Endogenous Multihoming and Network Effects: PlayStation, Xbox, or Both?
Management Science
Received: February 26, 2022
Accepted: April 19, 2024
Published Online: July 2, 2024
01
Research Team
Øystein Foros
Norwegian School of Economics, 5045 Bergen, Norway
Hans Jarle Kind
Norwegian School of Economics, 5045 Bergen, Norway
Frank Stähler
University of Tübingen, 72074 Tübingen, Germany
02
Research Background
If a product has positive network effects, the utility that users derive from it increases with the number of other consumers using the same or compatible products. Typical examples include the telecommunications industry and the video game console market, where users benefit directly from interacting with each other. Network effects can also arise from the data generated by users during product use, which can be leveraged by machine learning algorithms to enhance the perceived quality for other users. Recommendation systems are a prime example, widely found in streaming services like Netflix, HBO, Spotify, Apple Music, and YouTube. Building an efficient recommendation system relies on a large amount of user data, but once established, it can demonstrate extremely high efficiency. For instance, in 2017, approximately 70% of the content viewed on YouTube was driven by its recommendation algorithm. Additionally, Amazon introduced consumer review features as early as 1995, which became a significant characteristic of its platform.
A substantial body of literature on the interaction between network effects and pricing decisions presumes that each consumer purchases at most one network product, meaning all consumers are single-homing. Under this assumption, a key conclusion is that if different companies offer incompatible products, network effects may intensify competition, thereby driving down prices and profits.
However, in reality, in many product and service markets, at least some consumers choose to multi-home. This paper finds that in such cases, the impact of network effects on prices and profits may be precisely the opposite.
The development of digitalization has made it easier for consumers in the media and entertainment industry to multi-home. When purchasing multiple products brings additional value, consumers are often willing to pay for it. For example, to watch both “Game of Thrones” (HBO) and “The Crown” (Netflix) simultaneously, one needs to subscribe to both platforms. In 2020, 20% of Netflix users in the U.S. also subscribed to HBO Max, while 80% of HBO Max users also used Netflix. In the gaming industry, multi-homing is also quite common. At the household level, nearly half of Xbox One (Microsoft) users also own a PlayStation 4 (Sony). Microsoft and Sony offer exclusive games on their respective consoles, enhancing the marginal value of purchasing a second console. At the same time, network effects further strengthen the appeal of multi-homing. The gaming community has long recognized the value of multi-homing. For instance, in a post on Reddit.com, the author mentioned: “I own both an Xbox One X and a PS4 Pro. The biggest benefit is being able to play exclusive games on any platform. As for the downsides, aside from the initial purchase cost, there are almost none. Plus, it’s great to be able to play games with friends who chose different consoles.”
03
Main Content
This paper studies the competition between two platforms that offer incompatible network products, each with platform-specific network effects. Consumers can choose to purchase only one (single-homing) or both (multi-homing), depending on which choice provides higher net utility. This paper introduces the setting of “endogenous multi-homing” and finds that it has unexpected effects on market equilibrium and platform pricing strategies. In contrast, existing literature generally assumes that all consumers are single-homing, which may lead to misleading conclusions in policy-making and management practices.
This paper constructs a two-stage game model: platforms first announce product prices, and then consumers make purchasing decisions based on this information. In the presence of network effects, the more users a product has, the more attractive it becomes to other potential users. In the base model, this paper assumes that consumers have “responsive expectations,” meaning they use platform pricing as a cue to infer the future user base of that platform. Under otherwise identical conditions, a platform with a lower price is generally perceived as more likely to attract more users. Accordingly, this paper explores the impact of network effects on platform competition, pricing, and profits when allowing for consumer multi-homing. Additionally, this paper extends the base model in three ways: (1) endogenous decisions on exclusive content, (2) the introduction of product compatibility; (3) consumers having “passive expectations” (i.e., they do not infer user numbers based on price).
04
Main Conclusions
1. The relationship between platform pricing and profits exhibits an “inverted U-shape” concerning the strength of network effects. If some consumers choose to multi-home, moderate network effects can enhance prices and profits. For example, after introducing recommendation systems, HBO and Netflix improved user experience, leading to increased demand. When network effects exceed a certain threshold, users become more focused on joining the “big network” rather than choosing their preferred product. This tendency intensifies competitive pressure between platforms, resulting in decreased profits.
2. The number of exclusive features offered by each platform is crucial for market equilibrium. A single-homing equilibrium can only occur when exclusive content is scarce; when exclusive content is abundant, the marginal value for users increases, leading to a “multi-homing equilibrium.” When exclusive content is plentiful and network effects are strong, multi-homing is not only the only equilibrium but also results in higher platform profits than in single-homing scenarios. Conversely, if exclusive content is insufficient, multi-homing may lead to decreased platform profits.
3. Existing literature that allows for multi-homing predicts that platform pricing is independent in a multi-homing equilibrium. However, this paper finds that in multi-homing scenarios, platform pricing is interdependent and exhibits strategic complementarity, even when network effects are weak.
Full Text Link:
Øystein Foros; , Hans Jarle Kind; , Frank Stähler (2024) Endogenous Multihoming and Network Effects: PlayStation, Xbox, or Both?. Management Science 71(4):2993-3012.
https://doi.org/10.1287/mnsc.2022.00613
Recommended by:Hu Yulin, Xue Weili
Affiliation:Zhejiang University School of Management, Southeast University School of Economics and Management
Unauthorized reproduction is prohibited; for inquiries, please leave a message in the ISCOM WeChat public account.
Editor: Wang Yusheng
Reviewer: Li Yongjian
Supply Chain and Operations Management Division
United | Strive | Achieve