HITS Connections

Challenge for M&E Companies: Making Sure More Data Isn’t Too Much of a Good Thing

Thanks to the explosion in viewing on connected devices, media and entertainment companies today have access to vastly more data about how consumers are engaging with and responding to their content. But the industry is still in the early stages of figuring out what it all means and how it can best be used.

Speaking at the HITS Broadcast IT Summit in New York this week, Levels Beyond CEO Christy King, who spent several years as a distribution executive with Ultimate Fighting Championship, recalled the difficulty of making strategic decisions back in 2006 and 2007 due to the lack of data on the UFC audience and how they were engaging with the content.

Today, she said, the amount of data available to broadcasters is “phenomenal” by comparison. But that doesn’t mean decision making has necessarily gotten easier.

“Now that we’re able to get so much closer to the consumer we have so much more data. So now it’s a question of what weighting do you put against each data point. Which one of these means the most?” asked Ken Brady, who spent 20 years at Turner Broadcasting before a recent move to the financial services firm Bridgewater Associates. “Now it’s an aggregate question, it’s not just a snap question based on one or two data points.”

Added MarkLogic chief content officer Diane Burley, “What’s going on is that you have so much more data and you have so much data laying in the shadows because it’s really hard to shine the light of day on it.”

While it’s possible to track closely how people are engaging with a piece of content from what they’re clicking on and how they’re scrolling, she Burley said, figuring out how to integrate that information to improve decision making is still a challenge.

“We look at that data as a slice because we’re not seeing the whole 360 yet until we’ve built out those systems,” she said.

To date, most of the user engagement data being collected has been used to drive personalization algorithms, but it could also potentially be used to guide production of the content itself, King suggested. She recalled a case in which a “Hunger Games” movie was being previewed to an audience while the producers and cast watched an audience reaction graph as the movie unspooled. At one point, King said, the graph spiked and the producer turned to the cast and shouted, “more of THAT!”

“How do you use the data to actually affect the content being produced?” she asked. “If you’re not sure which data is really telling you the right thing, how do you not overwhelm a producer by giving them too much data? So it’s an analysis problem.”

According to Brady, changes in distribution patterns have further complicated the problem.

“One of the things that has made it more complicated is we now have full-season binge watching,” he said. “So what do you do if the data comes back after episode two or three and you find out, oh man, people really don’t like that character? But you’ve got an entire season of shows already in the can. You not going to go back at that point and change anything. But there are things you can learn as the season goes on that you can apply next time.”

More data can make decision making better, in other words, but it doesn’t necessarily make it any easier.