نفوذ بازار در میان محصولات ابتکاری رقابتی: مورد از سیستم عامل گوشی های هوشمند

Competition models have seldom considered the future development of products. However, to make decisions about products, it is key to understand their future demand and competition with other products. In the present study, we considered four Smartphone Operating Systems (OSs) (Android, iOS, Symbian, and Blackberry), and proposed an integrated model that combines scenario analysis and the Delphi to predict possible scenarios for the future development of the four OSs. Then we used the Lotka–Volterra competitive model and an innovation diffusion model to forecast the adoption volume of each OS over the next 5 years. We suggest strategies for decision makers.

Regarding sales volume forecasting, the logistic and Gompertz models are popular technology forecasting models, whereas the also-popular Bass model (Bass, 1969) is used to forecast sales for the timing of initial new product purchases. These three popular new product diffusion models all assume that the product in question is independent of any other product. However, high-tech products/technologies tend to have short life cycles and high competition and substitution effects. Therefore, estimating the sales volume among competitive innovation products/technologies has become increasingly critical in this era. Three models focus on sales forecasting among technologies, Peterson and Mahajan (1978) extended the Bass diffusion model (1969) to include more than one innovation simultaneously. They specifically conceptualized four extensions of the fundamental diffusion model, depending on whether innovations are independent, contingent, complementary, or substitutes. In 1993, Bucklin and Sengupta (1993) described the concept of co-diffusion, which is the positive interaction between the demands of complementary innovations that have separate adoption tracks. The interaction arises because the adoption of one innovation enhances the value of the other to the end-user. The Lotka–Volterra equation, which had been developed to model the interaction between two competing species, has been adapted to model competitive market situations (Kreng and Wang, 2009). For example, Lopez and Sanjuan (2001) analyzed the effects of competition among websites. Kim et al. (2006) explored the Korean mobile phone market and found a set of commensal relationships. These proposed models have focused on the competition relationship, however, the future development of competing products is contingent on various external developments not included in these models, such as sales volume forecasting, technological breakthroughs, societal backlash, and so forth. Moreover, the development of new high-tech technologies involves many considerations (including societal, technological, economic, environmental, and political issues), may be hindered by future uncertainty, and is often limited by a relative lack of data in early stage. At this time, how do experts forecast the sales volume of competitive products/services? Traditional scenario analysis is a qualitative method, which considers macro- and micro-economic factors as well as future uncertainty to present rich and complex portraits of possible future scenarios (Porter et al., 1991). Until, Wang and Lan (2007) and Tseng et al. (2009) combined scenario analysis (to address future uncertainty) and quantitative multi-generation sales forecasting methods to analyze the future development of new-generation technology. Generally speaking, researchers collect the opinions of experts when conducting scenario analysis; however, experts’ opinions often vary greatly. Therefore, some researchers combine scenario analysis and the Delphi method to generate future scenarios (Rikkonen and Tapio, 2009 and Czaplicka-Kolarz et al., 2009). Tseng et al. (2009) and Tseng and Wang (2011) combined scenario analysis with the innovation diffusion model and the Delphi method to analyze the development of a new technology. Dong et al. (2013) reviewed a generic step process for scenario development and suggested that there are three limitations to the current practice: the number of quantitative scenarios developed, the lack of probabilities attached to the scenarios, and the lack of transparency in how descriptive scenario storylines are converted into quantitative scenarios. Dong et al. (2013) tested scenarios with qualitative and quantitative techniques and identified two limitations of applied quantitative techniques: (i) the need to extend discrete scenarios to continuous scenarios to more completely cover future conditions and (ii) the need to introduce probabilistic scenarios to explicitly quantify uncertainties. They suggested that these limitations can be overcome by using computational algorithms to develop a large number of scenarios and then using Bayesian probabilities to narrow them down. Önkal et al. (2013) proposed using scenario analysis to aid in quantitative judgmental forecasting. Varho and Tapio (2013) used the Q2 technique to produce forward-looking and heuristic scenarios that combined qualitative and quantitative Delphi methods. Based on these studies of combination methods, the present study proposes an integrated model to analyze the relative development and sales forecasts among competitive innovation products/technologies with future uncertainty and limited data. We combine scenario analysis and the Delphi to develop scenarios and the experts’ forecasts, the diffusion model and the competitive model to forecast the sales volume of each product/service in the scenarios. Mobile phone types can be divided into basic phones, feature phones, and Smartphones. However, Smartphones have invaded the feature-phone market, due to their multiple features and applications, and their share of the overall mobile phone market has continued to grow (Chen, 2009). Previous studies have found that mobile services drove the growth in the adoption of the mobile phone. Sharma and Xiaoming (2012) developed and tested a value model for the adoption of mobile data services in international markets. Tojib and Tsarenko (2012) proposed a post-adoption model of advanced mobile services that emphasizes relationships between actual use and its antecedents: service ubiquity and experiential value. Moreover, as Smartphone sales have continued to grow, mobile phone providers’ innovation is concerned not only with telecommunications functionality but also operating systems (OSs) platform integration. Platforms have become a major battleground as device to rich user experiences (Gartner, 2009). Moreover, the successful development and future market share of new software and applications depend heavily on the success of their OSs. Hsieh, 2010a and Hsieh, 2010b noted that severe competition exists among OSs, especially iOS and Android. Karippacheril et al. (2013) analyzed how operator, device, and service provider centric platforms compete, collaborate, to deliver scalable services to the poor. OSs are likely to play an important role in Smartphone development. However, few researchers have focused on competitive diffusion among Smartphone OSs. Therefore, we chose Smartphone OSs to be our example. The remainder of this paper is organized as follows. “Competitive models” section describes the three competition models, “Methodology” section describes the proposed integrated model, “Empirical study: the development of Operation Systems” section provides the empirical analysis, and final section presents the research results and a discussion of the findings.

