CES 2019: ThinkAnalytics Talks Metadata, Viewer Behavior Solutions for the Viewing Experience

ThinkAnalytics is showing at CES “how its solutions understand metadata and viewer behavior to personalize the viewing experience and boost subscriber engagement,” according to the company.

“For us, it’s all about being a catalyst to drive the engagement economy and to start looking at what are the currencies of engagement, and how we can really fill those gaps,” ThinkAnalytics CEO Gabriel Berger told MESA at the show. “That’s really important to us,” he said.

Increasingly, if you look at things just in terms of a supply chain, “it’s not the volume that’s starting to kill people – it’s the velocity,” he pointed out, explaining: “It’s the fact that the supply chain is so truncated and so fast….whether it’s content recommendation or whatever part of” the ecosystem that one considers, “you have to be very, very nimble” today, he said.

Data mining led to machine learning and that has made way now to artificial intelligence (AI), he said, adding: “Now we’re at a point where, if you look at the data science part, we look at that as plumbing in a way. We’re very, very good at how you set this up. And what we see is that, as you kind of set this up, it’s the 14 years of learning of best practices of how to do this. And more and more, we’re seeing it’s about best practices: how you set everything up, how you’re doing everything.”

Once you get past the data science, he said, ThinkAnalytics is increasingly seeing that the information sciences part of the equation “is the differentiator – how you imbue content intelligence into this content.” And he told MESA, “we see that at” the over-the-top (OTT) “side, the operator side and, more importantly, more upstream, with the studios and the networks.”

Looking at the industry on an international basis, there is a “somewhat extraordinary” amount of content coming into the U.S. now via Netflix and other companies, he said.

“We’re doing all this content recommendation and search in 40 languages – 43 at this point,” and that’s “really helping,” he noted, adding: “What we see more and more, as much as we push the boundaries and the learning experiences of AI…clearly, the differentiator is in human intervention” – and “that is not going away – that is actually more and more important.” Tagging metadata, after all, is “not just a mechanical thing; it’s looking at the nuance of the storyline and literally bringing in” even, in some cases, “people that write scripts that can bring in their feeling of nuance to be able to do the right kind of tags … for all this content.”

He told MESA: “We think that’s really, really important – to start infusing more and more actionable metadata…to the content as it relates to search and discovery.” And this has been a “really big deal for us,” he said.

ThinkAnalytics is seeing a “huge market expansion,” he went on to say, noting Asia is “very big” now and “we just launched with Viacom 18,” in which “we run 40 million mobile devices in eight different languages in India.” Viacom 18 is a joint venture in that country between Viacom and Mumbai-based TV18. ThinkAnalytics also recently signed a deal with the British Broadcasting Corporation (BBC) making it part of BBC’s iPlayer global online streaming ecosystem, he noted. That “will launch in the next week or two,” he told MESA.

In the U.S., meanwhile, ThinkAnalytics is seeing more “premium content people” request better metadata, and that’s especially important for companies with relatively small content catalogs, he said.

Berger went on to make a prediction: “In general, everything we’re saying regarding content intelligence seems to really resonate. I think it’s going from the theoretical to the actionable. And I think this will be the year of content intelligence.”