How Streamers Are Using Data: From Programming to Experiential Marketing

“In the last 10 years, we’ve seen more changes than in the previous fifty, and it’s all due to digital technology,” said Hulu Chief Data Officer Jaya Kolhatkar while speaking at Collision, one of the tech world’s biggest conferences, held recently in Toronto.

By now, most of us have heard about streaming services’ deftness with deep reservoirs of data to make viewing recommendations and encourage extended viewing sessions. But what is less known is how data impacts almost every corner of the streamers’ operations and strategy. At Collision, Hulu’s Jaya Kolhatkar and Amazon Studios’ Head of Marketing Mike Benson offered up examples of the ways they’re using data that go beyond the personalized recommendations that keep us on the couch longer than we initially planned.

Distribution of binge viewers in the United States in 2017, by age

Credit: Statista

Netflix is undoubtedly the elephant in the room, with a 2019 content spend of $15 billion, about 3 times that of Amazon’s $5 billion and almost 7 times that of Hulu’s $2.5 billion. Also worth mentioning is that Hulu is a domestic platform and its core service is ad-supported. So for now at least, it’s playing in a very different space than Netflix and Amazon. With the recent full takeover of Hulu by Disney, the expectation is that Hulu will experience significant growth in the next 5 years and compete more directly with the other large streaming companies. Analysts estimate that, with the access to Disney’s catalogue, Disney’s upcoming shift to a direct-to-consumer model and its considerable marketing muscle, Hulu will grow from its current subscriber count of 28 million to 40 to 60 million by 2023 or 2024.

What Hulu does with its 15 billion daily data points

As CDO, Hulu’s Kolhatkar oversees a team that analyzes the fifteen billion data points generated each day, and uses this cache to inform almost all the decisions the company makes, from product development to programming to ad targeting. With 28 million subscribers, 85,000 episodes of programming and thousands of content partners, there are many decisions to be made, and on an ongoing basis. Kolhatkar explained how Hulu leverages the data to understand the particulars of its audience behaviour. For example, the average Hulu viewer is 31 years old, a full 20 years younger than the average broadcast or cable TV viewer, and therefore has very different tastes and expectations. Hulu knows these viewers well, armed with each one’s search history and second-by-second behaviours such as scrolling, rewinding, fast forwarding and pause frequency. To augment its profiles of audience clusters and members as well as sharpen its ad targeting capabilities, Hulu also purchases third-party data.

How does the company think about merging all this data with its decision-making processes? Kolhatkar explained that in big data circles, people talk about ‘the 4 V’s’:

  • Volume
  • Velocity
  • Variety
  • Veracity

In other words, how much data do you have, how quickly is it being accrued, how heterogeneous is it, and how accurate is it. Data scientists and analysts then mine this raw data according to these attributes, with the objective of creating a fifth V, which is value, specifically creating value for the viewer that in turn creates value for the business.

For example, there’s Hulu’s sophisticated CRM (Customer Relationship Management) system, which houses the data at the level of the individual viewer and customizes offerings both inside and outside the app, with features such as push notifications and email, designed to increase viewer engagement and satisfaction. The company also uses AI and machine learning to reward binge viewing (defined by Hulu as more than 3 episodes consumed in one sitting) with an ad-free viewing experience. Also being experimented with are less intrusive ad formats, such as those served on the screen when viewers hit the pause button, as well as ‘situational ads’, described as ads that “target binge viewers with a creative that is situationally relevant to their viewing behaviour”. The company’s plan is for half of their ad revenue to come from these new ad formats by 2022.

Amazon looks to experiential marketing

In some ways, Amazon Studios competes with the other giants in the streaming space and in some ways, it does not, as it’s a perk that comes with an annual subscription to Amazon Prime, a service that provides free shipping, along with access to digital music, video and book content. Even though it’s an add-on to what is principally an e-commerce offering, Amazon Studios is still spending billions on content each year and knows it's competing for its customers’ time across an array of services and platforms.

As Amazon Studios’ Head of Marketing, Mike Benson is tasked with figuring out how to use data to anticipate the size of a show’s audience and how to keep fans actively engaged in-between their binge viewing sessions. Benson used a classic principle of marketing to explain the mindset he uses: “You either think in terms of getting more customers to buy what you’re selling, or you can think in terms of getting existing customers to buy more, or buy more often.” Benson and his team therefore use data to understand how to put shows in front of the right people, to determine which ones are likely to satisfy niches and which ones are likely to be hits, and they then allocate resources accordingly.

That’s how Benson and his team decided upon an ‘in real life’ brand activation for the popular Amazon series The Marvelous Mrs. Maisel in late 2018. The concept: bring New York City’s legendary Carnegie Deli back to life, albeit temporarily. The deli first opened in 1937, experienced its real heyday in the 50s, and then closed its doors in 2016. It was known as a popular hangout for comedians in the 1950s and is itself a character in The Marvelous Mrs. Maisel. Benson and his team came up with the idea of a week-long Carnegie Deli pop-up, complete with 1950s prices, period-specific design elements, rotary phones that let you eavesdrop on recordings of imagined conversations from nearby tables, and even anachronistic messenger boys who delivered the news about the re-emergence of the deli to press and social media influencers.

As one might expect, the lines were often hours long, though 11,000 New Yorkers managed to get served during the Carnegie’s comeback. But the real upside was the buzz, the positive sentiment invoked, and the media impressions created by brand activation: over 3 billion impressions in total, with 70 million reached on social platforms through deli visitors’ posts on the likes of Twitter, Instagram, and Facebook; and an overall positive reaction to the week-long event pegged at 98%. As Benson put it: “Social scales the experience.” To build on the success, the activation was not a ‘one and done’. The pop-up also received coverage via traditional media, through outlets ranging from The New York Times and Vogue to the Today Show and Entertainment Tonight. Also, to keep the buzz alive, it was brought back as a mobile pop-up in 6 locations in New York City in May of 2019.

Several years into the move to streaming and data-driven entertainment, the biggest players are demonstrating how far they’ve come from pulling data from viewer behaviour and pushing thumbnails to screens. Today, a blend of on-screen and off-screen initiatives is keeping the data points flowing in both directions, from platform to person and from person to platform.

Leora Kornfeld
So far in life, Leora has been a record store clerk, a CBC radio host, a Harvard Business School case writer, a blogger and a crossword puzzle clue. Currently she’s a media and technology consultant, working with clients in the US and Canada.
Read Bio