E-commerce evolution based on the perspective of network effects: A multi-case study in China Peng Shao, Ping Hu, Jie Qi School of management, Xi'an Jiaotong University, Xi an, China Abstract This paper, which is based on network effect theory, focused on the key factors of two-sided network construction and e-commerce ecosystem formation in a big data environment. It takes six e-commerce companies as a research case and suggests how to improve the timing and spacing of service precision and business credit, reduce the experience gap between online and offline commerce, increase the members in e-commerce networks, and ameliorate the form of e-commerce ecosystems. Based on the several propositions, we discuss e-commerce evolution in four stages and propose new perspectives on two-sided network effect improvement in the big data era. Keywords: big data; e-commerce; two-sided networks; network effects 1 Introductions With the wide application of a new generation of information technology, such as social networking, cloud computing, and mobile commerce, global data is growing at an unprecedented speed [1]. Nature published a special "Big Data" issue in September 2008, and Science launched its own special issue in February 2011 to explain the importance of big data in scientific research. Meanwhile, the academics of China have also become interested in big data related areas, such as business management, technological innovation, and interdisciplinary research. There are large amounts of data in e-commerce that can be used to segment the market and meet customer needs with customized products and services. In a network situation, customer behavioral data in all channels and all stages of growth can be recorded or collected by enterprises, and 1 accurate and quantifiable marketing strategies can be designed [2]. E-commerce is a two-sided market; the research on two-sided markets usually takes pricing strategy as a starting point; it analyzes the impact of supply and demand characteristics on the profits of platforms, for which on optimal pricing strategy is developed. E-commerce platforms have a two-sided market effect and the interdependence of the two sides play a driving role in development of platforms. Unstructured data, such as relational, location, video data, and images, emerge in large numbers in the big data era and function as the driving resource in the e-commerce model transformation. The existence of a two-sided network effect is based on the demand to resolve big data, and new information technology can facilitate the matching of supply and demand in platforms. Based on the two-sided network theory, research propositions are made in this paper through a case study of six e-commerce businesses; further, the evolution of an e-commerce model is discussed in the four stages of the big data era, and a new perspective on the improvement of the two-sided network effect is developed. 2 Theoretical Review Network effects are largely influenced by the number of users. Katz and Shapiro (1985) [3] point out that there are direct and indirect network effects. The term direct network effect refers to increases in product value with increases in the number of users, including new users; this increase is known as a demand-side network effect. Indirect network effect refers to the product value derived from the number of complementary products. Indirect network effects will be strong when there are various kinds of complementary products, and the consumption of
core products and the value of other products will increase. In addition, network effect can be divided into local network effects and global network effects. Local network effects are caused by family and friends and global network effects by the installed base. As a consumer s utility depends on the number of interactions with his or her friends, rather than the size of the overall network [4], in markets with network effects, companies seek to remove features of its original products and sell degraded versions of them at low prices or for free [5]. There are two kinds of users in a two-sided market, and value is created by the interaction between consumers and businesses on the platform [6]. The revenue of participants on the platform depends on the number of participants. Cross-platform network effects should be considered when business strategies are formulated so that these can be rational. Generally, network product markets have the following characteristics in two-sided network effects [7-9] : First, user groups are often very concerned about the scale of the other side of the platform because their needs are interdependent. Second, each member of a group on one side is willing to join the network only when the group members of the other side are expected to join it. Third, the platform sponsor is often quite inclined to a price structure that matches the needs of the two-sides, since the demands of the users on both sides have an asymmetrical interdependence. Proposition 1: In a two-sided market, the utility of players on one side is dependent on the number of players on the other side. Proposition 1a: In a two-sided market, the utility of business is dependent on the number of consumers. Proposition 1b: In a two-sided market, the utility of consumers is dependent on the number of businesses. 