Data Science as a Team Sport: Building a Culture Open to Risk
The head of a data scientist team at Cisco discusses how he hires and manages his diverse group -- and why the ability to write software code is not a prerequisite.
- By Linda L. Briggs
- April 26, 2016
At Cisco, Robert J. Lake has turned the popular term "data scientist" on its ear and uses some unusual methods for hiring and managing his team. As the senior manager of data science, he leads a team who does "operational research" but says the group exists to help people make decisions, whatever method that involves -- not always traditional data science.
Upside: Data scientist is a new term that's generating a lot of interest now, but you've actually been a data scientist for many years without being called that, right?
Robert J. Lake: Right. I've been working in operational research since 1998. From a skill set point of view, operational research has been around a long time. I got introduced to data science -- or what we call data science now -- in my master's degree program in manufacturing systems.
We were doing operational research and simulation and optimization and design and Six Sigma work -- all those fun things that come together to help you understand using artificial intelligence, using calculations, calculus [and so forth]. Everything you do if you want to design or develop something ready for use.
Operational research has been a good skill and I've always used it. It's funny -- it was only two years ago or so that someone came up to me and said, "I finally understand what you do. It's called data science. I now recognize that's what you've been doing all this time." So now we have a label for it and people realize what it is. To me, it's just operational research.
What we do here is mixed in with other bits and pieces, but operational research has been around a long time. It covers the whole concept of how to build something, experiment, and use mathematical proofs to build a model, then be able to demonstrate how the world works within the constraints of that model.
How does that approach differ from that of a business analyst or another sort of business intelligence expert?
A BI expert tends to look in more of a descriptive area...describing things and putting things into reports. We take that a step further. We look forward and we look backward, but we're not just looking backward at where we've been. We're looking backward in terms of what caused us to be where we are. There's a cause and effect component in there.
What if all the different combinations of things that could have caused us to be where we are today actually came to fruition? What are all of the different things that could cause us to be in multiple different places in the future? We try to understand what those are. Then we try to understand how we can influence the future, and if we can't, perhaps, then how do we prepare for the inevitable?
Can you describe the skill set you look for in hiring data scientists for your team at Cisco?
A lot of people focus on technical skills. I'm really much more interested in soft skills, because at the end of the day, it comes down to the culture I expect for the team -- the culture I hold my team accountable for. The team culture describes who we are looking for.
We're looking for people who are willing to learn, and not just about the techniques and approaches that we use in the field. We need a willingness to learn about the environment they are going to be working in -- the subject areas.
They have to be passionate about learning and be willing to learn, and they have to have an open mind. If you have an open mind, it means you listen to people. You can't come in and arrogantly tell a business leader, "Oh, by the way, you need to improve your sales, and I know the answer." That can get you in trouble. The ability to learn is key, because that means you are listening.
What's another critical skill you look for?
Next is the ability to take risks. It's OK to fail, fail fast, recover, and learn. We come back to learning again. The more you fail, the more you learn. The less you fail, the less you learn.
You know, not failing has been drilled into me, into all of us, and I never realized it until perhaps five or six years ago. I used to hate failing. I hated failing in structured courses in the military, for example -- I always wanted to do my best. We're taught not to fail; we're taught that you have to succeed. I think we need to turn that around and allow people to fail, and then ask them, why didn't you do as well as you thought you would?
I try that with my children; I allow them to fail. They get upset, and we help them through the emotions involved, but they learn so much from failing, because they try, and they try again, and then they succeed. That's a really big point I look for in the people I work with on my team. I want them to be able to fail, and fail quickly, and then recover and get over it.
Here's the third thing [I look for in hiring], and this one is really important: they have to have an insatiable desire to solve problems. Not just any problems, problems no one else will touch. People will say, oh, you can't do that. It's too hard. It's impossible. I want people who say, hell no. Let's not walk away from this just because it's difficult. It's only difficult if we look at it from a particular way.
Maybe if we change the view, if we approach it in a different way, it's not impossible.
At the end of the day, I want people who, when you tell them something is impossible, want very much to solve the problem. They need to be able to look outside the box and to think about things in a different way. When you listen, and you think outside the box, you can solve problems. It's amazing how quickly you can do that when you know how to approach it.
Those are the three things I look for. The job needs passion for solving problems and making a difference, and it doesn't take 100 days for my team to solve things. They do it quickly and effectively, because they know how to break problems down. When I look for people, that's what I want.
That's quite a list of soft skills. Does it preclude someone just out of school? When I saw a photo of your team, they looked very young.
I have a very diverse team and I'm very proud of it. One of the disgraces in our industry is that at last count, the statistic was something like only 8 to 12 percent of data professionals are female. That's a shame. When I hire, I just look for the right people.
When you look for the right people, it's amazing how diverse they happen to be.
When you think about it, I don't just want people with experience. I want people who are new, who are hungry, and who want to accomplish things.
If you take someone just out of school, they have their own way of thinking. It may not be the way you think about things, but if you just listen and think back to the three important skills I described earlier, they hit those. They're thinking outside the box, they're taking risks, and they want to have a go at it. They don't mind failing. So you mix that, the younger group, with older and more seasoned professionals, and that's an amazing group.
The key to the whole thing is listening and being respectful. It's not about young team members coming in, listening, and then shutting up and doing what they're told. They fully expect to be part of the team, to be vocal, and to question other people's views.
The challenge to the more senior members is to listen, because they may think they know everything. When they listen to a younger team member come up with a new idea, they may think, "Well, that doesn't sound right," then suddenly catch themselves. "Hold on a second. What was that?"
It's that sort of interaction I'm looking for. Having a diverse team really encourages that kind of thinking. I don't want clones of me; I would hate my job if I surrounded myself with people just like me. I've seen managers who build teams that look and walk and talk just like them.
I want a team that challenges me, a team that pushes me, and a team that can do things without me there in the middle saying, "Look at me! I'm the leader!" They can get on with it and do a job that is effective and efficient.
Read Part 2 of our interview here.
Linda L. Briggs is a contributing editor to Upside. She has covered the intersection of business and technology for over 20 years, including focuses on education, data and analytics, and small business. You can contact her at firstname.lastname@example.org.