Psychological Safety: The Foundation to Build Your Engineering Team’s Success - Interview with Dan Rummel (Sr. Director of Engineering, One Medical)
Sometimes, teamwork gets tiring, communication gets weird, fresh ideas are rare and far between, and people are more likely to jump down each other’s throats. Sound familiar?
Building psychological safety could help you fix all that.
We sat down for an interview with Dan Rummel. He is currently the Senior Director of Engineering at One Medical, and he’s held engineer and engineering leadership positions at various Bay Area startups. Basically, he has tons of experience from the last one and a half decades.
He shares this invaluable insight with you on creating a workplace with high psychological safety, so you get to understand his way of thinking and pick up actionable tips you can apply at your company.
We’re covering:
What is psychological safety and why is it important in a developer team?
Ultimately, it's about whether individuals in a team feel safe enough to speak their mind, share their ideas and take risks without fearing traumatic impacts to their career or interpersonal relationships with other teammates. When you have safe teams, people can really bring their full selves to work.
There's a couple of ways to look at it, first from the individual side.
When you’re working in safe teams, it just feels good. You can show up and be yourself, express your thoughts and have your ideas truly heard. A shared belief, and the minimal interpersonal risk it facilitates are cornerstones of a high performing team.
There isn’t much fulfillment in going to work and keeping your mouth shut, or in being a cog in a big machine. If you're in a creative role, like software engineering or design, etc., you have to think outside the box to put the pieces together or to invent new ones.
Research shows, if you're in an environment without psychological safety, you're not comfortable bringing those creative ideas to the table.
The entire organization does better when people come up with new ideas, get productive and move the company metrics the right way.
A psychologically safe team is not the silver bullet to fix your team's or organization's problems. There are many other things you have to get right, but I think it’s one of the cornerstones for highly productive teams.
How to tell the difference between psychological safety and the comfort zone?
Brené Brown makes a great point about this. She uses the term “rumbling with vulnerability.” It means you’re not simply trying to make everyone feel good about themselves, but instead, you are building trust so everyone can really challenge each other and debate in a healthy way. I think she uses the term “rumbling” to get us to think differently about what a healthy debate is.
Psychological safety is the basis to enable the interpersonal trust necessary for this. Again, you need the other pieces of the puzzle as well, like dependability, clarity, and purpose. Otherwise, teams won’t be motivated enough, nor will they be able to drive things in the right direction.
If you focus only on psychological safety, you may fall into that comfort zone where things feel soft and squishy. But sometimes, getting things done and making an impact isn't about being soft and squishy. You have to dive in, find the rough edges and have the hard conversations to create the polished gems that have true value.
The difference is to not just settle for everyone having a smile on their face.
What are the must-have requirements for psychological safety?
1. Transparency
People look at an engineering manager or a leader like they have all the secret tricks in their back pocket—things they know, but no one else does. That’s counterproductive.
The first thing is transparency. Saying, “I really believe psychological safety is important. It’s everyone’s responsibility to promote this on our teams, and here is why.” So, you're up-front about it, making this a team value, and a team effort to build as a part of your company values.
2. Build the inner communications network
Once people understand what this is, and why it matters, phase #2 is creating a voice for your team. Make space for people to share opinions with their peers and managers by creating channels for individuals to talk to leaders via 1:1s and for individuals to talk to their teammates via retrospectives.
3. Equal talk time
The next step is creating equal talk time. Embrace the diversity of personality types in your group, and make sure managers are tuned to it. It needs to be transparent; you’re trying to get everyone to participate in this practice of communicating openly with one another.
4. Make room for mistakes
The final step is to make room for mistakes. Build a culture where if things don't go as planned, it’s part of the game. You have ways to learn from it and to treat it as an opportunity to improve. It beats calling someone out, only to dish out some punishment.
Learn the ins and outs of Monzo's renowned engineering culture! [Podcast + Blog post]
How did you build a psychologically safe environment for your engineering teams?
Give ownership of the processes
Some of the key components are building a foundation of education and championing ownership of this across the organization. As a leader, you should openly talk about it; send around articles and different ways people can learn about it.
The other part is that you should empower continued improvement for your people. Give them ownership of how they're doing things, and give them permission to change it, so they get better at it. It’s far more effective than just dictating exactly how they have to do everything.
Invest in 1 on 1s
The next step is creating the voice. I really believe in 1 on 1s. We invest a lot into them at One Medical. They always seem to be the easiest to reschedule and delay, but consistent 1 on 1s with your team members are key to building your own voice. It gives people a chance to talk to their manager, as it's their time to communicate to you, not the other way around.
It’s about their agenda, so take the time. Take a break, recognize you’ve been working hard for a week or two, analyze what went well or poorly. Then, you’ll see what you should double down on, or what questions you should be asking.
There are countless different types of retrospectives, so you should find which work for your team, and use them consistently.
It may seem like it doesn’t have a lot of value, because it doesn't have a direct impact on the customer, but it’s really about your acceleration measure. How quickly can you solve the problems you face and uncover and resolve frustrations? All these frictions slow you down and make things weird.
