تقسیم‎بندی بازار در بازار اینترنتی

This paper presents a model to describe the competitive dynamics of web sites on the WWW market and analyze the stability of the model which is composed of one powerful site and two small sites. One of the most important results that emerge from this simple model is that strong competition among websites does not necessarily lead to the demise of the small website on the WWW market. From the stability analysis of the model, we obtain a series of conditions in which small sites can obtain competitive advantages by using the market segmentation strategy.

Recent surveys show that a few web sites have grabbed almost all the customers while other sites become extinct [1] and [2]. This so-called ‘winner-take-all’ phenomenon was proved to come from increasing competition among web sites [3]. The sites with the maximum market share will usually win in the competition and the others will become extinct. Maurer and Huberman first presented a simple model [3] and pointed out that the model will lead to transition from a number of sites sharing the market to the winner-takes-all market, with increased competition among the web sites. The model can be described by the following differential equations: where fifi is the fraction of the market which is a customer of site ii, αiαi is the growth rate which measures the capacity of site ii to grow, ββ is the maximum capacity which expresses the saturation value of fifi and γijγij is the competition rate between sites ii and jj. The greater the fijfij, the greater the loss of users of site ii caused by existence of site jj will be. Further works will focus on two main aspects: (i) try to find new models which can better characterize the real market; (ii) try to find proper strategies to help small sites avoid extinction. Luis Lopez, Miguel A.F. Sanjuan [4] and Wang, Wu [5] show that when taking both competition and cooperation into consideration, the web site competition model can show some interesting phenomena which cannot be found in Ref. [3] and propose a series of strategies to help small sites survive in the competition. Wang and Wu [5] and [10] propose a cooperation–competition model for the WWW market with the assumption that the intrinsic growth rates and maximum fractions are equal. Jiang and Cheng [6] give the complete strategic classification for the model proposed in Ref. [5] without any hypothesis. Xiao and Cao [7] take the time delay in the process of competition into consideration. Li and Zhu [8] and [9] emphasize the ‘richer get richer’ phenomena in the process of e-commerce web sites competition, and propose a competitive model of e-commerce web sites. So, what we are interested in is: apart from cooperation, is there any other way for small sites to avoid extinction? Market segmentation is one of the most commonly used strategies in real market competition. Then how can small sites avoid extinction or even win in the competition by market segmentation? In this article, we present a model to characterize competition between web sites in a market with a market segment. And by analyzing the stability of the model which is composed of one powerful site and two small sites, we get a series of conditions for small web sites to try market segmentation strategy.

In summary, different from the model mentioned in Ref. [3], we introduce market segment into the known competition model and present a model that accounts for competition both in whole market and market segment. We have found that market segmentation gives rise to unexpected strategies for small sites to win. All the analysis above indicates that small sites can increase their market share and survive or win in the market by market segmentation, if one of the following holds: (1) If the segment is more than half of the whole market and the powerful site weakly competes against small sites, while small sites strongly compete against the powerful site, and the two small sites strongly compete with each other as well as the initial market share of the three sites makes system equilibrium at O23O23. (2) If the segment is more than half of the whole market and the three sites strongly compete with each other as well as the initial market share of the three sites makes system equilibrium at O13O13 or O23O23. (3) If the powerful site and the small sites strongly compete with each other, while the two small sites compete with each other weakly and the initial market share of the three sites makes system equilibrium at O123O123. When View the MathML sourceθα0α0+γ0>1−θ, small sites can surpass the powerful one and become winner of the market. (4) If the segment is more than half of the whole market and the powerful site and small sites weakly competes against each other, while the two small sites compete with each other strongly and the initial market share of the three sites makes system equilibrium at View the MathML sourceO123∗. When View the MathML sourceα0(α1−γ10)α1α0−γ10γ01<α1(α0−γ01)θα1α0−γ10γ01, the small site can even become the winner of the whole market by market segmentation. However, there are different aspects which are not involved in. Further research can be done to proof for the most general case which parameters of web sites are not the same remains open. To apply the model to simulate the competition in the real internet market, parameters of the model need to be evaluated by Econometrics method. Despite the imperfections, we believe that the model characterizes the market with market segment properly. And analysis of the three site system can actually exhibit useful predictions. This may be helpful for better understanding the complexity of competitive dynamics in internet market with market segment.