Effective Teams and Managers: What Google Has Learned
As the director of People Analytics at Google for the last decade, Brian Welle’s world revolves around data. He has found that those hard, cold numbers can, when used properly, uncover the key attributes that make people better managers and team members. Once identified, the attributes can then be cultivated and instilled to boost performance. Welle spoke recently with Cade Massey, Wharton practice professor of operations, information and decisions and co-director of the Wharton People Analytics Initiative, about applying data analytics in the office. Massey is co-host of the Wharton Moneyball show on Wharton Business Radio on SiriusXM channel 111, and this interview was part of a special broadcast on SiriusXM for the Wharton People Analytics Conference.
Welle identified eight attributes that define a top-performing manager: a good coach; empowers the team and does not micromanage; expresses interest in and concern for team members’ success and personal well-being; productive and results-oriented; good communicator — listens and shares information; helps with career development; has a clear vision and strategy for the team; and possesses key technical skills that help him or her advise the team.
An edited transcript of the conversation follows.
Cade Massey: You have a Ph.D. in industrial and organizational psychology from New York University and were a postdoctoral fellow at Harvard’s Kennedy School of Government. How did you end up at Google?
Brian Welle: I did the post-doc because I thought I might want to be a professor. Then I saw how hard professors actually have to work, and I decided maybe that wasn’t for me. This role I applied for in 2006 could not be described to me by my interviewer. He basically said, “We’ve got a lot of data and a lot of challenges on the people side, and we need smart people who know how to do research to help us figure it out.” That was the lack of structure I was looking for at the time.
Massey: You told a story about a scene from Star Trek, which you felt perfectly captured what you wanted to do in life. I would love to hear that story again.
Welle: I honestly believe that I am an IO psychologist today because I watched Star Trek: The Next Generation when I was in high school. There was one character that really stood out to me across the whole show. Oddly, when I ask people about Star Trek: The Next Generation now, that is the one character they never really think about. But it was Deanna Troi, and she was the ship’s counselor. She was the one who negotiated cross-cultural agreements, she was the advisor to the captain, she did the crew rosters, and she was the only person on the bridge that did not have to wear a regulation uniform. I thought, how could I be the ship’s counselor to an organization? I sort of feel like that is what my role is at Google.
Massey: How would you describe what you do at Google, and where are the connections between that and the Star Trek character?
Welle: Deanna Troi really had to understand why people were behaving the way that they did, and she had to sort of impose a logical explanation for what was going on so she could make those interactions better and save the crew and all sorts of other fun things. The amazing thing about organizations is, as human beings we all know how to do the basics. We know how to interact with each other. We can get through the day. We can work in teams relatively productively, but there is so much room to optimize all of that. If you can take a step back and understand the dynamics, you can actually help the system work better. You can give people the tools they need to self-regulate. You can put rules in place when you need to. There are so many opportunities for efficiency that a lot of organizations don’t even know they need, because they’re people who say, “I know how to do all of this stuff because I’ve been doing it my entire life.”
Massey: I know that Google is not a very rules-happy place. In fact, they’re anti-rules. Can you give us an example where you realized you needed to put a rule in place?
Welle: I think a good example is our approach to management. Hierarchy is like a necessary evil in most organizations. Of course, we had managers at Google, but we did not provide them with a lot of structure. We said, “We want your teams to be productive. We want you to do good work. We will not stifle you with rules on how to get that done.”
We were getting some signals that managers on the whole were not performing or at least not providing the experience to the people on their teams you would want to have when you come to work. We did research on what differentiates the best managers from the not-so-great managers and came up with eight attributes of success — everything ranging from having consideration for people as people to providing coaching and career development advice. We codified that set of eight, and we give every Google employee the chance to rate their manager on those eight attributes, and we provide them with the feedback. That is an example of a set of eight rules that is empirically based. We know it drives good outcomes, and if you can prove that a set of rules is leading to better outcomes, people will listen to that. It just needs to be based on data and evidence.
Massey: You’re very well known outside of Google for unconscious biases and work on diversity. You’ve been working on judgment biases within Google for probably 10 years. I feel like teams are a place where analytics hasn’t made much progress, so can we hear more about this teams research?
Welle: I think that there are two really big, important topics within organizations that are incredibly difficult to study. One is leadership and the other is teams.
Part of the complexity is that we all feel as human beings, we work in teams, and we look to leaders, so we feel like we intuitively understand it. Yet when you try to actually study it, you realize they’re messy. The rules are difficult to determine, so we leave a lot of it unstructured. We didn’t want to let teamwork just happen at Google because we thought there might be an opportunity to help teams be more effective and to create a better team experience.
There are a lot of requests coming from engineers who are doing work. Our engineering population tends to be very skeptical about the status quo anyway, so we’re asking, “What do we know? You have so much data about us. What can you tell us about what will help us work better together?” That was the genesis of the work. What we were expecting to find is that we would have a long list of individual characteristics that would help us determine the composition of an effective team.
You would want a 10-year mix. You want to make sure you have gender diversity, you want extroverts, you want introverts, you want highly conscientious people — you want that whole mix. If we could quantify that mix, we could come up with the ideal recipe for an effective team. What we found instead was that those individual factors did not predict team success. What predicted team success is what the team members themselves were actually doing in their interactions with each other. What kind of environment were they creating? How much could they depend on each other? How clear were their goals?
You can absolutely measure those things. Psychologists have been measuring them but didn’t necessarily put them together in one study. Now that we know that, we can distribute those measures to teams so they can take a self-diagnostic and at least have the language and the structure to make improvements to their team process.
Massey: What does this look like in action?
Welle: We created diagnostics. We tested it to make sure that it was valid, internally consistent, had a good user experience. We created an online survey. If you are a member of a team, you can input the names and the email addresses of all of your team members into the system. If everyone agrees that they are ready for this experience, the surveys get sent out, people fill it out and it gets aggregated. Each team gets a report on each of these dynamics that we found to be important, and that’s the starting place for the conversation.
Massey: How do teams react to that? I would think in most organizations, people are sensitive about giving critical feedback in that kind of setting.
Welle: It’s usually perceived very well. First of all, it is instigated by the team, so they want the feedback. Usually what the diagnostic does is gives you information you sort of felt already about the team, but it helps you channel that. If the one place where your team scores extremely low is on psychological safety, then providing you with that feedback can say, “It says psychologically unsafe. You people figure it out.” We usually need a trained facilitator to come in and broker that particular conversation.
Massey: One last observation I want to hear is a very practical tip for those working remotely. I hear you and Elizabeth McCune at Microsoft talk about what it means to work away from headquarters. You say you have a recipe for what makes this work. What’s your recipe?
Welle: I worked in headquarters for two years before I left, which allowed me to create really good, robust social network. You have to have the skill to collaborate in meetings over video conference, and that’s something you can learn over time. What’s great about our team’s research is one of the things I was afraid that we would find is that the distributed nature of team would negatively impact effectiveness, and they would force me to move back to headquarters. Fortunately, no correlation.
Massey: What are working on that what can we anticipate hearing about from you two years from now?
Welle: You might be tired of us hearing about unconscious bias and the diversity narrative, but we started this journey a few years ago with a concerted investment in unconscious bias training and other diversity-related activities. It has just skyrocketed since then. We have made some public declarations about diversity and where we want the tech industry to go. I sincerely hope that is what you’re going to hear us talk more about in the next six months, year, two years.