Introduction
The word âdata,â in my opinion, has some challenges in the context of the cultural and creative industries. It is such a generic term, covering any type of information, and it can be perceived as rather dry and extensive, almost the opposite of creativity. Yet those with the right skills can bring these data to life for people and find stories within it that have an interest or practical meaning for others. I have been engaged in ongoing exploration into the challenges, trends, and opportunities of big data in order to put audiences at the heart of our cultural institutions. This exploration has taken the form of technical experimentation, action research, and ethnography, in particular the following:
Aggregating and analyzing data for audience development
Rapid-prototyping through collaboration between tech and culture specialists
Ethnography into the cultural and structural barriers to data driven decision-making
When given the right tools, big data can help organizations build resilience while supporting creative risk taking. Furthermore, the rapid expansion in available information about who our audiences are (and are not) sets the stage for a radically transparent understanding of whom our public arts organizations are serving, who decides what goes on in them and how we communicate that to the public. Evidencing approaches to cultural democracy versus the democratization of culture should impact how we interpret the mission and policies of those that create and fund artistic activity. If, for example, Arts Council Englandâs vision is centered on achieving âgreat art, for everyone,â a conversation needs to be had about what âgreatâ and âeveryoneâ really means and how we measure it. Until then, we will lack as a sector the strategic framework and vision to use big data, so even the best audience, technological and organizational development will have limited impact. In the following, I discuss some learning from key projects that led me to this viewpoint. It should also be noted that while the findings have relevance across the creative and cultural industries, much of our exploratory work has focused on the subsidized touring companies, theaters, concert halls, museums, and galleries in England.
Barriers Are Cultural Not Technical
Throughout all the research projects, we have found that the most significant barriers to effective use of data within the organizations are not technical, but cultural and attitudinal. The data exist, but it is not always clear who has the means and rights to use them. We have found that this uncertainty is a common challenge across organizations, both large and small, and in every art form. Equally, there are pockets of excellence, but the cultural sector and, in particular, the subsidized part thereof are not necessarily of great scale or easily scalable. This is a fundamental challenge to the adoption of truly big data techniques, where size does sometimes matter when it comes to the data set. Many cultural organizations whose existence involves presenting new work to the public must also constantly seek to find a market for what are effectively new products. We are not like a supermarket or grocery store. Nobody needs what we offer in the same way they need food, and individually, we do not have hundreds of products on offer to suit every mood or situation. So while our analysis shows that the loyal core audience remains relatively stable across art forms and organizations, the churn of new audiences to any one venue is huge, in the region of 60% to 80% per year. The question of whether this is good or not cannot be answered by the data. The cost of maintaining such high churn rates can. We have found that business and technical terminology is itself a barrier. Terms such as âdata-driven decision-makingâ (DDD) can be off-putting to many leaders in the subsidized arts sector, who believe that the use of such language and techniques is the top of a slippery slope leading to a purely commercial, transactional relationship with audiences and the cultural sector becoming just another part of consumer society. When delivering a talk on DDD to chief executives of arts organization, Anthony Lilley1 was asked:
Youâre telling me a computer should programme my venue?
Participant in a leadership workshop
This is not a silly question. The hype of big data has promised extraordinary levels of automated decision-making. However, experts such as Nate Silver make a strong case for the human intervention in understanding the meaning of data analysis. In his book, The Signal and the Noise,1 Silver writes about the limitations of automated analysis techniques to prescribe the action that should be taken on the basis of any analysis.
Weâre not that much smarter than we used to be, even though we have that much more informationâand that means that the real skill now is learning how to pick out the useful information from all this noise.
We at The Audience Agency have come to call this the âso whatâ question, and I will talk in more detail about this later in the chapter. What is clear is the importance of context and experience that, combined with values-driven business models, cannotâat least for the foreseeable futureâreplace people as the final decision-makers. We are some way off the prescriptive analytics of the âGartner Model of Data Management Maturity.â2 What is within reach is the application of big data techniques to move beyond the merely descriptive and toward the predictive.
Aggregation and Collaboration
Forecasting the likely interest in an exhibition or event and the likely impact of a marketing intervention could drastically improve resilience in arts organizations while simultaneously increasing the capacity for creative risk-taking. However, truly big data in the artsâof the scale available to telecoms and financeâcannot be generated by a single organization within the cultural sector. There are large theater operators that might come close, but even they would be paddling in the shallow water in comparison to, for example, a supermarket chain. Furthermore, for the analysis to be of use, it must be actionable. In the case of theaters, our research shows that it is usually the local venue with which audience members feel they have a relationship and not some larger parent organization or visiting company, which has implications for who is best placed to use any analysis of captured aggregated data, as well as the previously described difficulty in achieving the volume o...