Over the last five years, I’ve had the chance to speak to a lot of executives, managers, and artists in the music industry and other creative industries around the world about big data. I’ve met folk in the UK, Europe and the US. I’ve asked them about their needs, their use and their understanding of big data. I’ve asked them about how much they value it.
Big data arises out of two key, new phenomena. Firstly, there is a vast amount of data created in every online interaction, as fans navigate the web, engage in social media and express opinions. In aggregate, the internet generates more data in a day than was previously created in a decade. Some people call this, the exhaust fumes of the internet. Secondly, such a vast quantity of data requires immense computing power to be able to analyse it. Only in the last five years, has hardware become cheap enough and the software smart enough to be able to start to make sense of all this data.
But when you go speak to sales, marketing and promo folk in the midst of it all, they tend to say, as one executive commented to me recently: “I’ve got way too much data. I’m drowning in f***ing data. I can make the data say almost anything I want. Please don’t give me anymore, I beg you!”
In the relatively recent past, Monday morning meetings at record labels were simple affairs, happy to juggle Chart position, Sales figures, YouTube plays and Facebook likes. Pretty much anything else was seen as irrelevant. But with the advent of mobile services, streaming services, and the flood of social media the trendscape has been transformed. The tools for promotion of music online are now so diverse and varied; it is baffling for even the most sophisticated marketer. The fact is that for any one recording artist online, in any one day, there is data to be collected about their fans on Youtube (views, comments, channel subscriptions), Facebook (likes), Twitter (mentions, followers, tweets and retweets), Meerkat, Periscope, Vine, Vessel, Vevo, Tumblr, Instagram, Wikipedia, etc, etc ad nauseum. We can find data about gender, demographics, geography, platform preferences, phone type, time of day behaviours, habits, anomalies and more. These together can easily add up to a thousand data points per artist per day – and then there is the real time analysis of data feeds which is just beginning to be possible!
We become seasick with all this data and how it might be interpreted, our heads swim, our hearts sink. Typically social data companies serve up all kinds of attractively coloured graphs and charts that attempt to provide some kind of insight through analysis of the data. Self-appointed data-scientists (the new alchemists of the mobile online world) hypothesise possible alignments between this social metric and that sales index. This is a rapidly developing field. The speed of development and sophistication is vertiginous. Large companies have been investing quite heavily in this space for some time, building in house platforms and portals to combine analysis of internal with external data, to supply it to in-house staff and external partners. Indies have been mostly figuring out how to ignore it.
Interestingly, though, the real direction of travel may not be how to manage the vast oceans of data, but instead it might just be the ability to strip out most of the data. What will turn out to be most useful is to be able to use the volume of information to zero in on one or two key indicators. More valuable than understanding graph after graph depicting trends, spikes and possible causes, is to find the individual data point that is a key to the fortunes of the act.
The question that no one agrees on yet, is which are the key data points that are most meaningful across all acts. In this fragmented world of overlapping communities and competing platforms, some will still argue that there can be no silver bullet. The next evolution of data analysis for the music industry will see a growing tension between industry standards and proprietary “secret sauce” solutions. Can the industry create new data standards, beyond the current chart based ones for analyzing trends, futures for bands or will each label continue to compete in nurturing its own private ways of seeing into the future?
Jeremy Silver is a director at Mediaclarity Digital, a London-based, global consulting and advisory firm, specialising in music and digital media. He is also Chairman of Music Glue. He will speak at a session called “Big data sets the pace of music”, at Midem 2015: more info here.
& follow Jeremy on Twitter @Jeremys1 or on LinkedIn here!