Reverse Network Effects
This article talks about the “Reverse Networks Effects”. What this term essentially means is that we normally would think that for social media networks, when the user base grow larger and larger, the social media networks become more and more stable; however, sometimes it is possible that even under an enormous user base, social media can suddenly shrink and lose their users rapidly. Examples include “ChatRoulette”, “Orkut”, etc. This seems unintuitive, since according to our model of network effects, the more users social networks have, the more probable that for a specific person, the percentage of his friends that uses that particular social media is high. When this percentage is high, that person is also more likely to use the social media because the percentage exceeds the threshold q, if that exists. Therefore a very large user base for a social media seems to be always be able to cause further increase on the user base.
The article, however, tells us that this is not always the case. In fact, the term “reverse networks effects” is not accurate. Social media with an enormous user base suddenly losing a lot of users is itself a networks effect. Think about the examples we talked about in class: when all people are using product B, how can we stimulate everyone to start using product A. This is essentially the same situation described in the article: when a lot of people are using a social network product, how could people suddenly stop using it? Remember that if many people in a concentrated group suddenly stop to use a product, this could cause the percentage for many individuals to drop, and thus creating a chain effect where a lot of people suddenly stop using the product. That is, if we can find certain explanations for why some people would suddenly stop using the product initially when there is a large user base already, we could explain this kind of “reverse networks effect”.
The article gives us three explanations to why this would happen. The first reason is based on “connection”. A lot of social media relies on connection between people, like Facebook or Snapchat. However, when too many users join the network, there would be unavoidable fake information or fraud, causing users to leave the social media. For example, for dating sites, if a lot of people use fake information and contact others basically to harass and disturb, it is likely that a lot of users suddenly decide to drop their account, and causing a chain effect. Another reason is based on “content”. This one is straightforward: if social media have a large amount of users without any kind of censorship, it is likely that there would be inappropriate content that drives people away – imagine if anyone can post any videos on Youtube. The last reason is based on “clout”. A lot of social media let users build up authority, like Twitter, where you can have many followers. This kind of model becomes problematic on a large user base because it is hard for new users to get followers, and thus they might build up an incentive to quit, causing a “reverse” networks effect.
I think the core of this “reverse networks effect” is to remind us that network effects are not about the current user base, but about the relationship between users and the threshold percentage. No matter how many people are currently using product B, as long as we could trigger a chain effect, everyone could switch to A immediately.