The logline reads:
“After a disastrous season, a desperate general manager at a losing major league team turns to a computer science statistical model written by a brainiac to help analyze players for his team, forever changing the business of baseball.”
Over the last year, since Jonah Hill played the Yale data nerd who developed Billy Beane’s statistical model that revolutionized baseball, America is significantly more aware of data science. Throughout 2012 and especially following the Presidential election, news and magazines have been publishing stories about data science and “Big Data”. As industries such as retail and healthcare begin adopting “Big Data” to make better inform decisions, the media and entertainment industry has a huge opportunity, of its own, to deploy data science.
Yesterday we gave you an introduction to data science. Over the next few weeks, we will offer a few posts with specific examples of ways data science, complex statistics and predictive analytics are bound to change media and entertainment; but today, let’s generally consider practical applications from data science for our creative industry. It is time to revolutionize our data and statistics practices, just like the Oakland As did. And in the movie about it, I am hoping that Ryan Gosling attaches to play me – it’ll be an Oscar-worthy role.
Let’s start with the obvious but important: computer generated algorithms are not going to take over Hollywood, writing the perfect script (or even to pick the perfect scripts to develop) or crawling YouTube to find the next Justin Beiber. However, there are some fundamental ways data science and statistics can help media companies to grow revenue, optimize cost spending and mitigate risk.
Every CEO wants a better way to predict revenue. And, if you are running a film studio or a music label, predictive analytics has the power to help you better understand the probable outcomes for your portfolio of intellectual property projects. For example, in film, this means likely box office performance and the corresponding calculations for Ultimates.
Film Finance Modeling
A handful of operations have used Monte Carlo models or black box models to provide film slate analysis. The success of these models is mixed. Certainly many models failed to forecast the pending decline in DVD. And, when the industry experts began to understand DVD sales, not every studio or slate (or investor) was fully informed of the implications for Ultimates. Statistical correlations between distribution windows will benefit from increased application of data science into the industry.
Which exhibitors in which locations will maximize a film’s box office performance? How many screens are optimal? Buyers are seasoned pros, but more tools and better data analysis could benefit the exhibitors and the studios in the location process. This could also have implications for foreign strategies – which territories should a studio retain vs sell? And, in foreign markets, I bet the Head of International Publicity would love more data on which stars of the film correlate with the best box office and where so his press junket is optimally planned. Which international television formats have strongest potential for successful cross-over? Which recording artists have the highest probability of making it on the international stage? Or which foreign market is best to launch the next boy band before their U.S. debut?
From Production Budgets to P&A, from development deals to Advances, there is a lot of room for improvement. And, data analysis and statistics offers some interesting insights. For example, if two equally talented actresses could be cast in a role, both statistically equal in their power to drive box office sales, you may give extra consideration the cheaper actress? And, what P&A spend is actually needed to support a specific film? Can we apply the brilliant Micro-Targeting lessons from the 2012 election season to better identify and convert our likely consumer to buyers of our content? And, what should we be willing to spend in that effort? These are only some of the interesting projects advanced analytics is equipped to address.
Cost Containment and Risk mitigation
We are starting to see advanced analytics applied to Twitter data and other social media platforms; but, there is ample room to apply these fascinating tools to strategic and tactical decision-making. Existing solutions focus on capturing consumer sentiment after products are released to assess how they are perceived and how they may perform. Knowing how sales relate to tweets enables us to try and adjust our spend, either adding resources to drive more sales, or cutting off spending to mitigate losses earlier. The USC Annenberg School and IBM hosted a panel, reported on by Variety, on “The Power of Crowds: Social Sentiment & The Future of Film” where a studio exec mentioned that he knew the movie Cowboys and Aliens was not going to succeed. Why not mitigate loss by cutting incremental spend the instant you know the ROI of that spend cannot be recouped? The real value in advanced analytics is seeing the red flag before the major expenditure. It’s not helpful to tell the captain that conditions are right for hitting an iceberg after it has punctured the ship. The right analytics can help avoid such Titanic disasters.
Intellectual Property Valuation
Are you getting the right value from your Pre-Sale? Is an Actor / Actress getting his market value for his projects? Advanced analytics has the power to crunch the numbers and give you the evidence you need to deliver the best results from your negotiation. In a world where no deals are alike and compensation structures are complex with lengthy timelines to recoupment, the ability to understand and analyze the data is a critical comparative advantage.
Intellectual Property Investment and Development
While there is no fear that an algorithm is going to start picking box-office winners or platinum artists, data science can be helpful for development executives. Writers have known this for years, as they study and replicate the lessons from “Save the Cat.” Data science doesn’t offer a magic bullet, but it can help inform studios and labels to focus resources and investment where the greatest likelihood of return exists today. Micro-targeting can help development executives understand who is the likely audience and how can they be converted to a viewer.
As the business models for online distribution take shape, data science can help inform pricing, release strategy, marketing strategy, and countless other variables. This goes well beyond utilizing Google Analytics and social media analytics platforms – which many marketing departments have mastered. Digital remains the Wild West, and data science has the power to provide road signs along the route to success.
At the end of the day, data science, forecasting, and advanced analytics are incredible tools. They will complement and empower organizations in the media and entertainment sector. But, they are merely part of the equation. Content is king, and data won’t change that. The creative geniuses who have built the industry should never feel threatened by advanced analytics. Instead, they should consider how they can grow and develop their own artistry alongside better data on consumers, the market, and trends. Statistics and data science can help the artist to better distribute their work to the broadest and most receptive audience. Similarly, data science won’t replace media buyers or marketing departments; it will make them smarter more powerful with their decisions. It is a tool that we can all rely on in order to continue growing the consumption of media, entertainment and art.