The concept of network effect (in general) is by now well understood: a flywheel type situation where a good or service becomes more valuable when more people use it. Many examples out there from the telephone system (the value of a phone increases if everyone has a phone) to Facebook to many marketplaces (with some nuances for the latter).While they produce many of the same benefits, data network effects are more subtle and generally less well understood. Data network effects occur when your product, generally powered by machine learning, becomes smarter as it gets more data from your users.
Source: The Power of Data Network Effects | Matt Turck
I really like how Matt puts this. Here’s my way of thinking about it:
- Network Effects: The more people that use the service, the more useful it is because there are more people to use it with
- Data Network Effects: The more people that use it, the better the service actually becomes
Scenarios
- With a network effect, Jim doesn’t really like using FaceSpace because his friends aren’t on it, and once they start joining he likes it far better
- With a data network effect, Julie doesn’t like using her personal assistant because it doesn’t predict her desires very well, but after the service starts gaining more users it starts predicting everything
In the first case the service didn’t change, it just got more popular. In the second case, the service itself improved.
That’s the key difference.
Notes
- There really should be a different name for “data network effect”, as it’s too easily confused with something around data networks, and too similar to “network effect”. I propose “data learning acceleration” or something like that.