Tapan Khopkar 

PhD 2008

After seven years of doctoral studies at UMSI, Tapan Khopkar graduated in 2008 with a PhD in information science focused on the economics of information. His career path in the data science field includes lead roles in marketing science and product engineering.

Now executive director of marketing sciences at OMD, the world’s largest media agency network, Khopkar encourages students to get their hands dirty: Take on projects involving the kind of big, messy datasets that professional information scientists actually encounter.

Could you talk about your career journey and how you’ve seen the information industry transform?

Absolutely, all the experiences were interesting in their own ways.

Professional roles in information science evolved and became more mainstream in the last 12 years. Back when I was first starting out in 2008, data roles were esoteric and only big companies would have openings for someone with my degree.

My first company was in India, and this was before meeting remotely was a thing, so you had to go to clients’ offices to meet them. A lot of consulting in the early days was focused on relationship building rather than actual data science, as clients didn't understand data science and didn't understand the value we can provide for them. Now I think every company understands what they want from an information specialist. They want someone who can dig into data and create value.

But one thing has not changed: the clients’ data quality. Clients think that many things are possible, but the clients’ ambitions are never commensurate with the quantity and quality of data that they have. Nevertheless, I think an information graduate is much more central for any organization now than 10 years ago.

Now information professions are more mainstream, which is a mixed blessing because competition to get in the field has also increased. It may be difficult to stand out in the beginning. But having a holistic view of information will set you apart once you start establishing yourself. There is a difference between someone who has learned how to do a gradient boosting algorithm versus someone who understands how to apply it on a range of data. That difference becomes apparent once people start working with you. However, to get to that position, competition has become intense. 

As data science becomes more mainstream, new graduates are finding entry roles competitive. What do you think about this?

There are many people who have the qualifications, yet we struggle to find people who can actually do the job. The entry market has become saturated in the sense that there are many people who take a one-year data science program and have the qualification. But when we interview these folks, we realize that their skills aren’t a fit for what we need. They would have to be trained. How data is in the real world is different from how it is in a well-structured problem with a nice, tidy data set.  

If you have a degree from UMSI, then you'll be taking a holistic approach to any situation that is thrown at you. You understand that this is a problem in the real world where real people make real decisions. It is not an abstract academic exercise. 

In academic courses, you often get well-defined problems and nice data sets, then all that is required of you is to run some model and get a result. But UMSI teaches you to take a holistic view of the data that you receive. What is the process that produced this data? Can we trust this data or not? The outcome that you derive from processing the data, what does that translate to in business decisions? This is an area where my training at UMSI helped me understand the entire life cycle of information and data much better.

One thing that really worked for me is the interdisciplinary nature of everything we do at UMSI, in thinking, structuring the research problem and the solution. Observing my teachers and interacting with my cohort worked for me. Their incredible work ethics rubbed off on me. I worked very closely with Paul Resnick. My dissertation was on reputation systems, which was one of Paul's core interests at that time. Cliff Lampe was a great mentor during my years of PhD.

What does a day in your life look like? 

In a media agency, there is no such thing as a typical day. My company, OMD, is one of the world's largest media networks. At OMD, we help clients meet their marketing objectives through a combination of data, technology and creativity. Clients often come to us and ask, “Which of my marketing efforts truly drive results?” That's where the marketing science team comes in.

We are a small team of five specialists. Two big buckets of what we do: first, media marketing mix modeling and multi-touch attribution, and second, budget allocation and budget setting. The nature of the work is fast. In our recent project, the day of first communication to the day of final delivery only spanned two and a half weeks. It's fast-paced and exciting.

My weekly schedule typically involves a mix of project work and providing support for people across the network. OMD works in more than 100 countries, and our marketing science department is responsible for supporting people in the EMEA countries. In a week, I speak with people from two or three different countries across different domains.

How do you balance work and life?

Managing work-life balance is something you need to do. Work can consume you in media agencies. OMD has a great work culture: We help our employees manage work-life balance, and your life is taken as seriously as your work is.

What’s your advice for recent graduates entering the information field? 

A degree at UMSI prepares you to be a specialist who can be a generalist. A degree in information is going to make you a relevant candidate for many positions in many domains. What you learn can be applied in many contexts. But that's a mixed blessing.

Familiarize yourself with recent developments in software engineering (for instance, cloud computing platforms on Google Cloud Platform or Amazon Web Services) so that you can apply your learnings. If you are interested in a specific domain, I encourage you to find hands-on experience there. The best thing you can do to get noticed is have your repositories available and downloadable on Github or GitLab. Your public persona as a developer is important, as it is more real than things you can put on your resume.

What are your recommendations for students seeking to gain hands-on experience?

Kaggle is a good starting point to create your repositories. Try to participate in one contest every quarter. There are many contests on Kaggle that you can partake in. Now, success on Kaggle has become quite gamified, so you can get a great rank if you know certain tricks. Still, participation on Kaggle will give you skills applicable to real problems.

In business problems, data will often be messy. The data set will be large. You will have to really struggle with the data. Well-structured problems and nice results are not what the world looks like. Get your hands dirty and really struggle with realistic, big data. You have to apply some software engineering principles to solve those problems, because your resources are going to be limited.

Finally, it is necessary to make your profiles publicly available. When I get a resume, I look them up on Github more than on LinkedIn. Having a good LinkedIn profile is also necessary – it is not going to make you stand out, but the absence of it will hurt you. In a nutshell, when somebody looks you up, make sure they are able to find you.

What are your recommendations for recent graduates navigating an uncertain time in the tech workforce?

You have to keep learning, keep your profile public to demonstrate what you can do, and hope that you will be found. You need a little bit of luck, which you can't will yourself to get. But you can make yourself visible and that increases the likelihood of luck. Continuing to acquire skills and hoping for better times is all that we can do.

— Tuesday, Sept. 3, 2024