On Lehr (2021):
Network effects
The author gives a nicely packaged definition of network effects. Benefiting from network effects means that the system gets more valuable as it amasses something.
- Direct network effects: value comes from users
- Indirect network effects: value comes from groups of users attracting other groups of users. I.e.: developers bring more users (bc apps), and more users bring more developers (bc more money)
- Data network effects: value comes from the data generated by users
User costs
When thinking about retention, it's useful to think about:
- Switching costs: how expensive it is for users to drop the system
- Multihoming costs: how expensive it is for users to use multiple systems at the same time
Thinking in layers
The author proposes that company value comes from how deep it is within the layers stack (as opposed to thinking value comes directly from Network effects).
Using Apple as an example:
- its walled garden increases switching costs (User costs)
- it sells hardware, so it's deep: it owns the layer over which OS and software run
- it occupies as much real estate as it can (AKA the home screen), to avoid multihoming
The deeper we get into the stack, the stronger the grasp. The author uses the idea of 'truth layers', in which there is no multihoming (identity is a truth layer). So systems should aim to embed themselves there.