Ashish Thusoo, CEO and co-founder Qubole, previously ran Facebook’s Data Infrastructure team. “Facebook was one of the earliest companies to capitalize on big data, and I had the
privilege of having a front row seat to its transformation into a data-driven company,’” he explains.
He founded Qubole in 2011 taking what he learned at Facebook about
creating an open data infrastructure and applying to the public cloud.
Charlene Weisler: From your perspective, how has TV measurement changed?
Ashish
Thusoo: The world of television has experienced three fundamental changes over the last few years. First, TV has become digital….In today’s world, you know who is watching, what
they are watching, how long they are watching, and how they are behaving, which means the types of things you can measure has grown exponentially.
Second, TV has gone multichannel….
These days there are an infinite set of distributors and broadcasters, so viewers can watch content through cable TV, stream through subscription video on demand services such as Netflix or Hulu, or
access websites and apps like YouTube and Snapchat…. And they can consume on multiple devices simultaneously. This variety creates endless ways to measure video consumption, but also
creates challenges.
Finally, it’s not just about TV. You can correlate Web, social and offline behavior to build an even more complete profile of each viewer. Those three changes
have led to an explosion of data and increasing complexity about how to create meaningful measurement. While companies have new opportunities to collect data, the challenge now becomes how to get
value out of it.
Weisler: Tell me about Qubole and where it sits in the TV measurement world.
Thusoo: We provide big data-as-a-service, meaning we
help organizations in the media, advertising, gaming, e-commerce, TV and entertainment industries turn their data into business insights. Our cloud-based platform, Qubole Data Service, runs on public
clouds such as AWS, Azure, Oracle and Google Cloud Platform, and addresses the challenges of processing huge volumes of structured and unstructured data. For instance, our solution helps companies
process all the data collected by tracking and analyzing consumers’ consumption of video content.
Weisler: Are you able to help develop new metrics for TV
measurement?
Thusoo: In the past, TV viewership was measured on historical information. This evolved into real-time measurement, as advertisers and distributors could find out
who exactly was watching a show on a real-time basis.
The next measurement for TV will be predictive. Not only can you learn historical and real-time information about a viewer, you’ll be
able to predict what he or she will watch next.
Weisler: What role do you see data playing three to five years from now?
Thusoo: In the next few
years, technology will emerge that can determine precisely who is watching something, through biometric face recognition or eyeball tracking.
The challenge today is that media companies
aren’t able to know 100% for certain who is watching something. A middle-aged father, for example, may watch the first half of a football game in the living room, but then leave to make dinner
and his young son watches the second half without him. And yet, the same commercials will play during the son and father’s separate viewing experiences, although they aren’t relevant to
both of them. Biometric tracking could be something we see as a tool for the media to know precisely who is watching something and optimize his or her experience.
Additionally, we’ll see
the creation of hyper-personalized viewing experiences by combining a user’s past, real-time and predictive experiences. For instance, predictive analytics will anticipate when a consumer will
prefer to watch a commercial and then change a consumer’s viewing experience in real time.