The trick with a network effect + stored value model is solving the "chicken and egg" problem. You have to make sure your service is valuable to a scalable base of users on Day 1, and that the value proposition to users doesn't depend on there being an enormous base of users (network) or a wealth of stored value (content, data, etc.).
Often, this means starting with a niche. Ideally, a niche that is willing to invest a lot of time and attention up front, thereby creating value for the more casual and follow-on users. Wikipedia is a good example. The idea of using an encyclopedia that is completely devoid of content probably wasn't very appealing to the casual masses in the early to mid 90s. So Wikipedia relied on a smaller mass of users who got a kick out of cataloguing, writing, and organizing information. These users tilled the soil, so to speak, making it more fertile for a mass userbase.
Often, this means starting with a niche. Ideally, a niche that is willing to invest a lot of time and attention up front, thereby creating value for the more casual and follow-on users. Wikipedia is a good example. The idea of using an encyclopedia that is completely devoid of content probably wasn't very appealing to the casual masses in the early to mid 90s. So Wikipedia relied on a smaller mass of users who got a kick out of cataloguing, writing, and organizing information. These users tilled the soil, so to speak, making it more fertile for a mass userbase.