Build a blameless workplace culture
Make room for mistakes by having a blameless culture. I intentionally use the word “mistake” and not failure. I think we overuse “failure,” and in organizations, you don't want full failure, when everything comes crashing down. Shutting the company down is a failure.
Doing post-mortems, or as some people call them instant response retrospectives, help a lot. When something goes wrong at a blameless structure, you get a group together and just talk it through. It's not about pointing fingers, negative consequences to whoever may be at blame, but bringing curiosity to the table, to learn how you got to this, what you missed, and how can you protect yourselves in the future.
So, build safety nets for people to take risks in the future.
Some of these are practical safety nets, like being able to roll back infrastructure quickly.
Others are psychological safety nets that turn mistakes into learning opportunities for the organization. Having a great post-mortem process can be a cathartic experience for somebody who accidentally, let’s say, brought your service down. .
For example, I fat-fingered something, or I pushed a bad piece of code. It can be tough to deal with, and it’s useful to have a process you can point to and say, “This is how we deal with these things.” It makes people more comfortable to get through that and to continue to push hard in the future.
In my experience, it increases employee engagement, even though often when things go wrong, you see people drift away and disconnect. These are some key things you need to invest in, maintain and not sacrifice if you want to maintain high psychological safety and high team performance.
Lead by example
Leading by example is often neglected. It’s a real challenge for engineering leaders and engineering managers because their typical personality types can take up a lot of space in the room. I think leading by example means making space for others.
We need to actively make space for others to chime in during meetings. It also helps a lot to show vulnerability, talk about some of your challenges, or throw out the occasional wild or even silly idea. Even if you know it will likely be shot down, this lets your team know you’re fair game for debate as well.
Perhaps the most important thing to do is admitting mistakes or if you're wrong about something, and not make it a big deal. Just saying, “I was wrong; you've got it,” helps the healthy debate and makes it comfortable to rumble through ideas and to find the best outcomes.
The reason healthy teams produce better outcomes is that there's a diversity of experiences and perspectives that come to the table. When people can build mental models with everyone else's experiences, it facilitates organizational learning, your outcomes get exponentially better, and you get beautiful results from those debates.
How often do you have 1 on 1s in your engineering team?
The frequency of 1 on 1 meetings and skip level meetings depend on the relationship and the reports. It's hard in a growing organization, as people are added in and are instantly overloaded. I do it weekly when it's a really important relationship and biweekly with others. There are lots of ways to do it.
Consistency is the most important, so people can count on it. If you have a bad day, and you want to discuss it with your manager; you might be stressed and distracted if you are unsure when you will have time to talk. However, if you already know you’ll have another meeting in a week, it helps you put the issue aside and focus. It's not ideal, but knowing it's coming gives people confidence. The same goes for team retrospectives.
Even when the shipping schedule is tight, and the team has work to do, make sure your scheduled 1:1s happen. When people know you’ll take the time to talk to them, it lets them put the issues aside and focus on the work at hand to remain effective teams. All in all, I think cadence is less important; it’s more about the reliability of that communication channel.
How do you create a psychologically safe workplace as an engineering leader?
If you feel like the organization is lacking in it, you have to talk about it all the time. Set up anchor points in your weekly and daily practices to push it forward. It's been fun to see as we've built our teams at One Medical. Once people really understood, it became ingrained in the culture, and it only takes a small effort to keep the flywheel spinning.
Think about where the organization is. If you feel like people are disengaged, they can’t communicate and no one brings new ideas to the table. In that case, I’d be pushing it all the time. I’d be talking about continuous improvement, psychological safety, and equal talk time, and I’d probably do more 1 on 1s.
We have a practice where every other week, we have our teams tell us about what kind of continuous improvement they are investing in currently. Asking, “What have you learned?” “What worked, and what didn't?” We all learn from their process improvements.
On the one hand, we can share that in the broader organization; on the other, it lets people know this practice is important for us. It's all about dialing for the amount of change you want to make. Once people feel the importance of this and the productivity you gain, it becomes natural, and it gets easy to maintain.
We need to find ways to keep teaching as new people enter the organization, or as some teams lose track, but it builds momentum and keeps itself going eventually.
When the team psychological safety is high, they are very productive. Meetings where you try to create this voice might be the first thing you'd want to sacrifice, as they don’t appear to produce value for the customer. It's a false intuition to think that if you cut it, continuing to talk and iterate means you’ll be more productive. The faster you are traveling in a certain direction, the more important it is to stop and adjust your course; otherwise, you can easily end up way off the mark.
It’s all about maintaining the discipline, rather than falling into the false feeling that you need to cram core work into every part of your workweek. Work smarter, not harder.
What You Should Do Now
👉 If you are serious about becoming a great engineering manager, you should subscribe to our newsletter or download the state of software development 2020 report.
🚀 If you need developer help for your project then click here for a FREE consultation.
About the author:
Karolina Toth has been working with engineering managers for over 5 years. She is an internal coach, working on the most pressing management-related issues tech companies face. She is the host of the Level-up Engineering Management podcast where she talks with accomplished tech leaders of fast-growing tech companies.