Tarun Katial on 3 Vs of hyper-personalisation - Video, Vernacular & Voice
Tarun Katial, CEO, ZEE5, talks about the age of hyper-personalisation and AI, and how they have evolved with the rise of digital
Published - Sep 25, 2019 9:17 AM Updated: Sep 27, 2019 2:26 PM
The e4m Conclave 2019 saw many key industry leaders gather together to have an evening of learning. Tarun Katial, CEO, ZEE5, spoke about the age of Hyper-personalisation and AI.
Speaking about how things have evolved with the advent of digital, Katial reminisced, "I think most of us, advertisers, marketers and agency folks, started in the late 90s or early 2000s looking at target audiences. We were told we were to look for an upper cap of 25-44. You could add gender, geography to it. Then there were some studies which added lifestyle attributes to it and so on, and so forth. Then came digital, about five to seven years ago which started to do user-level activation and tried to understand user cohorts and segments. Then we got excited that Google could tell us more than we know about our audiences."
Katial then explained the need for hyper-personalisation in this digital world. "Why do you need hyper-personalisation? Most of our user journeys today are taken for granted. When you get on to a food delivery platform, you expect it to understand what you ordered last, what are your likes and preferences, your prescriptive data recommendations, and much more. All that is taken for granted because each individual today believes to be served at an individual level rather than at the segment level.
"Machine learning is a broad subject. In fact, ML and AI are two words that are most abused in data sciences today. Everybody believes that their product or platform is well-equipped with each of them. What do we expect some of these things to do? How do we look to uplift numbers when we use them? Is only descriptive data good enough? Or are recommendations the way to go or are they getting into data privacy? What happens when GDPR and data laws come to a country like ours? These are questions that all of us will continue to answer. Today you allow access to much of your data on an ongoing basis."
He also spoke about how there is more data about how conversations are being recorded by some companies. "Are you willing to give as much as the user is giving to you today? If you don't believe that you want to part with your personal data, then how will the machine learn and come back to you? The balance between machine learning, data privacy and descriptive data is something that all product owners and marketers today are struggling to keep," Katial revealed.
Citing the example of ZEE5 and how they have included personalisation, Katial said, "When we got onto the platform, we realised there were things that were essential to hold on to users to be able to create better DAUs, MAUs, retention and time spent. This is what we call the strategy of three Vs. The first one among them is Video. India is a country of a humongous amount of devices. So it starts with something, which is not even part of many countries, which is a feature phone that delivers video. The Jio feature phone which you can buy for Rs 50 monthly or Rs 1500 one time cost, also delivers video. Smart TV which also delivers 4k also delivers video. The adaptive bitrate technology really needed to be put in place to understand what platform the video is going out to, what's the bandwidth, it's ability in terms of serving video, the memory of that phone, lower and lower exits before the video starts, and the quality of experience. That's the technology that India or the world for that matter hasn't seen. We worked towards it in the last 24 months."
Continuing further, he said, "The second one is language what we call Vernacular. We realised that India lived in its languages. Most of you may not know that Google and Youtube are available in 6 languages. We are available in 11-12 languages. To be able to display those languages, to make UIs for those languages, and most importantly to have searches in those languages. That was a big ask because most phones didn't support so many languages on their keyboard. We had to then move to Voice which was our third wheel. So we created voice search in those many languages for everybody to do content discovery."
Speaking about reaching out to consumers in every application possible, Katial revealed, "This I must tell you is an ongoing experience that has got us to have not just one, but 15 applications in the market today. I will also talk about another tool. We partnered with Minute.ly to create an auto preview technology. You have a number of curated times on video OTT platform, auto preview scrubs 30 seconds from a 30-minute video that you are more likely to see, create an auto preview of that."
Wrapping up the session, Katial spoke about a time when there could be a personalised payment plan on OTT. "At some point in time, we want to get to a subscription plan that allows your own mode of content and have your own subscription plan."For more updates, be socially connected with us on
WhatsApp, Instagram, LinkedIn, Twitter, Facebook & Youtube