Harvard Business School

Assessing the Strength of Network Effects in Social Network Platforms

Network-specific traits, such as the degree of clustering and the prevalence of multihoming, influence the strength and competitive impact of network effects. However, network size alone is often misleading, and network effects should be examined on a case-by-case basis.
By Marco Iansiti, Professor of Business Administration at Harvard Business School
March 21, 2021   /   2 Minute Read
Network Effects in Social Network Platforms

Originally published on March 21, 2021 in Harvard Business School’s “Working Knowledge: Business Research for Business Leaders.” Below is the overview from this paper:

Network effects have risen to the forefront of platform competition discussions (e.g. the House Judiciary investigation of competition in digital markets, claiming that Facebook, for example, is entrenched due to strong network effects and high switching costs). While newer literature has developed much more sophistication in characterizing network effects, common regulatory perspective often assumes more simplistic views.

Older literature tend to simplify the issue of network effects and focus on size as a primary determinant of their strength and impact on competition. The historical characterization of network effects as constituting “winner-take-all” systems is inaccurate. More recent work shows more nuance and considers factors such as network structure and ease of multi-homing which may significantly reduce the strength of network effects. Ultimately, network effects do not necessarily increase in line with network size. For example, network structural traits may weaken overall network effects depending on the degree of clustering on the network. This leaves highly clustered platforms particularly susceptible to competition. Such traits may be specific to a given industry, platform, or even platform feature.

The Facebook network, for example, like many other social networks is characterized by a large number of relatively small and largely separate local clusters. This indicates that network effects may be weaker for Facebook than the sheer size of their user-base may imply. The more tightly clustered a network is, and the more segregated these clusters are, the easier it is for competitors to enter the market with focused solutions. This significantly reduces the likelihood of an individual social network gaining dominant share and adds pressure on incumbents to innovate and compete to retain users.

Moreover, competition in social media is evidenced by the prevalence of multi-homing amongst social network users, the frequency of entry and success of competitors, and the necessity for innovation by incumbents. Multi-homing reduces user reliance on any platform, exposes users to competitors, and suggests broadly that other competitive options exist. Such multi-homing forces constant innovation by incumbents, despite the inevitable costs of such innovation. Ultimately, the entry and success of new competitors such as TikTok and Snapchat serves as compelling evidence of significant competition.

Access the working paper here

Find out why top tech firms, Fortune 500 companies, and global law firms partner with Keystone.