Information Cascades, Word of Mouth Marketing, Astroturfing, and Viral Internet Content
Sources: http://news.stanford.edu/news/2014/april/viral-photo-cascade-040314.html, http://www.forbes.com/sites/kimberlywhitler/2014/07/17/why-word-of-mouth-marketing-is-the-most-important-social-media/, http://www.theguardian.com/technology/2014/feb/26/secret-to-viral-video-marketing
What exactly makes an image or video go viral on the Internet? Oftentimes I’ll be on Facebook or sitting with friends and all of a sudden my news feed will show something a friend of mine shared. Oftentimes it’ll be a popular music video or a funny clip no more than 3 seconds long, but they’ll always have thousands of views on them each. Why is it that I end up seeing these things with thousands of views already attached to them? Based on what we’ve learned in the course, information cascades would be the primary culprit. For me, I would not be watching those videos unless they had those views already. I’m sure that for many other people, they would share the same sentiment. How, then, do videos like that become viral in the first place? If a video is watched because they have views, and a video gets views from people watching it, how do we resolve that circular paradox? Obviously there may be some astroturfing involved – the Guardian article details how many businesses and companies are the driving forces in getting certain videos popular, especially those that feature their products. One such (in)famous example was the Dove Real Beauty Sketches (https://www.youtube.com/watch?v=XpaOjMXyJGk). I’m sure that if Dove didn’t create the initial media hype that other people would not have shared that video. As explained in the Guardian article, “[The Dove Real Beauty Sketches] was also backed up by a rock-solid media planning, distribution and public relations strategy.”
What if a video didn’t have such resources devoted to it? What if a video was just simply a video of a bunch of corgi puppies running across the camera (https://www.youtube.com/watch?v=Dsg8JccRZCw)? Surely there’s no possible way for such a small video to gain any such traction, could it? Yet here I just showed that a 6 second clip has over 5 million views, so it must be possible. But how? Researchers at Stanford believe that information cascades are the driving forces behind such phenomena. Although their research involved photos, I believe that viral photos are highly correlated with viral videos in the way they become viral. The researchers believe that it could be possible to predict if a photo will go viral based on information cascades. The main issue, however, was that “It wasn’t clear whether information cascades could be predicted because they happen so rarely.” Even with that kind of statement, scientists (including our very own Jon Kleinberg) were able to accurately predict, 8 times out of 10, when a photo would double in shares. Much of this was based off of observations in the rate and speed with which the photos were shared, and the structure of sharing (where multiple networks created stronger cascades). Therefore, the faster the sharing, the higher likelihood of a photo becoming viral. This idea could just as easily extend to viral videos.
Ultimately, what the main point is from examining information cascades and astroturfing is that word of mouth marketing is one of the most powerful tools any one can utilize. From the Forbes post, “92% of consumers believe recommendations from friends and family over all forms of advertising.” What would be extremely interesting is the kinds of people that these videos and photos are shared with. I’d wager that there is a higher chance that videos and photos are shared among friends and family on social media than through random links on various miscellaneous websites. If anything, I’d be extremely interested in the social ramifications research would have in this field. What would this reveal to us as as humans? What would this mean about our online behaviors now that the Internet is becoming more of a required utility rather than a luxury? Could it really be possible to forecast fame?