3 Study Design TABLE1 SAMPLE DESCRIPTION type Sample Description Differential & advantage Data acquisition Meituan.com, which was founded in March 2010, is Services accuracy: location based service, Interview with Marketing PR the No.1 website in the group-buying industry in information push Director on August 10, 2012; Group-buying & China. secondhand data collection mobile commerce Komovie.cn, which was founded in July 2011, is a Services accuracy: based on the remaining Interview with CEO on August 10, startup in the mobile commerce industry; it s main time, dynamic pricing on the remaining 2012; secondhand data collection business is movie tickets with dynamic prices. number of seats Dhgate.com, which was founded in 2004,is a foreign Credit: mechanism of honesty and guarantee, Interview with vice-president on trade platform for small businesses, based on online seller management, credit files, sunshine August 10, 2012; secondhand data Foreign trade trading plans collection e-commerce Globalmarket.com, which was founded in 2000, is a Credit: eight items of global manufacturer Interview with CEO on July 27, international e-commerce platform with certificate 2012; secondhand data collection self-developed audit standards. 500ccc.com, which was founded in 2011,is based on Minimize the experience gap between online Interview with vice-president on Combining of online the model of one city onenetwork andtransaction and and offline: one city one network to solve August 9, 2012; secondhand data & offline distribution services offline. the last mile problem of large appliances collection Alibaba.com, which was founded in 1999,announced Ecosystem:resource integration Interview with director of Taobao E-commerce the launch of seven business groups in 2012, it Extensive,third-party service providers settled President Office on August 19, ecosystem acquired shares of Sina Weibo in 2013. in the ecosystem 2013; secondhand data collection Based on the network theory, research propositions are Semi-structured interviews are proposed, based on theory made in this paper through the case study of six and literature review after the cases are selected. There are e-commerce businesses, and a new perspective on the four interviewers for each interview, one of whom is the improvement of the two-sided network effect is developed. local contact person; one is responsible for questioning The case study is the elementary method in management and clarification, and the remaining two are responsible research [10], and it is suitable for observation and research for photographing and recording. After the interview, on the series changes of businesses [11] ; it coincides with cases were saved in Word format, based on written notes the study of the evolution of e-business models in this and audio recordings. The findings of interview recording paper. The criterion of sample selection in the case study are combined with those of published studies, such as is particularity rather than the generality; four to eight is internal reports, journal articles, and speeches in order to the appropriate number of cases in a multi-case study [12]. ensure the reliability and validity of the collected data. 2
4 Analysis on the e-commerce revolution 4.1. Services accuracy improved by data mining With the emergence of new technology, such as blogs, social networks, location based services, and the rise of cloud computing, networking, and other technologies, data is growing and accumulating at an unprecedented rate. There are many kinds of data; the most common are unstructured data for fairly long periods, such as relational data, location data, images, and videos. The founder of komovie.cn: We know the positions, phone models, general consumption periods, and the consumption capacities of consumers. And we also know the time they enter and come out cinemas, and what kinds of movie they like. All this data can be used to provide precise marketing services. The Marketing Director of Meituan.com: The two-sided market will appear gradually. Omn the consumer side, Meituan.com want users to be able to enjoy good services at anytime. on the business side, Meituan.com is establishing a long- term relationships between businessesed and consumers. A large amount of data is generated every day by e-commerce, not only those that serve as records of objective phenomena or numerous unordered values, but also data with special meaning and value. Marketing management combines art and science; the scientific part depends on various data collection methods and marketing databases. Depending on the characteristics of consumption and behavior, modern technologies, methods, and strategies are adopted by enterprises in order to achieve the goal of marketing communication among consumer groups. The two-sides of an e-commerce platform involve consumers and businesses; further, an important issue concerns the precise matching of supply and demand. Information mining by mass businesses to meet consumer needs requires not only various types of data analysis, but convenient channels for information transfer between consumers and businesses. Proposition 2: The impact of consumer numbers on business utility can be enhanced by improvement of service accuracy in the two dimensions of time and space. 