Over the last several decades, the tools used by entertainment and media executives to gauge the prospects of new products and services have evolved. Early pioneers in music, film, and television famously relied on their instincts as they developed hits. Next came data that was relatively crude and not particularly timely — Nielsen ratings, box office sales, bestseller lists. Today, there’s the collection of big data, the mass of information on eyeballs, downloads, purchase, time-spent on pages, shares, likes, and comments that our powerful computers and servers hoover up daily.
In the minds of industry executives, there’s no doubt that big data can be a powerful resource for predicting what content will succeed, and even for creating new content that has the attributes for success baked in. But all companies involved in making, selling, and promoting movie and TV content face a similar conundrum as they seek to combine new sources of data with primary research: how to harvest the unprecedented volume and richness of consumer data while separating out the noise, and how to use the insights to make content more compelling, reach desired audiences, and ultimately contribute to profitability?
We believe the answer lies in the next logical evolutionary step: applying data science to the massive daily catch that our technological nets harvest from the ocean of data. With the proper application of data science, one can efficiently identify the most important pieces of information, leading to faster and improved decision making. Moreover, in the new digital age where platforms are fragmented and present low barriers to entry, this blend of art and science leads to optimized profitability.
Here’s what we mean.
Every minute of every day, countless millions of people worldwide take to websites, apps, platforms, and social media to express and act on their views — often about entertainment and media content. Liking and commenting on Facebook, tweeting, viewing trailers on YouTube, sharing links, blogging, and more. That’s even before you take into account searches on the likes of Google and Bing. As they do so, they leave behind a large data trail on consumption, preferences, and habits.
Beyond providing data about specific existing content, these same activities can provide crucial information and interactions out there about future content. We’re talking about actor and director scores on IMDb; the number of Twitter followers a star has; trailer views/shares; and box office data per actor, director, and genre; posts on consumer and industry blogs, and Facebook comments about future storylines.
Traditionally, content companies relied on primary research to measure awareness and intent to purchase, and to monitor campaign performance. But this research can be slow, skewed, or lack precision. In many cases, results from primary research alone — from a focus group, early purchase behavior, or surveys — comes too late to allow for any substantial changes to marketing plans. An additional risk with primary research is that it is often very difficult to fully capture what a consumer really thinks (just ask our election pollsters).
However, applying sophisticated analytics can help eliminate much of this bias and provide more real-time feedback, thus allowing companies to shape campaigns quickly. With the power of the analytical tools and platforms available today, you don’t just get to see the number of eyeballs looking at a content site or review at 10:30 a.m. on the first Monday of the month. You also get to see, measure, and often help direct a consumer’s online journeys and actions — clicking through, sharing, posting, lingering, commenting, and more. As a studio builds its analytics factory, it will increase its ability to predict the performance of a film precisely and accurately in the weeks leading up to release.
The opportunities go further. In isolation, each of the crumbs of potential future-related information you feed into your analytics factory might not be particularly nourishing. But when combined with other data sources and predictive analytics, these crumbs can be combined into a powerful set of indicators for forecasting the performance of media offerings that haven’t yet been produced. These indicators can then be used to guide decisions about which projects to invest in and about elements such as casting, storyline, and characterization.
The common thread in the potential applications of big data analytics in making and promoting content is clear. The ability to sift through a mass of apparently unconnected data with precision in order to derive insights and find opportunities that others have missed is like having a powerful magnet that pulls needles from haystacks.
Thanks to these capabilities, big data analytics has the potential to help content developers and producers avoid following the herd by pinpointing and meeting consumer demands that no one else has detected. By opening up new frontiers for content, the science of big data analytics will enhance the effectiveness, appeal, and profitability of art.