We focused on the development of Smartphone OSs during the next 5 years. The key factors influencing the development of Smartphone OSs are “consumer performance,” “degree of development with application stores,” “variation of global market growth,” “trend of global economic,” “quality of source software,” “open software development kit,” “degree of competition within operating system,” and “share of development information.” We determined driving forces, grouped high-impact and uncertain drivers into three axes (“competition among OSs”, “openness and cooperation among industry chain,” and “service providing”), and developed three scenarios (optimistic, pessimistic, and most likely). OS companies should note these key decision factors and driving forces when making decisions. The competition model revealed a predator–prey relationship between Blackberry and Android, where Android is the predator in the pessimistic scenario. Symbian suffers from the existence of all other platforms. The diffusion of Blackberry, iOS, and Android have no effect—positive or negative—on other platforms, and there are no significant substitute phenomena among the tested platforms. However, these three OSs substitute the market of Symbian. According to sales forecasts, Android will grow rapidly and, in all three scenarios, exceed Symbian sales in 2011 to become the leading OS in the market. In the optimistic scenario, Android captures more than 50% of the OS market by 2015, with a sales volume of more than twice that of iOS, while iOS will surpass Symbian in 2012. The peak of Symbian sales will occur in 2011. In the pessimistic scenario, the sales volume of Android will be about 1.84 times that of iOS in 2015. The sales volume of Blackberry will begin to fall in 2013 and that of Symbian will begin to fall in 2011. In the most likely scenario, the sales volume of Android will be around 1.97 times that of iOS in 2015, and iOS will surpass Symbian in 2013. Before 2014, the sales volumes for the four OSs will be as follows, in order of most to least sales: Android > iOS > Symbian > Blackberry. This study suggests future work that will extend our research. An analysis of the competitive relationships among OSs at the same price level could illuminate the advantages and weakness of these OSs. In the current study, limited data prevented the use of the Lotka–Volterra model to simultaneously compare five platforms. Future studies using larger data sets will permit us to more thoroughly explore the competitive relationships among the five platforms. Besides, future studies can richer models of innovation and market competition both in the agent-based method and system dynamics method.