4.2. Transaction risk reduced by business credit rating Big data means high value, which may result from the mass convergence of data with minimum value and the 3 useful data needed for management decision-making. E-commerce transactions are conducted between suppliers and consumers who do not know each other; information asymmetry is an important factor that affects the development of e-commerce. Credibility identification is particularly vital in foreign trade e-commerce because of asymmetric information. The Vice president of dhgate.com: Sellers om the management department will review business licenses, registered capital, products, and other information to build up credibility files. Various phenomena of false credit are listed in the "unshine" project, which was set up in 2012; the project asks businesses to conduct self-examinations. If the same complaint is received again, a more severe punishment is applied the businesses. The CEO of globalmarket.com: It is hard to distinguish sellers with credit; the credit system (GMC) was launched to solve this problem. A survey is conducted and the result shows that there are 100,000 Chinese enterprises with GMC standards, accounting to 90% of exports in China. Globalmarket.com uses their energy to service valuable customers, including Dell, IKEA, and so on. Online service providers create better services by the online information that they provide, but they can also use the information to despoil users; thus, the key issue is to establish ethics on the Internet and to ensure that these ethics do not inhibit innovation [13]. Buyers are not only concerned about the number of sellers but also by comments on goods. The more positive comments, the better the products and services provided by sellers; in this way, direct network effects become greater. An online reputation is an important factor that affects e-commerce development; credit conditions should be enhanced in e-commerce transactions between buyers and sellers in order to increase the rate of online trading. More traders will join a platform through active participation when fair credit environments exist on them. Proposition 3: The impact of business numbers on consumer utility can be enhanced by the improvement of business credit. 4.3. The experience gap minimized by the integration of online and offline commerce The relationship between online commence and traditional commence is the hot issue in e-commerce research [14]. The level of e-commerce development varies by area, as does the e-commerce model and the perspective on the relationship between online and
traditional commerce. Overall, existing research findings can be divided into four perspectives: alternative, promotional, complementary, and independent [15]. With the rapid development of e-commerce, customer migration from stores is unprecedented; the consumer transformation has become a reality. Small bits of data are merged into big data. E-commerce can more easily accumulate and mine data than offline channels and should provide services based on this data processing. The founder of 500ccc.com: There is a big difference between online and offline in appliance selling. 1:10 is the difference between online and offline in the selling of clothes and 1:100 is the maximum in appliances. Order online and service Group divided the e-commerce business into three categories and 25 business units; the first is vertical businesses BU, including Taobao, Tmall, and Juhuasuan; the second is the infrastructure platform, including cloud computing, payments, logistics, and so on; the third category is the sharing platform, which is responsible for information sharing between vertical businesses and the platform. The coordination between vertical business and the platform is achieved through internal open source implementation; the individual needs of vertical businesses and the common needs of the foundation platform can be fully met. offline, the one city-one network model can solve the last mile problem of large appliances. In the e-commerce era, the dual-channel strategy will be the best choice for manufacturers to improve the market competitiveness of enterprises and increase customer demand. The integration of online and offline commerce enhances the ability to penetrate markets and creates a new value transfer mode [16]. The integration of online and offline commerce depends on shortening the experience gap shorten these two sectors. The experience gap will shorten by the improvement of customer experiences and service levels, such as through online bookings, online payments, online counseling, online customization, online complaints, and other service experiences. By shortening the gap between experiences, more consumers will purchase offline products or service online, thus increasing the network effect of the impact of the number of businesses on consumer utility. Proposition 4: The impact of business numbers on consumer utility can be enhanced by minimizing the experience gap between online and offline commerce. 