Thanks to internet-speak, Viral is a word we associate with photos or videos and now there’s a new tool which can predict if your Facebook post will become immensely popular or not. A bunch of Stanford computer researchers have created an algorithm which can measure the chances of a photo going viral on Facebook based on how shares play out.
Apparently, only 1 in 20 pictures posted on the social network in question, manages to get a single share. 1 in 4,000 photos is reported to get more than 500 people sharing it. This is based on data provided by Facebook to university scientists.
How the system works:
At the upcoming International World Wide Web Conference, the Standford team will elaborate on how the algorithm works. For now, we know that the team was able to predict with 80% accuracy as to when a photo cascade could double in shares. That is, they were able to guess whether an image which enjoyed 10 shares would get 20 in the future, and if a picture shared 500 times would hit 1000 shares and so on.
150,000 Facebook photos which were shared at least 5 times were studied in order to formulate the algorithm. At any point that a snapshot goes through the sharing process, there was observed to be a 50-50 possibility of it earning double the number of shares. The rate and speed of sharing, the way pictures are shared (including reposts in various networks) and other factors were entered into the analysis.
Doubling of shares can be predicted best if images are shared many times. The speed of sharing is thought to be the next best indicator of whether the photo will go viral or not. Lastly, the more networks pictures are shared over, the better its chances of becoming hugely popular. Here’s the full report for those of you who want all the minute details.