In this paper, we set up a game theoretical model in which an incumbent and an entrant choose their respective R&D strategies to compete with each other. Our paper contributes to three major debates regarding a firm's R&D strategy; the incumbent's and the entrant's choice between a radical R&D project and an incremental one, the incumbent's decision of whether to pre-empt the entrant, and the trade-off between product quality improvement and time-to-market. Our model considers three decisions that both the entrant and the incumbent need to make: (1) the amount of investment in the R&D project, (2) timing of new product introduction, and (3) the magnitude of performance improvement. These three decisions completely define a firm's R&D strategy. As for the product performance improvement, the entrant has to decide whether to introduce an incrementally improved product or a drastically improved one. With the incrementally improved product, the entrant can enter the market earlier at a low R&D cost, but it will lag behind the incumbent in product performance. On the other hand, with the drastically improved product, the entrant may be able to leapfrog the incumbent in product performance, but the downside is that the entry needs to be delayed or more investment in the R&D project is needed. The entrant's strategy is further affected by the incumbent's potential reaction. The incumbent, anticipating the entrant's entry, can react by either improving its product incrementally or drastically. The incumbent can even pre-empt the entrant's entry by introducing its new product before the entry occurs. We find that, when trading off time-to-market against quality improvement, both the incumbent and the entrant should emphasize quality improvement over time-to-market. Specifically, the entrant should enter the market with a drastically improved new product, even if it means that the entry has to be delayed. The incumbent, anticipating the entrant's move, should react by introducing a drastically improved new product as well. Therefore, in the debate about the relative importance of time-to-market and quality improvement, we side with the school of thoughts that emphasizes the latter. Furthermore, we find that there is no need for the incumbent to pre-empt the entrant's move, i.e., the incumbent should introduce its new product only after entry even if it is certain about entry at the very beginning. This is different from Gilbert and Newbery's (1982) finding that the incumbent should pre-empt the entrant, and it is also different from Kamien and Schwartz (1972) who show that the incumbent will delay its R&D indefinitely in the face of competition. As for the incumbent's and the entrant's choices of either a radical innovation or an incremental one, our finding is consistent with the conventional wisdom that the entrant should introduce a radically innovative product (Day and Shoemaker 2000). However, our findings on the incumbent's choice of R&D project differ from any existing studies that advocate a cautious approach (e.g., Reinganum 1983). In recent years, many leading companies in various industries have come to the same conclusion and began to invest in more drastic R&D projects (see Chandy and Tellis 2000), as advocated by our findings. Our findings are also different from other studies in that we have produced clear and easy-to-understand findings. In contrast, many existing studies (e.g., Ali, Kalwani and Kovenock 1993; Cohen, Eliashberg, and Ho 1996; Bayus, Jain and Rao 1997) provide more nuanced but also ambiguous results that depend on market conditions that are difficult to empirically validate. The intuition for our major finding that both the incumbent and the entrant should engage in radical R&D projects is relatively straightforward. Consider the entrant first. If the entrant enters the market with an incrementally improved product, such is not forceful enough to challenge the incumbent's dominant position; hence, it is a dominated strategy by a drastic R&D project. And, knowing that the entrant's product will be always drastically improved, the incumbent should react likewise to protect its leadership position. The optimal entry time for the entrant and the reaction time for the incumbent are determined according to the trade-off of R&D costs and firm profits, and in equilibrium, the incumbent will not pre-empt entry to minimize its R&D costs. We also study the scenario where the incumbent does not anticipate the entry. In this case, in order to take advantage of the incumbent's delayed reaction, the entrant should accelerate its entry, but should not change its product strategy, that is, is should still introduce a drastically improved new product. Although being caught off guard by the entrant's surprise entry, the incumbent should not react hastily. Rather, it should proceed with the same pace in its R&D process as in the previous case and introduce a drastically improved new product.
