In 1985, Alivin Roth proposed a new concept “two–sided matching market”, aroused by the case in labor market (Roth, 1985).

Within the studies about markets with network externalities, two-sided markets (in a sense) were just considered to be markets characterized by a special type of network externalities. These externality do not depend on consumption of end users in the same group (for example, consumers of the same product), but on consumption of different, but “compatible”, end-user on an opposite market side.

Jullien (2004) argued that, by reducing the gains from interaction, the total price level affects participation. Increasing prices would mean reducing participation of both sides of the market. From this point of view, externality is then difficult to distinguish between one-sided or two-sided usage.

Wright (2004) described two-sided markets as two distinct types of users, each of whom obtains value from interacting with users of the opposite side over a common platform. Armstrong (2004) defined markets involving two groups of agents who interact via “platforms,” where one group’s benefit from joining a platform depends on the size of the other group that joins the platform as two-sided markets. Rochet and Tirole (Rochet & Tirole, 2004) gave a rigorous definition for two-sided markets under the condition that costs may not be passed trough from seller to buyer. Definition of two-sided markets from Rochet and Tirole (2004) was: there is a platform which charges usage fee aB and a S per-transaction from buyer side and seller side. When the transaction volume V varies with a B when a is constant, this market is a two-sided market. where a = a B + aS and a is the aggregate price level from the two sides. Rochet and Tirole (2004) definition is linked to the platform pricing structure.This definition reveals that platform can affect the volume of transaction by changing its pricing structure. This definition is the canonical and widely accepted one, despite the fact that it does not include platform charging membership fees (or fixed fees) or two-part tariffs from two sides and it is not a complete definition. Generally the buyer side is taken as the side 1 and seller side as the side 2 in two sided markets. For my thesis, I will apply the definition of Rochet and Tirole (Rochet & Tirole, 2004). A market is two-sided if the platform can affect the volume of transaction through its asymmetric price structure to the two sides by an equal amount of the total price level. Platform has to get both the two sides on it.  App store platform is designed to get both developers and users on board in mobile app market.

Network externalities in two-sided markets mainly mean the cross-side effects or indirect network externalities between two sides. Sellers benefit from the number and product or service usage of buyer side. Buyers benefit from the number and product or service supply of seller side. Both sellers and buyers benefit from the interaction between the two sides. Parker and Van Alstyne (2000 , 2002 ) classified cross market externalities and inter-network externalities into indirect network externalities in two sided markets.

Indirect network externalities are classified into membership externality and usage externality (Rochet and Tirole 2004). Membership externality means the effects of one end-user being associated in membership within one side to end-users from the opposite side. Membership decisions generate membership externality. Usage externality means effects of one end-user being interacted with another opposite side end-user to end-users from opposite side. Usage externality arises from usage decisions. If I benefit from downloading an app from Apple App store, then the app developer exerts a (positive) usage externality by supplying more apps to Apple App store. Platform usually charges fixed user-specific and paid ex-ante fees as membership fee. Usage fee is usually a variable interaction-specific and paid ex-post fees.

There also exists the inter-side effects or direct network externality in two-sided markets. Direct network externality indicates the increased benefit to one user with the increase of the number of other users who use the same products or compatible products like standard telecomm network or razors and blades markets (Jean Tirole, 1988).

Platform’s pricing structure is non-neutral and the mark-up for sellers cannot be passed to buyers. The price allocation between the two sides has impacts on participation of two sides on platform and transaction volumes. Platforms in two sided markets compete both to facilitate the transactions and to get more participation of end users. End users in two-sided markets are both users and “input”. Each end user’s participation can create value for others. Indirect network externality, multi-product pricing, non-neutral pricing structure and transaction volume affected by pricing structure are the distinct properties of two-sided markets.

There are plenty of two-sided markets in the real world today, Such as Application stores for iPhone and iPad; Cloud Computing platform; Mobile payment system, credit card platform; Video game platform; computer system software; portals; dating club; TV network and so on. We can find two-sided markets easily in Internet industry; computer industry and payment card system.

Before any formal study of two-sided markets, there had been some papers which had addressed specific issues of some two-sided markets. Markets with network externalities first attracted the attention of economists. Markets with network externalities have been widely analyzed, especially since the contributions by David (David, 1985), Katz and Shapiro (Katz & Shapiro 1985), Farrell and Saloner (Farrell & Saloner 1985), and others .

And then around the year 2000, the debates which were triggered by a series of antitrust cases against some international credit card networks (Visa, MasterCard) had pushed the practice of setting an interchange fee by cooperative credit card networks. Katz (Katz,2001), Rochet and Tirole (Rochet & Tirole ,2002), Schmalensee (Schmalensee ,2002), Wright (Wright,2004) (2003a), (2003b), Gans and King (Gans and King,2003),all these authors agree that credit card services have special characteristics, and that the conventional practices of antitrust policy were not totally applicable to this industry. And then the same characteristics were noticed in other markets, such as in media industries (Ferrando et al., 2004; Kaiser and knight, 2004; Reisinger, 2004) or electronic intermediaries (Caillaud and Julien, 2003; Julien 2004).Progressively, a general theory of two-sided markets emerged .

There was a surge of interest in two-sided markets with the appearances of papers by Armstrong (2004), Caillaud and Jullien (2003), and Rochet and Tirole (2003a). We can call Rochet and Tirole the founders of two-sided market theory. They had not just given a reasonable definition and structure of two-sided market but also introduced membership fee and usage fee to solve the pricing problem in this market. There were also researchers who considered series of factors which dominate the pricing in two-sided market such as elasticity of demand with respect to the total price; network externality; single and multi homing; product variety needs of consumers, etc. some other papers provided general introduction and lessons to be drawn (Evans, 2003) or general theoretical and framework (Rochet and Tirole, 2004).

Two-sided market pricing study has caught people’s attention again when the totally new business model of Apple App store introduced in 2008. Mobile app market has been paid more and more attention as time goes on and resulting in a dynamic increase of application stores. This has greatly changed the industrial pattern in telecommunication industry, traditional E-commerce industry, portable intelligent device manufacturing industry and other areas.