4.4 Mutualism promoted by the ecosystem of e-commerce The characteristics and attributes of biological ecosystems are helpful to understand business ecosystems [17]. Every player has its own task in an e-commerce ecosystem, which are intertwined to form a complete network. After several decades of development, the Alibaba Group has accumulated hundreds of millions of consumers, sellers, and numerous third-party service providers; all these players promote logistics and online payment development. In January 2013, the Alibaba 4 Figure1. Alibaba s ecosystem Alibaba and third-party service providers settled in the platform carry out their duties for e-commerce services, and a complete ecosystem is created. (1) The social ecosystem: Alibaba acquired the preferred shares of Sina Weibo in 2013; this deal, which marks a milestone in the history of e-commerce, extends to online social networks. Sina Weibo has hundreds of millions of users; information can be quickly spread on the Weibo platform, which is a weak-tie social network; a strong-tie social network, named Laiwang, was also launched by Alibaba. (2) The financial ecosystem: Alibaba uses the credit and behavioral data accumulated by B2B, Taobao, Alipay, and other e-commerce platforms, network models, and videos; a credit investigation mode was introduced to confirm the authenticity of customer information. Customer behavior data is transformed into personal credit evaluation, and then those who are usually unable to obtain loans in the traditional banking channels can achieve small loans
online. (3) The logistics ecosystem: Cainiao was jointly established by Alibaba and other logistics companies in 2013; an open social storage facilities network was formed all over China through self-build cooperation reconstruction. In addition, Alibaba become a shareholder of amap.com; it opened up location information and user data in order to promote the development of the O2O business. (4) Information services: Alisoft provides personalized software services for SMEs, based on the SaaS model. Alimama built a platform for small websites and e-commerce for advertising cooperation. Aliyun and Yahoo facilitate the obtaining of information for users; thus, consumers are able to search for thousands of products, and search cost is reduced. Customer-side The number of consumers Network effects of Customer side P1b P3 Business Credit Services accuracy P2 P5a Platform s ecosystem P5b P1a p4 Gap between online and offline Business-side Network effects of Business side The number of businesses Figure 2. The two-sided network effect model of e-commerce Proposition 5: The two-sided network effect can be enhanced by the formation of an e-commerce ecosystem. Proposition 5a: The impact of business numbers on consumer utility can be enhanced by the formation of an e-commerce ecosystem. Proposition 5b: The impact of business numbers on consumer utility can be enhanced by the formation of an e-commerce ecosystem. 5 Discussion and Prospect The formation of an e-commerce ecosystem has impacts on business; enterprises specialize in different areas to offer products that reflect their capabilities. Enterprises need communication and collaboration with other enterprise in the system to improve its value. The propositions developed in the case study emphasize the different stages of e-commerce. The discussion conducted on e-commerce evolution is based on four stages. 1) Introduction stage: New technology, such as mobile commerce and location based services, record consumer behavior in the dimensions of time and space. Marketing accuracy can be improved by data mining during this stage in order to attract businesses and consumers and boost sales. Big data is able to provide sufficient space for personalized business applications, based on individual consumer behavior and preferences data, the future enterprise may provide personalized products and services according to different interests and preferences for each consumer [18]. 2) Growing stage: Unexpected advantage is achieved by technology,,but only the most fit survive. It is hard to avoid the growing stage as mixed participants join the plantfrom. This is also the transition phase of platfrom, when it moves from barbaric growth to brand reputation building. Negetive comments are easier to spread than positive ones. Rigorous review mechanisms are semt to the market to avoid adverse selection problems in the market for lemons. 3) Adjustment stage: With the developmnt of e-commerce, online purchases are acheiced by PC and mobile commerce. Offline purchases become bottlenecks in the development of e-commerce. The traditional market faces changes in the interacion of online and offline commerce. In this suitation, business systems open up. Online and offline system integration become an important issue in order to solve O2O problems. 4) Mature stage: There are contestable assumptions in market theory: freedom in entering, no technogical inferiority of later entrants,and no sunk costs for those who leave. As the barriers to entry in the e-commerce ecosystem are almost zero, many business are involved in the e-commerce ecosystem. The existence of the scope economic effect promotes the opening up of the platform, as, for example, in the open API interface; this encourages third-party developers to create applications and to provide privacy processed data to research institutions and corporations. 5
