Schijndel – Paaspop, The Netherlands

Pop music is a very interesting field of research. The topic is part of most people’s everyday lives and is also economically relevant. Today’s music industry offers a huge volume of digital data which not only tells us about listening preferences but can help to improve public safety during live music festivals. Hannes Datta studies this data and recently gave a talk about his work during a major festival: Paaspop in Schijndel.

The music market has become almost exclusively digital over the past thirty years. The vinyl LP gave way to the CD, which in turn has been overtaken by digital downloads. But it is not only the music itself which makes up the vast body of digital data. There is also a wealth of information about the public’s listening preferences, their purchasing habits, and their behavior during live pop festivals. All these ‘ones and zeroes’ are certainly music to the ears of researcher Hannes Datta, who regularly ‘mines’ the digital data in search of hidden treasure. Online services such as iTunes and Spotify offer all kinds of interesting information about the economics of pop music. Which artists are most popular? How long does it take a new act to achieve international status? Can artists whose success was forged during the vinyl era look forward to a second ‘digital’ career on iTunes or Spotify? The online music services jealously guard their valuable data and it is by no means clear how they themselves use it. Nevertheless, much information is available in the public domain, often free of charge.

The Business models of iTunes and Spotify

Datta frequently trawls the social media sites and supplements his own datasets with information drawn from ratings, charts and the artists’ own figures. He then sets about analyzing the data. “iTunes and Spotify have different business models,” he points out. “iTunes allows the consumer to buy individual tracks, while Spotify is a subscription streaming service. You pay a fixed monthly membership fee and can listen to as many tracks as you want. One of my research questions is whether this affects people’s listening habits. This is certainly relevant information for the music industry, and would enable companies to adapt their services in line with consumer demand.” Datta has found that consumers who pay a flat rate subscription appear to spend twenty per cent more time listening to music, and are more willing to try music with which they are not familiar: there is twelve per cent more variation in their choice of artists. The data requires further in-depth analysis, but Datta is already prepared to state that the subscription model is not in the best interests of the established recording stars. “In essence, they earn less money.”

Pop festivals

Hannes Datta’s research centering on live music festivals represents an even more innovative use of the available data. Most festival organizers have a significant quantity of raw data derived from their visitors and other music enthusiasts, but are not able to put that data to any useful purpose. Big data without research and analysis is absolutely worthless. Hannes Datta has compiled a database of Instagram posts and Twitter messages (‘tweets’) sent by visitors to South by Southwest, a major music festival held in Austin, Texas, each year. There are some 800,000 tweets alone, offering a veritable goldmine of information. They not only reveal what visitors thought about certain bands, but further analysis reveals ‘metadata’ such as who was where at any given moment. When linked to Google maps and weather charts, this information offers new and very interesting insights. Where do people go if bad weather is on the way? How do the crowds move during the course of the day? Can any conclusions be drawn based on factors such as age and gender?

Safety

“The safety of both visitors and artists is clearly a major concern for festival organizers,” Datta says. “We can all recall events which went tragically wrong due to storms or overcrowding. Organizers therefore wish to know how their visitors will move around the site, and where the potential obstacles lie. By linking the Twitter data to GPS information, I can chart exactly where the visitors are at a given moment and how they move around during the course of the day. At one point during last year’s South by Southwest, a major storm broke over Austin. I was able to produce maps showing that visitors ran to various locations to take cover. This information enables the organizers to ensure that shelter will be available in the right places at future events. They know precisely where to erect the tents and marquees, and they know how many people are likely to use them so that they can plan capacity more effectively.”

Trends in group behavior

The music industry is interested in big data for various reasons. Combining several different types of database is likely to bring commercial benefits. For example, Hannes Datta was able to classify the ‘migration patterns’ of South by Southwest visitors by gender. This information is of particular interest to sponsors: it tells them where best to place advertising for brands which specifically target female consumers, and where it will be less effective to do so. Because many festival visitors now wear an electronic wristband, GPS location tracking data reveals precisely when they visit one of the refreshment areas. This information can then be linked to till receipts to reveal who orders what and when, which allows more efficient purchasing and stock control. “It also becomes possible to predict the peaks and troughs in the day’s business so that managers can plan their staffing requirements. It is the overall picture they are interested in, not what drink Fred ordered at 1pm on Sunday. They are eager to see the trends in group behavior.”

Linking data

As an academic, Hannes Datta does not allow himself to be distracted by what may appear to be the most commercially relevant questions. He studies the possibilities of big data from a purely scientific perspective. “If I am working on a causality study, my aim is to demonstrate causality – nothing more and nothing less. I try to identify generic trends and patterns based on the data to hand.” Datta graduated and earned his doctorate in Maastricht, joining the staff of Tilburg University as a researcher and lecturer in 2013. He feels very much at home here. “Tilburg has a long tradition of analyzing large volumes of data. Moreover, our university is among the top ten in the world for marketing research. Big data has an important place at the convergence of these disciplines.” Datta believes that the use of big data in the music industry is still very much in its infancy. He notes that the quality of the data is improving all the time, but the most significant development is likely to be the way that various data sources are interlinked. “There are many smaller companies working to develop interfaces which will enable entirely disparate, incompatible datasets to be linked and compared. This is likely to reveal many unexpected relationships and new insights.”

Source: The University of Tilburg
Credit: Rutger Vahl