Using Data to Foster Psychological Safety [Webinar]

Psychological safety is a hot topic across all industries these days, but in the fast-paced and dynamic world of software engineering, it plays an especially important role in driving innovation and high performance. Yet, what exactly is psychological safety? Why is it so important for software engineering teams in particular, and how can it be fostered?  

To gain real-world insight into this complex topic, Code Climate’s SVP, Customer Organization, Karthik Chandrashekar sat down with Anjuan Simmons, Engineering Coach at Help Scout, Heidi Waterhouse, Principal Developer Advocate at LaunchDarkly, and Lisa Van Gelder, VP of Engineering at Avvir.io, to discuss the importance of safety, the strategies they use to implement it, and how data helps them achieve it. Here are some highlights from their virtual roundtable. 

Psychological Safety Heightens Speed and Performance 

Lisa: [Psychological safety] gives engineers the ability to fully participate in a team. To give their ideas, to debate ideas, to feel free to take risks, to experiment. And if people don’t feel free to do those things, everything slows down. 

Anjuan: I boil psychological safety down to two things, truth and trust. People can live their truth – the truth of their experiences, their history, even the truth of their performance.

Heidi: When I come at psychological safety, I frequently come at it from the sort of scientific side, realizing how much of a difference it makes to the performance of a team. The faster a team moves, the more psychologically safe they are, the more they feel like they can take risks, but also the more they feel like they can take risks, the faster they move.

Learning Your Team’s Lived Experiences Helps Cultivate Safety 

Anjuan: I think that leaders have to establish authentic relationships with their directs. One of the easy tactical things to do is to have weekly 1:1s. If you’re a manager and have people who you are responsible for, if you’re not meeting with them weekly in a 1:1, then you’re not engaged in a psychologically safe environment, because you just don’t even know where you are. 

[During 1:1s] talk about more than just status or work. Discuss their career and personal goals and see what you can do to be supportive. By having those 1:1s, you begin to get an idea of their background and their profile. You’ll probably find that a number of the people who you work with and who are your directs have gone through trauma brought on by work, especially if they are members of marginalized groups. Knowing [someone’s history] is really important to understanding what team members may not be feeling psychologically safe and steps that you can take to help with that. I’ve never found a one size fits all approach to this. You have to tailor it to the individual people who you’re working with. 

Heidi: The more senior we get, the more trauma we have. To this day I have a fear reaction when I’m one-on-one with my manager and they close the door. There’s nothing wrong with closing the door to give me feedback. That is in fact the responsible thing for a manager to do. But I got fired enough at crucial times in my career that I have a fear reaction. And so if I tell my manager [about my fear] they know that, and they preface the conversation with “nothing bad, just a little feedback,” and then I can relax and actually take in the feedback. 

I think the idea of moving through the world as a trauma-informed person enables you to be considerate of the scars people are carrying around, becoming a force multiplier for your ability to be a good friend and manager, and to get the best out of the people that you’re working with. 

Lisa:  It’s not always obvious if your team is having issues with psychological safety. Ideally, they will tell you, but quite often, especially if you are a new leader on a team, they don’t trust you enough to let you in and tell you what’s going on. You kind of have to go into debugging teams mode, observe them, and see what happens. See if there are strange things that they’re doing that you don’t quite understand, because often there’s a safety issue underneath that. 

I once had a team that used to hide around the office on Fridays. They weren’t answering emails, they weren’t answering Slacks. They wouldn’t tell me what was going on either. I did try and ask, but I also sat there the rest of the week with them to try and see what was happening. I realized that people were trying to give them work on Fridays and then they would blame them if it wasn’t done by Monday. On the outside it looked like the team was slacking off, when in reality, it was a safety issue. So when I say look for odd things you may have to do a little bit of detective work to see if something’s happening. 

Put Metrics in Context 

Lisa: When you think about introducing data or metrics onto your team, I encourage you to think about what those metrics are nudging the teams to do in order to meet those metrics, because it’s really easy to accidentally introduce a metric that would harm the psychological safety of your team. 

A lot of the time when I come onto a new team, people will say the engineering team is moving too slow, that they want them to go faster and they want to measure this through the velocity of the team. However, it comes with connotations that the team isn’t working hard enough, and can impact the quality of work being delivered. 

A metric I love to use instead is Cycle Time. I love it because it’s a full-team metric. [For example], if it takes an engineer two days to complete a story and seven days to deploy it, you have a better idea of where to start digging to spot areas that can be optimized. Cycle Time encourages you to find the bottlenecks in your system as a team and address them. It also encourages you to break stories into smaller pieces because the smaller your stories, the quicker you can get them deployed. They’re good nudges that help a team work together to solve problems. 

Heidi: Metrics are useful for building safety and consideration around your team while still figuring out where your gaps and your problems are in a blameless way. Saying like, okay, here’s where things are sticking. How do we unstick them? Is that actually the first cause? 

Anjuan: As an engineering leader who’s often called to figure out how to measure my teams and how to support them in getting better, there are a lot of metrics that are well-meaning, but are often counterproductive, like how many lines of code people do each week. 

Some languages are simply wordier than others. So you can actually end up penalizing people working in languages where you may need 50 lines in one language but can do it in three in the other. It can also drive developers to add more crud to the codebase because they’re trying to gain in number of lines of code. 

So you really need to take a step back and say “what are we measuring? What metrics are we trying to use?” to see if they are serving us well. And, “are they creating an environment where our engineers feel that they can inspect themselves, learn from their experiences, do their best work, or are we doing something where we’re actually harming them?”

Small Changes in Syntax can Improve the Environment 

Anjuan: As I mentioned before, you always want to hone in on the lived experiences of the people who come through the doors of your company, whether they’re physical or virtual. There’s so many things that you can do to help take our industry, which is still steeped in colonialism, misogyny, transphobia, and all these horrible things, to begin debugging your work. So for example, at Help Scout, we took the common scrum term “backlog grooming” and changed it to “backlog refinement” because the word “grooming” has negative connotations. That small change was a nod to the fact that we understand that some words can be triggering to some people. 

One other quick example: master/slave is used in weird ways in our industry when there’s better terminology out there. You can use “main,” for example. So again, if you look for these terms you can make changes to make a better environment. 

Heidi: I love the example of the master/slave terminology because it’s also a safety check for your team. If somebody fights you really hard on that, there’s someone to keep an eye on because they are not considering how other people on the team might feel about it. They are attached the way things were. 

Anjuan: We also talk about how, when people use the word “guys,” right? People in Slack will say, “Hey guys, I have a question.” Why are you assuming that everybody in the Slack channel are all male, right? Just [changing] little small things like that, you’re debugging your environment to be more inclusive, to be more open. 

To find out more about using data to help foster psychological safety on your team, contact an engineering data specialist.