Sramana Mitra: We do content-based marketing largely. Everybody that we encounter finds us through our content. We essentially run a media property and all the ancillary distribution of our media work. We put in as much finesse as possible into that work.
Diaz Nesamoney: If you think about personalization, it’s essentially content-driven. We’re just saying, “If this person is interested in this kind of car, why not share with them content around that instead of showing them generic things?”
Sramana Mitra: I think that content is becoming the brand these days. It will increasingly be that as people figure out how to do that. It’s not so easy when you’re doing a toothbrush. How much content can you produce around a toothbrush? It’s not easy to do.
Diaz Nesamoney: I’ve seen an interesting transition and opportunity in the world of creative. Creative is not just about making something look good. It’s just as much about making something relevant. Not that looking good is bad, but because a certain part of us don’t buy rationally. So is relevance.
Creative agencies are starting to think more of relevant content. It’s not one-size-fit-all content that looks beautiful. A beautiful-looking ad that is completely irrelevant to you isn’t going to get you to buy, but if you can produce content around the product that’s tailored to different buyers of even that one product, then you’re more likely to get people engaged.
Sramana Mitra: That’s where television advertising fails.
Diaz Nesamoney: Yes. It’s one-size-fits-all. In digital, we have a tremendous opportunity to do so because it’s possible. You have more data. Technologically, it’s possible and it’s cost-efficient.
Sramana Mitra: I think that’s better for consumer products which sells through retail because either way, you don’t have the data to be able to do great personalization.
Diaz Nesamoney: True, but we’re seeing a lot of transmission there too. Even if they don’t have the data, using third-party data, they can tailor and customize. It may not be as precise as what our retailer can do, but it beats the heck out of having the generic ad. The product is also getting very nuanced. Do you take your marketing budget into tiny little pieces? It’s very challenging. Even if the data is not as precise and as rich as you would like it to be, relevance beats the heck out of one-size-fits-all.
Sramana Mitra: Last question, where are the open problems? If you were starting a company today, what would you start?
Diaz Nesamoney: People would say that AI is overhyped.
Sramana Mitra: I don’t think so at all.
Diaz Nesamoney: I don’t think it is. I think AI a very fundamental and foundational transformation of software and society. Having been through generations of software and technology, we always talk about Moore’s Law. I think we’re reaching the end of the scale there.
We’ve always used deterministic and algorithmic ways of solving problems. We’re reaching the tail-end of that cycle because the kinds of problems that we have to solve right now that remain unsolved are very difficult to solve deterministically. If you look at the breakthroughs in facial recognition and voice recognition, they’ve been around for ages. People have been knocking their heads against it.
The breakthrough came when they said, “Let’s not try to write rules about how you recognize the face. Let’s use data and let the machine learn.” There is plenty of data. It’s getting so complex that a simple deterministic rule-based approach is not going to work anymore. If you look at self-driving cars, we can’t do that algorithmically. I don’t care how much hardware you throw at it. You’ve reached the limit.
Sramana Mitra: You have to model the data. You have to learn from the data.
Diaz Nesamoney: I think technology is going to have to start mirroring humans because we’re not programmed with rules. We figure stuff out instinctively.
Sramana Mitra: There is this notion of evolution in our systems. You take inputs and evolve. The fact that, today, we are enabling software to do that is the most powerful transition.
Diaz Nesamoney: If companies don’t transform themselves quickly, they’re not going to be the large successful companies they are today. If I were starting a company again, it would be a company that is applying AI and machine learning to solve an existing problem. It’s just solving it better.
Sramana Mitra: Great. Thank you for your time.