News

How A Scientist Is Helping the Music Biz Unlock The Power of Data

SK Sharma
Illustration by Ollanski

SK Sharma

This story is part of Billboard's annual 40 Under 40 list, recognizing the music industry’s trailblazing young executives.

SK Sharma
Chief analytics officer, Ingrooves Music Group

At 23, Sharma graduated from Caltech with a PhD in chemical physics and physical chemistry. Sixteen years later, he’s far from the lab — but using the same skills leading the UMG-owned distributor's insights and analytics team. Since joining the company in 2016, he’s used artificial intelligence to create its Smart Audience tool, which analyzes streaming behavior to identify listeners who are more likely to become fans of, and stick with, an artist. (Both the tech and the tool are now patented — a landmark achievement in the music industry.) Sharma, 39, explains his unusual career path.

Your career has taken you from lab research to consulting to intellectual property venture capital. How does that inform what you do at Ingrooves?

My research emphasis was in therapeutic drug design, and a lot of the analytical reasoning and rigor that we bring to the work we do today stems from my early experiences as a scientist. We want to ensure that we provide every opportunity possible for [our artists] to be effectively heard, and to me, that draws from aspects of portfolio optimization — not in a clinical, corporate finance way, but how you drive engagement. It just so happens that you can drive engagement these days with techniques that come from quantitative physical sciences.

Can you walk me through an example of what that work looks like?

We have tons of data points [from digital service providers], but data by itself doesn't really mean anything. It's about what you do with it. Well, you could start to build these listening clusters to get a sense of how people's [listening behavior changes over] time. You can start to understand the decay process of a new release, you can understand the growth profile. You can understand how consumption and engagement shift as social signals change, or with a Netflix synch. You can do all this really cool math and put together something that is predictive based on information learned, but also based on the principles of causality: Did this cause this? Did this move the needle? With the Smart Audience platform, we're looking to optimize marketing spend in ways that return high-value audiences, and in ways we can measure — as opposed to saying, “I threw a bunch of money at this, this is what I think I got.”

What is a typical day like for you and your team?

Our team is fully integrated into every aspect of the business; we don’t work in silos. Depending on the day of the week, we might be on a call with our engagement team or our commercial analytics team. [We might be] working through a really hard problem, trying to understand how to make a particular model smarter so that it's going to be even more successful and able to correlate some offline event to online engagement on a wide range of platforms. We might get on a couple of meetings to talk about new releases and get a really boots-on-the ground sense of what's coming down the pipe — what the new release cadence is, how we’re thinking about back catalog — and using all that to stimulate conversations and ideas. It’s a combination of scientific testing, assumption testing, gut testing, but really in an integrated fashion.

Seven years ago, you were toying with the idea of early retirement — and then you were recruited for this role at Ingrooves. What convinced you to take on the job?

I was drawn to our CEO Bob Roback's vision. He had just joined Ingrooves in 2015 after a successful time running Fender, and then in 2016, he articulated his vision to me around the very things we're doing today — really focusing on the engagement economy and borrowing from scientific rigor to build a totally different way of thinking about differentiation, not just in the product space, but in the industry.

Was the idea of building this insights and analytics team always part of the pitch to you?

It was certainly a core part of Bob's strategy to succeed in the independent music market space. And that was of interest to me because that's exactly where I thought we were going to be able to really create value, not just for us, but for the entire industry to really change the way people do things and the conversations they have. Had that not been a core component of the plan, I certainly would not have had any interest in the industry at the time.

I told him, “We're gonna have to hire differently, we're going to have to hire PhDs, people with a background in science.” And he said, “That sounds great, let's do it the right way.” Dashboards are sort of static — they show you information you already have. What we really want to be able to do is change behavior. Even four or five years ago, [that attitude was] relatively rare in the music industry outside of Spotify.

What was the biggest culture shift as you entered the music industry?

One of the things I noticed about the industry was that, historically, people were judged by their success. There was this element of “I did this, therefore I have credibility.” But in the scientific process, you accept failure as a natural part of growth. You have to be obsessed with the culture of hypothesis testing — of being comfortable with being wrong so that you can learn how to be right more often.

You took a fairly unorthodox path to this role. What advice would you give someone looking to get into this area of the industry now?

I would say it's really important to appreciate the rigor in science and math. We know that they're very complementary [subjects] to music in general — there's a lot of neuronal studies that show a high degree of correlation, so I think appreciating that and really understanding that this culture of controlled experimentation and hypothesis testing is incredibly powerful. It's not just about pushing buttons, about coding something and releasing it. That's not a scientific approach to really trying to do valuable things.

Being highly collaborative is not being afraid to be wrong. Being okay with elements of uncertainty that you're looking to improve upon over time is a very critical component of success — and so is ultimately being very comfortable with the fact that you work in a diverse, collaborative industry where multiple opinions are needed. Algorithms are not truth all the time. You need to sit down with people and connect with them in real life in order for something [you create] to be effective, persuasive and ultimately adopted.

It’s unfortunate the way AI is talked about by those who don’t really understand science or mathematics — like, “The robots are coming for you!” We just don't look at it that way in the music industry. [At Ingrooves] we don't do any AI content creation — we're music lovers first and foremost. It’s really about using the best parts of AI to think about how we can solve problems in the music industry that no one else has solved yet.

What role did music play in your life before Ingrooves?

I grew up dirt poor in Compton — Eazy-E was my eighth-grade commencement speaker! Music gave me hope that one day I’d end up away from that day-to-day negativity. Growing up in the beautiful diversity that's L.A., you turn on the radio like, “I'm going to listen to [hip-hop station] 92.3., but I'm also going to slide the scale to listen to KLOS, classic rock — what the heck is that?” And then you take it further to KROQ: “Oh, The Offspring and Nirvana, they’re pretty cool.” It really was a broad education through the radio and the culture of living there. Music has been transformative in my life.

This story originally appeared in the Aug. 28, 2021, issue of Billboard.