Introduction stage Growing stage Adjustment stage Mature stage Evolution feature Evolution origin Consumer behavior mining : Dynamic Pricing on time; Location Based Service Business model revolution driven by technology : Big data era; mobile commerce Set up credit business environment : Quality supervision; feedback and exit mechanisms Establish Survival of the fittest driven by market : fake qualifications; difficult in supervise; negative comments spread Easily Improve the service capability from online to offline: Open up the business system; data sharing; Online and offline systems integration Channel Integration driven by value Chain: Traditional stores suffer shock from electricity business; Bottleneck of Offline Experience Adaptive ecosystem : open platform Construction; Provide API interface; Data open for third party applications Diversity driven by scope economies: Long Tail market need varies categories Network utility promote platform to be Evolution direction Attract consumer, improve product selling Retain customers, build reputation of platform Satisfied customers, integrate related resources Mutualism of participants Figure3 revolution of e-commerce model 6 Conclusion E-commerce is a business model with network effects; thus, how to utilize network effects is the important issue in e-commerce development. Based on network effect theory, a case study was conducted on six e-commerce corporations and two-sided network effects in e-commerce were explored. The proposed propositions included the following: the impact of consumer numbers on business utility can be enhanced by improvement of service accuracy in the two dimensions of time and space. The impact of business numbers on consumer utility can be enhanced by the improvement of business credit and by the minimization of the experience gap between online and offline commerce. The two-sided network effect can be enhanced by the formation of an e-commerce ecosystem. The research model was based on the above propositions, and e-commerce model evolution was discussed during the four stages of big data era. With the applications of web2.0 and the emergence of big data, there are many areas still worth exploring and improving in research and practice: redundancy in personalized recommendations and visualization in retrieval; fake credit and credit evaluation mechanisms based on network relationships; the emergence of network behaviors and supply chain coordination between online and offline; herd behavior in networks and coordination of the e-commerce ecosystem. ACKNOWLEDGMENT This work was supported by the Electronic and Information Department of the Chinese Academy of Engineering in 2013 (2013-XX-03) 7 References [1] Lynch C. Big Data: How do your data grow? [J]. Nature, 2008, 455 (7209): 28-29. [2] Hui S K, Fader P S, Bradlow E T. Path data in marketing: An integrative framework and prospectus for model building [J].Marketing Science, 2009, 28(2) : 320-335. [3] Katz M, Shapiro C. Network externalities, competition, and compatibility [J]. The American Economic Review, 1985,75(3): 424-440 6
[4] Corrocher N, Zirulia L. Me and you and everyone we know: An empirical analysis of local network effects in mobile communication [J].Telecommunications Policy, 2009, 33(1): 68-79 [5] Csorba G, Hahn J H. Functional degradation and asymmetric network effects [J].The Journal of Industrial Economics, 2006, 54(2) :253-268 [6] Wright J. The determinants of optimal interchange fees in payment systems[j]. Journal of Industrial Economics, 2004, 52:1-26. [7] Hagiu A. Two- sided platforms: product variety and pricing structures [J]. Journal of Economics & Management Strategy, 2009, 18(4):1011-1143. [8] Weyl E G. A Price theory of multi-sided platforms [J]. American Economic Review, 2009, 100(4):1642-1672. [9] Bolt W, Tieman A F. Heavily skewed pricing in two-sided markets [J].International Journal of Industrial Organization, 2008, 26(5):1250-1255. [10] Yin R K. The case study crisis: Some answers [J].Administrative Science Quarterly, 1981, 26:58-56. [11] Pettigrew A. Longitudinal Field Research on Change: Theory and practice [J].Organization Science, 1990, 1(3):267-292 [12] Eisenhardt K M. Build theories from case study research [J].Academy of Management Review, 1989,14(4):532-550. [13] Marc B. Interview with Hal Varian Chief Economist at Google [J]. Communications & Strategies, 2012, 88(4):127-128 [14] Welteveden J W J. Substitution or complementarity? How the Internet changes city centre shopping [J]. Journal of Retailing and Consumer Services, 2007,14(3): 192-207. [15] Mokhtarian P L. Telecommunications and travel: the case for complementarity [J]. Journal of Industrial Ecology, 2002, 6(2):43-57. [16] Tedeschi B. Compressed Data: Big companies go slowly in devising net strategy [N]. New York Times, 2000, 3-27. [17] Iansiti M, Levien R. Strategy as ecology [J]. Harvard business review, 2004, 3:68-78. [18] Gobble M M. Big Data: The Next Big Thing in Innovation[J]. Research-Technology Management, 2013 :64-66 7