What Matt Damon taught me (unintentionally) about viral social media and neuroscience

I am not a digital native. How do digital natives simultaneously promote their work, abilities, lifestyle and beliefs coherently and effortlessly, across multiple platforms, as if was the most natural thing in the world? Sure, I have a website and Facebook, LinkedIn and Twitter accounts, but they're not exactly the life of the social media party. In September 2015 I decided to wade a little into the social media pond and try it out.

Here's what happened.

Testing the water

I created an account for myself on Quora, because I had noticed it appearing in some media articles. I set some keywords related to things I know about - virtual reality, startups, neuromorphic engineering, Zurich (where I live), Australia (where I was born) - and started answering a few questions. Pretty soon I was answering questions that had nothing to do with work. Then, one day, this question came along (by Ferdison Cayetano):

How much money has been spent attempting to bring Matt Damon back from distant places?

It caught my attention because I thought it was funny. Around the same time, the movie The Martian was playing in cinemas. I decided to answer the question in two ways: adding up the budgets of movies where he gets rescued, and estimating the real costs of saving him in each case. This was a bit tricky due to several of the movies being science fiction, but I came up with some nice big numbers. The total came to just over $900 billion.

After a little while, my answer became a minor hit on Quora. Thousands of people clicked on it. About 1% of the viewers also upvoted it (for my posts it seems to be a fairly consistent pattern that about 0.5-1.5% of people upvote the popular ones).

You can see from the graph that there was a strong initial response, followed by an elevated period lasting about a week, before returning to a baseline of about 500 views per day. I was pretty happy with this. It was small compared with some of the other Quora posts which had hundreds of thousands of views, but I had managed to generate a minor social media response. Thousands of Quora people had read my stuff, cool! Some people even left comments, including some useful feedback that I used to edit the post slightly.

Going viral

I didn't think much about my Matt Damon post after that. I kept occasionally answering Quora questions about all sorts of topics, just for fun. Then, three months later, Julie Ann Exter from Quora's media publishing team contacted me. They wanted my approval for my post to be featured on some tech website. Sure, I said. Why not? It would be fun to be "published" on a media website.

Fairly rapidly, things started growing ahead of what I could understand. A few days later, I received a message from Quora user Carla Attenborough:

I opened Facebook as I sit here on my parent's sofa 10 minutes ago and trending #2 in ALL of Facebook is your response to the question of how much money has been spent in movies to "save" Matt Damon.

I didn't even know that trending topics existed on Facebook (I think it's not available in Switzerland). I went to check and saw this:

OK, now that was strange. My name is not exactly common, so all of those hits were probably about me. I wasn't sure what to do about it.

Propagation; searching for the source

I had no strategy for dealing with what came next. Shortly afterwards, I was contacted by journalists from BuzzFeed and the New York Post. They wanted to know if I worked for NASA or something super-cool like that. Unfortunately I had to disappoint them, but I do work in technology research, which was enough to provide a tiny shred of credibility for the numbers. National radio broadcasters in two countries also wanted to interview me, as long as I responded immediately. I declined the interviews - what on earth was I going to say?

Things bubbled along over the following hours and days. Quora had managed to get about ten media sites to take the story - Time, Mashable, USA Today, Yahoo, etc. This led to a serious bump in the Quora traffic, or so it seemed:

Note that the second peak was about twice the height of the first peak. However, it was also about half the width (about 5 days), and it was also more symmetrical. The total number of views was roughly the same in both peaks. This meant that, although millions of people now knew about the post, only a tiny percentage clicked through to the original source.

A review of the Google search revealed that, in a very short period of time, hundreds of other sites had posted replicas of the post. There were a few translations into other languages too, in particular Spanish and German. What caused the media surge that led to the second peak? Is there an all-powerful bot that decides what "news" we are all going to see? Or are all of those content editors looking at each other, individually yet collectively deciding the same thing, waiting for the first to jump? Are we all just sheep with emergent conformity, or are we being subtly controlled by the System?

