How do you measure network effects in markets?

1 Direct vs indirect network effects

The first distinction to make is between direct and indirect network effects. Direct network effects occur when the utility of a user depends on the number or quality of other users of the same product or service. For example, a phone network becomes more valuable as more people can call each other. Indirect network effects occur when the utility of a user depends on the availability or variety of complementary products or services that are compatible with the main product or service. For example, a gaming console becomes more valuable as more games are developed for it. To measure direct network effects, you can use metrics such as user growth, retention, engagement, or referrals. To measure indirect network effects, you can use metrics such as the number, diversity, or quality of complementary products or services, or the degree of standardization or interoperability.

2 Network externalities vs network internalities

The second distinction to make is between network externalities and network internalities. Network externalities are the benefits or costs that a user imposes on other users by joining or leaving a network. For example, a new user of a social media platform may increase the value of the platform for existing users, but also increase the congestion or noise. Network internalities are the benefits or costs that a user receives from other users by joining or leaving a network. For example, a new user of a ride-hailing platform may enjoy lower prices or faster service, but also face more competition or lower quality. To measure network externalities, you can use methods such as surveys, experiments, or simulations to estimate the willingness to pay, the marginal value, or the social welfare of different users. To measure network internalities, you can use methods such as regression, matching, or instrumental variables to estimate the causal effect, the elasticity, or the marginal cost of different users.

3 Network size vs network structure

The third distinction to make is between network size and network structure. Network size refers to the number of users or nodes in a network. Network structure refers to the pattern of connections or links between the users or nodes in a network. For example, a network can be dense or sparse, homogeneous or heterogeneous, centralized or decentralized, clustered or random. Network size and structure can both affect the value and performance of a network, but in different ways. To measure network size, you can use metrics such as the total number, the average number, or the distribution of users or nodes in a network. To measure network structure, you can use metrics such as the degree, the centrality, the density, or the modularity of users or nodes in a network.

4 Network dynamics vs network stability

The fourth distinction to make is between network dynamics and network stability. Network dynamics refer to the changes or evolution of a network over time. Network stability refer to the resilience or robustness of a network against shocks or disruptions. For example, a network can grow or shrink, adapt or resist, innovate or imitate, cooperate or compete over time. Network dynamics and stability can both affect the sustainability and competitiveness of a network, but in different ways. To measure network dynamics, you can use metrics such as the growth rate, the churn rate, the innovation rate, or the learning rate of a network. To measure network stability, you can use metrics such as the reliability, the redundancy, the diversity, or the security of a network.

5 Network effects vs other factors

The fifth and final distinction to make is between network effects and other factors that may influence the value and performance of a network. Network effects are not the only source of competitive advantage or market power for a firm or a platform. Other factors such as quality, price, reputation, design, or regulation may also affect the demand and supply of a product or service. For example, a network may have strong network effects, but also face high switching costs, low differentiation, or strict regulation. To measure network effects relative to other factors, you can use methods such as conjoint analysis, discrete choice models, or hedonic pricing to estimate the relative importance, the trade-offs, or the preferences of different factors for different users.