روش شبکه های بیزی برای بررسی عوامل کلیدی موفقیت بازی های تلفن همراه

As mobile game business becomes one of the most lucrative as well as fast-growing businesses, examining key success factors in this industry is of great interest. Utilizing a research method called Bayesian network, this paper models and tests interrelationship among product, marketing, consumer and competition variables. The current study surveys experts who launch many games in Korea. The three most crucial factors for successful games turn out to be targeting, awareness and consumers' willingness to pay (WTP). Many of the other factors influence the performance of games via these three factors. This paper not only investigates into the sensitivity of game performance to targeting and awareness levels but also examines the influences of product/marketing variables on consumers' first impression or willingness to pay. The findings on the roles of product or marketing factors that affect consumers' perceptions and responses, thereby competitiveness and success, will help game makers and distributors make reasonable decisions in allocating corporate resources more efficiently.

The mobile game industry is growing fast. Due to rapid diffusion of mobile phones all over the world, the mobile game holds a number of exciting possibilities for new business models and growth strategies (Sharp & Rowe, 2006). The combinations of high-end devices, large developer communities, and ever-expanding retail platforms provide industry participants with various opportunities. The recent hyper-growth of the smart phone market also creates a fruitful basis for the leap of mobile games. Highlighting the significance of the mobile games is the fact that almost a third of the available titles on the application stores, such as Apple's AppStore, are games. Social network services such as Facebook also become one of the key platforms of so-called social games. Despite the importance and popularity of the mobile games, little research explains what the key success factors are in this business. The purpose of the current study is to explore the interrelationships among various factors at product and consumer levels, thereby examining their direct and indirect effects upon the success of individual mobile games. The structure of the paper is as follows. First, this article outlines perspectives of game-related studies as well as issues in assessment of new product performance. The following section explains a useful research method called Bayesian networks approach. The next section presents the results of an empirical analysis on mobile games using Bayesian networks model. Some of the sensitivity analysis results follow. Finally, the paper concludes with a discussion on the implications and the limitations of the present study.

This study utilizes a Bayesian network approach to identify and examine the success factors of mobile games. BNs are useful in that the posterior probabilities help researchers determine the effect of the predictor variables on the target or the performance variable. One of the key advantages of the method lies in the fact that it can nicely handle the situation where the variables are correlated and thus interdependent. BNs enable us to understand the true impact of causal relations between certain variables in terms of their conditioned probabilities. BNs also allow us to perform what–if analyses for different hypotheses in relation to some variables of a model, and to determine the value of information in terms of the reduction of the level of uncertainty. The Bayesian nature of the network enables planners to improve the accuracy of the networks as their experience and expertise grow, to update information, and to simulate the impact that changes of factors have on the prospects for the new product. The conventional approach to knowledge development is largely theory driven. Researchers test hypotheses about the relationships among the variables of interest. Research without a theoretical base is often considered lacking in intellectual merit and analytical rigor, but the current environment demands more problem-oriented research and feasible methods to explore the vast quantities of disaggregated data. Given the increasing amount and variety of data, a BN approach provides an efficient tool for managers to extract and update knowledge in a timely fashion (Cui et al., 2006). However, limitations remain in this method too. Above all, BNs are discrete in nature. Although discretization simplifies the learning process and the resulting model, there should be a loss of potentially useful information, and the model may not fully capture all the details of the relationships. The integration of existing domain knowledge using supervised learning is a fruitful venue for future research. The current study has aspects to improve upon due to its exploratory nature. The authors collected important variables based on the literature and the industry experts' advice. Although examining whether the variables are exhaustive and mutually exclusive might be helpful, gathering all of the data worthwhile for considering is beyond our capability. For example, the current model does not take into account other factors such as genre characteristics or game platforms. Finally, it is still unanswered for why awareness does not play a role in medium product-target fit. A further study is warranted to explain the exact mechanism behind this effect.