Cascading effects

Another thing I looked at was second-order clicks. Below are the clicks on an answer that I wrote to a different question on Quora. The first peak was at the time of my original post; the second peak was around the time of the Matt Damon post going viral. Again, the second peak was roughly twice as high as the first, and the second peak is also more symmetrical than the first.

Next I checked my own website. I only had data for the second peak, but again it was quite symmetrical. A very tiny fraction of readers (maybe 200) had gone looking specifically for me.

The next graph shows the relative frequency of Google searches for the phrase "Matt Damon". An interesting factoid: people generally search for more Matt Damon on weekends. This pattern was broken when my post went viral; the Matt Damon searches stayed at an elevated level. The peak at the very end was probably due to the announcement of the next film in the Bourne series. So Google saw the effect of my post in two ways: in its indexing of websites carrying the story, and in the search patterns of users.

Feedback and interaction

Needless to say, in additional to these graphs, there was an avalanche of text comments and feedback all over the interwebs. I deliberately ignored it, reading only a few comments. The little that I did see was enough to infer that the same questions were being asked over and over again. I asked two colleagues active in media and communications for advice - Heather Berlin (also a neuroscientist; a bit more about neuroscience later) and Daphne Spyropoulos. Their tips included two key points:

• You posted it, you own it. Don't back down or walk away, or you lose credibility.
• This type of wave doesn't come very often. Enjoy the ride, see where it goes.

Finally, when I was told that Ashton Kutcher had asked about it on Facebook, and I found out that he has 17 million followers, I decided that I had to write something. I think it's called "getting ahead of the story" before it goes off in an unexpected direction. Six days after the original post went viral, I wrote a post on Medium (another new site for me), explaining how I had calculated the largest numbers, plus a little bit about my qualifications. I then linked to it from my original post and my Facebook feed.

Again, there was a similar response. A small number of people clicked through and read the new article. This time the curve was less symmetrical, with a slower fall-off. Perhaps it was because this time I was tweaking things and posting the link in a few extra places as I went.

At this point I couldn't really think of anything else to do or measure. I had gained a small first-hand taste of the ethereal beast that is the collective social media consciousness, and escaped with no significant damage. I went back to posting random stuff on Quora as a hobby.

Analysis and analogy

Why did I look at all of the graphs? It turns out that these social media graphs are very familiar to people working in neuroscience, particularly with neural models and recordings. I'm part-time on the faculty of the Institute of Neuroinformatics at the University of Zurich and ETH Zurich. Here, I get to see a lot of smart people worry about things including the analysis of "spike trains" (sequences of neuron firing patterns) using various methods, including histograms which sometimes look a bit like the ones shown above. One of the key questions in systems neuroscience is how and under what conditions neural activity propogates from within a local cluster (microcircuit) of neurons to another microcircuit in another part of the brain. The data from various systems seems to suggest that the links between clusters are weak (a few %), yet highly specific. This vaguely echoes the results I found in the social networks, where a very small fraction of users propagated from one site to another. Yet, somehow, all of these small and seemingly random interactions caused my particular post to cross boundaries of web platforms, countries and languages to a brief period when it had a global audience. So if we can understand how it works in brains, then we might also know more about how to understand and shape social media.

So what did I learn from all of this, apart from how to use Quora and Medium?

• If you say anything in social media, you become the designated expert - so it's best to make sure that you really are the expert.
• Social media management is hard, time-consuming, around-the-clock work.
• Timely response in social media is essential. The caravan moves on very quickly.

So will this experience turn me into an active social media person? Probably not. But it was fun and instructive to try it out anyway.

Epilogue

There was still one more breath of life in the story. Ten days after the story went viral, Matt Damon was interviewed on TV (he was nominated for a Golden Globe award), and for a bit of fun the interviewer asked him in detail about the post. I couldn't believe it. A succession of improbably random events, starting from a question that I had answered as a random joke, had led to the real-life target of the question. It was all a bit surreal.

Disclaimer

I have zero track record in social media research; my results above are those of an interested non-expert observer.