What is network effect?

A network effect exists when every new user adds value for the existing users. It matters because it can make a leading product pull further ahead, and it explains why some markets end up dominated by one or two firms.

How it works

The value of a network grows with the number of connections it can support. With n users, the possible pairs number n times (n minus 1) divided by 2, so value can grow faster than the user count. More users make the product more useful, which attracts more users, a reinforcing loop.

The loop works in both directions. A network that loses users becomes less useful and loses more. The early stage is hard, because few users means little value, so a product often needs to give subsidies or seed one group. Beyond a point, growth slows as the most likely users have joined.

Example

Say a messaging app has 100 users, and each pair can talk, giving 100 times 99 divided by 2, or 4,950 possible connections. When it reaches 200 users, the pairs number 19,900. Doubling the users multiplied the connections by about four. If each connection is worth 0.01 per month to the users, monthly value rises from about 50 to about 199. The numbers are an illustration.

How to model it in Machinations

Create a Pool called Users with a Source of new sign-ups. Make the sign-up rate a Register that rises with the number of users through a state connection, and add a Drain for users who leave at a rate that falls as the network grows. Add a Register for value per user. Play from a small starting Pool and from a larger one, and chart users to see when growth takes off.