This study investigates the issue of pricing and innovation for technology-intensive products when there are overlapping introductions of successive generations with rapid technological improvements. The goal of the research is to discover the optima/pricing strategy focusing on the relationship between the prices of old and new generations, the quality improvement between the two generations, and the length of introduction intervals. Our model is distinguished from the prior work by capturing two important features of high-tech product markets with sequentially introduced generations. First, the consumers make intertemporal purchase decisions on whether and when to participate in the market at each introduction. Second, the manufacturer's profit depends not only on the sales of the new generation but also on the sales of the old generation products still in the market. We found three possible equilibria and identified the one that a high-tech manufacturer would choose in the real world based on current market parameters. In this equilibrium, the manufacturer chooses a strategy that makes the older generation attractive to consumers, which means the firm chooses to protect the older generation products from cannibalization by the new generation. Another characteristic of this equilibrium strategy is that the introduction price of the new generation increases with the quality improvement between the old and new generations but decreases with the length of the interval between two introductions. Thus, the manufacturer has an incentive to increase the price of the new generation product with higher consumer willingness to pay when the quality improvement raises consumers' perceived obsolescence in the older generation, and when consumers' cost of waiting is reduced by a quick arrival of the new generation. Empirical support for the theoretical result is provided using Intel microprocessor pricing and quality data. The findings of this research can be applied as a useful pricing decision tool for various technology-intensive products such as computer hardware and software.
Keywords: Technology Strategy, Technology Management, Obsolescence, Successive Generations, Technological Innovation, Pricing Strategy
Moore's law predicts that the number of transistors in an integrated circuit, and along with it the computational power of the circuit, doubles about every two years (Moore, 1965). The high-tech industries, especially the semiconductor industry, have kept the pace that Moore's law forecasts since it was first proposed. Moore's law also implies the exponential decline in price per transistor or, in other words, price for performance (Moore, 1965; Aizcorbe, 2005; Aizcorbe and Kortum, 2005). According to the data reported in Aizcorbe and Kortum (2005), the average introduction price of Pentium I, II, and III families remained similar--$516 for Pentium I in 1993, $587 for Pentium II in 1997, and $506 for Pentium III in 1999--while the growth rates in number of transistors in a chip between Pentium I and II and between Pentium II and III were 53% and 45%, respectively. With a simple glance at this data, one may conclude that the price of a new processor family is attractive to consumers when the new generation is introduced, considering the rate of technological improvement from the previous processor family.
However, consumers face a more complicated purchase decision problem than the one described above. Many high-tech product manufacturers continue selling previous generation products along with the newly introduced generation as long as the older generation produces profit (Aizcorbe and Kortum, 2005). Thus, the consumer decision problem encompasses the following elements: whether to buy only the first generation at the time when it is introduced, to wait and buy the first generation when its price drops, to wait and purchase only the second generation, or to buy each generation as they are introduced.
The price of high-tech products declines over their lifetime. For instance, by the time the new Pentium II chips were introduced, the price of Pentium I chips may have been reduced enough to compete with the introduction price of Pentium II chips even considering the quality and performance improvement. Pricing sequential, overlapping generations of a technology-intensive product line is a complicated task. If the price of the new generation is too high, it will fail to attract enough consumers. In contrast, pricing a new generation too low may result in a cannibalization of the old generation of the product.
We develop a game-theoretic model where a high-tech product manufacturer introduces successive generations and consumers make inter-temporal decisions regarding their purchases. The model considers a monopolist who sells sequential and overlapping generations, that is, the firm sells a first generation product along with a second generation product. Our model is distinguished from prior work by capturing several important features of high-tech product markets with sequentially introduced generations. First, the consumers make inter-temporal purchase decisions on whether to buy the currently available product immediately or in the following period. Second, the manufacturer's profit depends not only on the sales of the newer generation, but also on the sales of the older generation products. Third, we use data on Intel's microprocessor sales to present an extensive empirical test that supports the equilibrium solution of the model.
The model is designed specifically for high-tech products with rapidly improving quality. Technology-intensive products such as computer components or software are characterized by the rapid, sequential introduction of new and improved generations. The successive introduction of a technology-intensive product should be distinguished from that of other durable goods because of the significant performance improvement at each introduction (Dhebar, 1996). This innovation in technology-intensive products causes significant product obsolescence, and it subsequently leads to frequent purchases by consumers. Consumers usually obtain a new generation of non high-tech durable goods when the current one does not function properly anymore. For instance, a consumer may buy a new model of a refrigerator when the current one needs a replacement due to performance deterioration. On the other hand, a consumer may purchase a new personal computer (PC) because the current PC is not powerful enough compared to the new PCs in the market, even though the PC performs as well as when originally purchased. In other words, the motivation for repeat purchases of non high-tech durable goods is replacement due to breakdown, whereas the motivation for technology-intensive product replacement is obsolescence. This obsolescence is external to the product in question, and occurs due to economic and technological changes in the environment. Levinthal and Purohit (1989) define product obsolescence as the relative loss in value due to styling changes (style obsolescence) or quality improvement (functional obsolescence). Repeat purchases due to the obsolescence of technology-intensive products are highly associated with the extent of the improvement in the new generation.
Another facet of the consumer problem in buying technology-intensive products is the time value of availability. Given the fast technological improvement, consumers face a "buy or wait" decision problem (Kornish, 2001). Briefly, they have to decide whether to buy the currently available product, or to wait for the improved product available in the future. In such consumer decisions on inter-temporal tradeoffs, a discount factor plays the key role in assessing the present value of the future asset. We utilize the discount factor to represent the product introduction speed throughout the paper. In general, when the time interval between introductions is longer, consumers will discount the future product more, and the discount factor increases.
We found three possible equilibria and identified the one that a high-tech manufacturer would choose in the real world given the current market parameters. In this equilibrium, the manufacturer chooses a strategy that makes the older...