If, like me, you work in the technical end of the professional spectrum, for the past two years all you will have been hearing about is the advent of big data and how it can be used to change our lives for the better. Even if you have no idea what big data exactly is, the countless news articles and opinion pieces on the subject that were forthcoming once Edward Snowden revealed to the world that each individual has personal data, and (if you have ever had electronic contact with someone in the US) that this ‘metadata’ flows directly to the NSA definitely gives you an idea of the concept. So basically we as a society are big data, and through the maintenance of online profiles and social media we contribute towards the biggest dataset in history, which unfortunately for us consists mostly of #neknomination videos, filtered pictures of food, and declarations of love for One Direction. I doubt even the processing power of quantum computing will help us even broach the sociological implications of how people live their online lives, but of course that job belongs to the NSA: they wanted it, let them organise it. I have followed studies on online behaviour closely for the past decade, and have never really been impressed with any results that have been published. That is, until this week when I saw that PornHub released an interactive graph that showed the responsiveness of online porn traffic to major global events.
I tried vigorously to embed (giggity) the full interactive graph here in this post, but unfortunately this is not possible and you will have to click here and visit the graph on the PornHub blog if you want to play around with it yourself (ooh sir!), or maybe you will be fully satisfied (I will stop soon, I promise) with my discussion of the results here and not wish to seek any more knowledge on the secret lives of others. Anyway, I would highly recommend checking out the full thing for yourself, the results are fascinating. Firstly, by major global events, I mean cultural benchmarks that bring people together en masse, and in this situation specifically they target sports events, holidays, news events, and American TV schedule lynchpins. If you are disappointed with that definition of major global events, I am sure there is a less depressing world out there for you somewhere, but in the meantime I will move on to the data.
Starting in South America, you can see below the graph showing PornHub traffic during the Copa America final in Summer 2011 (it says 2012, but the PornHub guys are American so don’t know about football), where Uruguay played against Paraguay. Overall in South America, there was a 6% drop in porn consumption, while the figures from the individual countries are much greater. Paraguay showed a drop in 20%, while Uruguay reduced their porn needs by just 12%. Uruguay were heavy favourites for the game and indeed went on to win comfortably, which might explain why the Uruguayans went back to their normal schedule sooner than their optimistic, and ultimately decimated neighbours.
This is to be contrasted drastically to the following years European Cup final, where Italy played Spain. If you look at the individual countries, it is obvious a fair amount of people were watching football instead of going about their usual PornHub routine. What is interesting first is that the drop in porn consumption is much greater during important football games in European countries involved in finals than in their South American counterparts, which might have implications about how comfortable Uruguayans are watching sweaty men on one screen, and sweaty men and women on another simultaneously. What is more interesting about the European data is that overall porn traffic is up 7%, which means that not only are the other European nations (who didn’t reach the final) watching more porn than usual during the highlight of the football season, but that they are watching so much that it actually compensates for the loss in a third of the Spanish/Italian market. Who says European integration is dead?
There are so much more priceless nuggets of information contained in that infographic, so much that I would go so far as to name it the best infographic of all time. The release of the first iPad coincided with an upturn in porn traffic of 9%. The Christmas period sees a big downturn for PornHub, except in Japan, where Christmas Day sees a spike. Environmentalists, beware, as Earth Day (where people should use electricity and therefore internet-enabled devices sparingly), sees no change in porn consumption whatsoever. I said 18 months ago that Felix Baumgartners space dive was a pile of wank, and the data confirms it: +4% worldwide. And finally, nothing kills a good porn buzz like the Jewish High Holidays and New Years Eve, which each inflict a traffic loss of over 40% in assorted territories. Check out the full graph, there is something there that will surprise you, I promise.
What is important about all of this data and the subsequent findings is that it is actually trustworthy. PornHub is a virtual platform and therefore operates within a two-sided market, which means it earns revenue through gaining user hits that contribute to its advertising revenue (I assume: I am not familiar with such base sites, however I did write my Master’s Thesis on the theory behind how such a business would operate, which I blogged about incessently here). PornHub gets paid for each set of eyeballs that view their site, and therefore they need to be very aware of how many eyeballs they possess at any given time, as this data is needed in order to set prices for their advertisers. PornHub will charge advertisers more for an ad running during a potential apocalypse (see data in the chart about the Mayan Apocalypse in 2012), and lower for an ad on New Year’s Eve. Although I assume the New Years Eve crowd would be a much more niche group and therefore there would definitely be some targeted marketing opportunities. Anyway, the point is that no matter the source, the data itself is pure, and therefore the results are open for analysis.
The reason the results from this dataset are so interesting for me is that they describe deviance from a normal schedule (yes, in this case deviance means “abstaining from watching porn”). It therefore tells a story that direct actions or statements from the general consumer would never capture: let’s face it, no one would admit to turning off PornHub just to watch news coverage of Osama Bin Ladens’ death. The PornHub traffic data captures revealed preferences, rather than stated preferences, which is the economic theorists dream (yep, I have blogged about this subject before, here). In economic theory, what people reveal willingly about themselves is not as important as what their unconscious behaviour reveals, and this information is hard to come by (yep, a final sex joke). The PornHub data reveals that there are instances and events in our shared cultural history that cause a marked deviation from normal behaviour, and this is interesting on many levels. This years’ World Cup in Brazil will prove the benchmark in such a data analysis, as it will involve viewers from all over the world as spectators, but with an emphasis on the two countries competing in the relevant game. PornHub will release the data on this, and I actually do think that more than a few social scientists should be interested in using it for their research. I will leave you with this other excellent graph from PornHub, which shows the online porn-watching exploits of fans resident in the areas home to the two teams involved in last weeks Superbowl: the Denver Broncos and the Seattle SeaHawks. The SeaHawks had won the game by halftime: see how fast the Denver Broncos fans got over the loss and got back to their normal ways?