis of Households' Personal Computer Choices

Abstract:     
Using a rich, micro-level dataset on households and their personal computer (PC) purchases, this paper analyzes PC purchasing behavior at the household level. Following the literature on experience and search goods, we begin by building a simple learning model of households' brand choices over time. We then describe the model's empirical implications for the relationship between brand choice and market experience (measured as a household's lifetime number of PC purchases). In our empirics, we have three main findings. First, we find that households generally exhibit inertia in their PC purchases, meaning that a previous purchase of a given brand increases the probability of a current purchase of that brand. Second, we find that the level of inertia is increasing as a function of experience on the PC market. That is, experienced households are more likely to stick with the brand they purchased most recently. Third, we find that, for households switching brands, the likelihood of buying an "obscure" brand (i.e., a brand with low market share) increases with experience, regardless of the brand of the previous purchase. The last two findings pertain to the relationship between brand purchasing and market experience, and are consistent with the two predictions of our learning model. We then combine these findings with further analysis to rule out several alternative models of purchasing behavior. These findings extend our understanding of how market experience affects purchasing behavior to an important technology product, with implications that may apply to other similar products (e.g., televisions and mp3 players). The results show that experienced households require larger incentives to switch from their previous brand purchase. Consequently, PC brands will be perceived as more differentiated and, ceteris paribus, barriers to entry may increase over time (as more households become experienced with the PC market); however this latter effect is tempered by a higher propensity to try obscure brands among experienced switchers. Further, the results are consistent with the idea that market experience helps households establish their favorite brands, and if such learning is indeed occurring, they suggest that inducing a household to switch brands (e.g., via a one-time price cut) likely won't result in any notable long-term returns.

Keywords: Technology good, personal computer, inertia, market evolution, brand choice, true vs. spurious state dependence, durable good, consumer learning

JEL Classifications: D12, L63, L11

Working Paper Series