Does Big Data Matter to Indie Filmmakers?
In the age of the internet, a rapidly growing ocean of data is being generated, collected, and in some cases used by marketers to target customers.
And it isn’t just active online user behavior that is being mined by marketers.
Mobile phones are also creating vast amounts of information (e.g., even if your phone is “off,” your phone company keeps "a database of where phones are likely to be" and therefore has a pretty good idea which movie theater is closest to you right now).
And the so-called “internet of things” - enabling data collection from previously dumb appliances and objects – may soon change marketing forever (e.g., it won’t be long before you’ll own a refrigerator that texts you when your milk has reached its sell-by date - but information from your fridge - like treats purchased for a date or upcoming televised sporting event - may also soon trigger movie suggestions in your Netflix queue).
What does all this mean for filmmakers?
We already know that Netflix is mining big data to make user recommendations and to tweak their content acquisition decisions.
And we are beginning to see businesses built around extracting useful information from the vast amounts of online video (e.g., security camera and cellphone videos) "based on what’s going on rather than how they’re described."
But how will this new world of so-called “big data” impact individual small budget filmmakers?
Does big data even matter to indie filmmakers?
Isn’t big data just a tool for marketing to mass audiences?
A crutch that will kill creativity?
The enemy of indie film?
The idea that data can be analyzed in a creative way to help make marketing decisions is something that has been axiomatic in the big studios for decades.
(Long before there was an internet, my first steady gig in Hollywood was shuttling videotapes and questionnaires to-and-from test screenings conducted by the studios back in the 1970s. We called it “audience studies” back then – and I know from first-hand experience that the data we were collecting was deeply flawed.)
Still, it’s a mistake to allow Old World ideas about the flaws of “audience studies” to cloud our vision of how big data might help us now.
Today the flood of information being generated by digital devices and their users is poised to change how indie films are financed and marketed in fundamental ways.
For indie filmmakers to benefit from the New World, we first need to understand how big data works and to abandon our biases and misconceptions about market research.
The first misconception to discard?
Isn't big data just that stuff being collected by the government and multinationals?
While it takes big money to collect and organize vast amounts of data, the term “big data” also refers to how specific datasets are used – for example, how an indie filmmaker might analyze information about what movies cable subscribers are watching on VOD to yield predictive information based on key correlations (e.g., fans of gore who live within 5 miles of a particular theater).
In other words, the trick to using big data may not be in having the resources to collect it and store it, but in unlocking its value for our unique use cases.
Some forward-looking entrepreneur – with access to someone else’s large datasets and an appreciation of the challenges faced by filmmakers – might just be be the one to unlock big data’s promise for indie filmmakers.
Besides identifying fans of gore near a certain theater, how else could big data pinpoint customers who would love a particular film?
How will the predictive models of big data be different from the Old World tools?
For indie filmmakers, the dream is that big data might minimize marketing costs by narrowly targeting just those potential customers with key psychographic identifiers (markers that indicate the highest likelihood of enjoying a particular film).
With big data, it should be possible to alert a small core of fans – fans who will spread the word about a film via social media.
But how are cash-strapped indie filmmakers supposed to access a huge dataset to find just those potential fans who will become rabid backers?
My best guess?
It will probably require a well-financed coder to come up with an easy-to-use tool that filmmakers can use to get at information that “would have been hard or impossible [to get] with smaller samples [and] fewer variables.” Liran Einav and Jonathan Levin, “The Data Revolution and Economic Analysis”, Prepared for the NBER Innovation Policy and the Economy Conference, April, 2013, p. 2.
And the challenges don’t stop with simply building a big-data-for–indie-filmmakers tool…
The marketing of such a new tool will also require money and creativity too.
So if it will take time and money and expert coding, why bother?
The rewards of targeting new films to core audiences could be huge.
And, while the coding and marketing challenges will be big, the datasets the developer will need for filmmakers to identify potential core fans probably already exist.
That’s because a smart developer could use data that was not originally collected with the goal of helping filmmakers.
(As Mayer-Schönberger and Cukier have observed, the initial reason for gathering data may have little relation to how that data is used: “With big data, the value of information no longer resides solely in its primary purpose.” Mayer-Schönberger and Cukier, “Big Data: a revolution that will transform how we live, work and think”, Houghton Mifflin Harcourt, 2013, p. 153)
So an algorithm that teases out a number of factors that correlate with potential rabid enthusiasm for any given film (e.g., theme, genre, actor, story, etc.) might be applied to a dataset that already exists.
Any developers out there interested in discussing likely datasets and algorithms for indie filmmakers?
Randy Finch's Film Blog:
Thoughts from a film producer about making and distributing films.