Will Your Tweet Go Viral?
This article analyzes the sharing and spread of information on Twitter. Specifically, it uses network theory to estimate the likelihood that a tweet will go viral. Tweets that are shared rapidly and widely, so called “social contagions” carry an extreme amount of influence. As a result, predicting their spread is a valuable tool in understanding societal behaviors and opinions. Authors explained that models of contagious disease have been theoretically used to describe the spread of information, but real data is rarely incorporated into these claims. The article summarizes a data-based study from Beihang University in China, which analyzes real data – including 12 million tweets and 1.5 million retweets. In examining how infectious tweets become infectious, they found that by analyzing the network structure of the first 50 tweets was enough to indicate if a tweet would go viral. The model articulated in the article aimed to predict “retweet cascades.” They also expanded the model to indicate the decline of a tweet’s popularity after it becomes older.
The boom-or-bust concept that was taught in lecture directly applies here. On Twitter, once one person retweets a tweet, it is shared to their friends, and because the first friend liked it, it is likely that people in the second group of friends also will. This provides exposure of the tweet to new audiences (a new group of followers) where there will likely be more people who will also like the tweet, and share it with their followers. The boom-or-bust concept occurs for both direct-benefit and information based reasons. The concept of following someone on Twitter already assumes there is an information-based benefit to knowing what they like. When someone retweets something, it conveys information that will be helpful in determining if you will like it and if you should share it with your own followers. If they found it worth sharing, there is some indication that you should watch/read it also, and that your followers may enjoy it as well. Additionally, there are direct-benefit reasons for retweeting something that other people have also retweeted, as retweeting something popular may contribute to increased status on Twitter. Even more, even if the content is not worth while to you specifically, many find it valuable to still be involved in viral content, just because they are a part of the Twitter community. The article indicates that by the network spread of the first 50 retweets, the model can determine if the tweet will boom or bust. If it does not gain the necessary momentum, in both timing and network dynamics, with these retweets, it is likely to bust.