How can the science of data best be used in spaces as subjective as customer experience? For Gregor Stewart, a longtime fascination with that complex question was a big reason he joined Medallia. Below, he explains the problems his team is solving, shares how he thinks about leadership, and discusses the passion project that often brings him to the office before the sun is up.
I’m VP of Data Science, which means I lead a team of researchers and engineers who are responsible for three things: First, we drive the development of the infrastructure needed to bring AI to Medallia’s platform. Second, we build core capabilities that sit on top of that infrastructure and make it easier to use—so, for example, we can develop a single model that works for multiple human languages. Third, we define and deliver intelligent features that help Medallia’s clients do their work by helping the product teams understand what’s possible. Sometimes what seems easy is really very difficult. But nearly as often, the opposite is true—something people consider magic is actually pretty straightforward to do. Helping people at the company understand what’s possible is the most satisfying—and probably the most valuable—aspect of our work.
Immediately before Medallia, my wife and I both took about a year off to be with our newborn son. Before that, I worked for a company called Basis Technology, which is a leader in natural language processing (NLP)—the branch of artificial intelligence that helps computers understand and make use of human language. I’ve been fascinated by NLP since early childhood, when I first played the text-based adventure game Zork and read about ELIZA, aka the mother of all chatbots. Since then, our understanding of human language and our ability to build systems that do useful things with it has increased immeasurably. Most recently, statistical and deep-learning techniques, which I studied when I went back to school in 2008, have allowed us to unify our understanding of language, imagery, and behavior.
One of the reasons I joined Medallia is that the domain of customer experience offers an abundance of unique data. I also liked that the domain is still taking shape. Even now, customer experience is about a lot more than collecting scores and closing the loop. We’re learning to model how people experience things based on circumstances and who they are, and how that understanding can shape the experiences our clients offer their customers. And we’re operationalizing a value I hold deeply, which is to ask questions and listen closely, emphasizing communication as important as observation. That’s exciting to me.
Essentially, we help Medallia’s clients understand the experiences of their customers and take action to improve them. We do this by focusing on questions like, “Who and what is it important for us to focus on or know more about?” and “How should we get that information?” A simple and effective approach could be for a client to reach out to everyone who scores them six or less in a survey. But for some of our clients, that might be hundreds of thousands of people per month, many of whom don’t actually need to be called. Instead, we can use our knowledge of language and people to help clients decide whom to contact first, saving time and expense.
We also help clients figure out what questions they should ask—and when to stop asking, too, to avoid survey fatigue. Then after the initial feedback has been gathered, we help follow up in an intelligent way. If your business has a mediocre net promoter score, for example, improving it is going to be about more than just closing the loop with a customer. Data can help you diagnose broader issues and suggest solutions to try.
Our scope of work is broader than it is for many data science teams, but this is entirely in keeping with Medallia’s culture of creativity and innovation. And Medallia is unique in that it already has millions of users, many of whom use the platform daily to measure progress and learn how they can improve the lives of their customers.
So we’re not maniacally focused on optimizing a small number of metrics like click-through or sign-up rates. We don’t go fishing in our datasets for “five weird tricks” to exploit. We start with, “What are our clients trying to do?” Then we work to answer, “What data and techniques can help them to do those things?” Rather than constrain our thinking to just algorithms and models, we get to design collaborations between Medallia’s platform and the people using it every day.
Finally, while we have to build systems that learn continuously, we also have to respect the data of each user, client, and customer. We’re trusted with a massive collection of information about how people experience the products, services, and workplaces of the world’s most successful organizations, many of whom compete with each other. Protecting that data is a top priority.
I write short sci-fi stories—although that’s not always technically outside of work, because I do a lot of it at Medallia! Our son wakes up early, so to find quiet time for work, writing, and research, I get up around 3 a.m. and head to the office for a few hours. Then I go back home to spend some time with my family before the regular workday starts.
Inspired by an old Star Trek episode called “Wink of an Eye,” my most recent fascination is with the ways machines help humans in barely perceptible slivers of time. I love to consider all the clever things cars do to keep us safe in a crash, for example, and to explore what happens when you add agency to that intelligence. I try to capture a vision of the future in which, through learning to better protect us from harm, the things we’ve built may come to love us.
I think leadership is primarily about cultivating relationships. When you can build a relaxed and free-flowing connection with someone, you both become more creative and resilient problem solvers. And when the people on your team know they are intrinsically valued by you and their teammates, they pull hard and long at the oars—together. My best mentor, a former colleague at Basis, is constantly working to establish and maintain authentic relationships of this nature by staying curious about people and their situations. I try hard to do this, too, whether I’m working with a teammate or one of my mentees in WIRED, which is Medallia’s program for developing female leaders. I’m learning not to overcompensate when someone is struggling, and I’m getting better at supporting my colleagues without getting in the way of their development.
As a leader, I have been extremely lucky at Medallia. Everyone I work with loves what we do and is highly motivated. The best thing I can do is to help my team find exciting and impactful work.
This story was created in conjunction with Job Portraits, a San Francisco-based creative agency that helps teams scale using culture-focused content.