David Stenglein

My name's Dave Stenglein, and I am a solo consultant in the New York area. I focus on platforms and human factors.

Erin Doyle

And I'm Erin Doyle. I'm a staff engineer at Lob.

Conversation

Dave: So how we got here — I met Erin when we were both speakers at QCon SF last fall. And she had a really interesting talk about an outage that happened and sort of what it meant to be a staff engineer during that. But I saw other possibilities around it with psychological safety.

Erin: Yeah, when we discussed, it really became apparent much later that so many points throughout that experience that I had really tied directly to a lack of psychological safety. So it's kind of become a passion of ours that we like to talk about — improving culture and psychological safety of our workplaces. So that's why we're here together today.

So we're not here to talk about toxic workplaces, bosses who throw things or anything like that. We're here to talk about the subtle impacts with psychological safety, and how you might be creating an environment that lacks that kind of comfort unintentionally.

So we should probably define psychological safety, even though it's a fairly common term. Erin, what does psychological safety mean to you?

Erin: To me, psychological safety is the comfort to be open and share your thoughts, ideas, questions, concerns, mistakes, and experiences with others without fear of negative consequences such as judgment or punishment. And with a trust that what you've shared will be heard and treated with respect, even if others may disagree.

Dave: So if we're gonna talk about psychological safety, we really need to bring up Amy Edmondson and all the research that she did starting back in the nineties. And she coined the term "team psychological safety" after the term came about in the sixties of just "psychological safety." And Google published some research around 2012 that they did in trying to figure out what makes a high-performing team. And they found that the single biggest factor in making a high-performing team was psychological safety. And that kind of kicked off the public knowledge of this.

So we're gonna do a little bit of some of our experiences having to do with psychological safety, a few different aspects of it. And in order to start that off, there's kind of two different ways. So there's two kinds of dangers that you experience when you don't have psychological safety. One is the risk of loss. You could cause something bad to happen by not having psychological safety, and the downstream effects could be potentially life-threatening. The other is a missed opportunity, or an opportunity cost. If you do not have psychological safety, you may not achieve high-performing teams, and therefore you don't get as much as you could in terms of business outcomes.

So for a first one — Erin, it's always really been neat that you worked at Kennedy Space Center, and you were also there during the Columbia disaster. So could you tell us what it was like being there when that happened?

Erin: Yeah. So over the seven years that I was at the Kennedy Space Center, both before and after the accident, we were told that safety was to be our top priority in everything we did. There were safety posters everywhere. We were constantly taking trainings, we had procedures to follow. We were all very well aware that lives and billions of dollars were at risk.

But underlying all of that was this really palpable, strong pressure constantly to stay on schedule and under budget. And so it really didn't feel like safety was our top priority day-to-day. It really felt like if one were to raise concerns about safety, they wouldn't be taken seriously, especially if they put schedule or budget at risk.

Dave: So the basics of the disaster are known — that a piece of foam came off of a booster and impacted the wing of the shuttle, and that's what caused it to break up. You may not have known exactly what happened behind the scenes at the time, but through research after the fact we found out. What did you find out was going on that led to this?

Erin: Yeah, so over the course of the mission, as you would expect, numerous emergency meetings and sessions were taking place to try to assess the impact of the damage and to come up with an appropriate contingency plan. And during that time, a number of people may have had concerns, ideas, suggestions, that just didn't feel comfortable speaking up. And those that did speak up — their concerns were downplayed, they were diminished, they may have been ignored altogether.

And so it really was a lack of proper testing, proper contingency planning, and this really strong reluctance to alter the mission schedule, that sort of resulted in a "it'll probably be fine, let's hope for the best in landing" approach — which now we know resulted in catastrophic loss.

Dave: So since that was such a big factor, did it come up in the official report?

Erin: Yeah. So when the Accident Investigation Board published their findings — as they were doing their root-cause analysis, for every technical issue that contributed to the accident, they were able to find and tie together a lack of psychological safety that failed to prevent a technical issue. And so really decades of this schedule and budget pressure, and defining success for the organization based on schedule and budget, permeated the organization with a lack of psychological safety.

Dave: So given that, and this fairly public and well-read report, did they do anything about the situation?

Erin: So yeah, when the Investigation Board published the report with the recommendations, my company came out with this Timeout Policy. They printed out these cards, and we all were to put them in our badge holders. And we were told, anytime anyone feels uncomfortable about the safety of an operation, they could pull out their card, they could call a timeout, they could halt the activity, and give everyone an opportunity to discuss before things proceeded.

But despite all of that, that schedule and budget pressure remained so strong that people still didn't necessarily feel comfortable speaking up. And so what this illustrates is — just because you put a process in place and you tell people they should feel comfortable, doesn't mean they automatically will. We really have to prove to them that if they speak up, they will be safe and they will be heard.

Dave: So that was a somewhat extreme example of the effects of psychological safety. More relevant maybe to people here is the talk that you gave, and the incident that happened, and how that played out. Can you tell us?

Erin: Yeah. So this incident that I spoke about was the most severe that I've been a part of in my career. It took us three days from discovery to final resolution, with a number of follow-up tasks that sort of trailed off after that.

And when zooming out and looking at it, pretty much every misstep we made along the way was due to a lack of psychological safety. From the initial commit that caused the incident — was done by an engineer working in an area of the code that they weren't familiar with, and they didn't really understand the impact of their changes, but they didn't share that knowledge with anybody. They didn't speak up, they didn't ask for help.

There were two engineers that approved the PR — I was actually one of them. And honestly, neither of us fully understood the change either. But for differing reasons, we each didn't feel comfortable speaking up and saying, "hey, I need another set of eyes, I don't actually understand this." And so the change was merged, deployed, caused an outage.

And from that point on, throughout our entire incident response, we had multiple opportunities where maybe someone needed help and they weren't getting it. Someone had concerns, someone had problems, someone saw something we could have been doing differently — and nobody spoke up. A lot of it was, we just didn't wanna hurt anybody's feelings. We didn't wanna sound like we were criticizing someone who we knew was doing their best — but we could see something they could have been doing differently.

And so because no one spoke up, we missed multiple opportunities here. We could have prevented the incident in the first place, and we had multiple chances to resolve it much more quickly, just if someone had felt safe enough to speak up.

Dave: So we talked earlier about two dangers — the risk of loss if you don't have psychological safety, and the risk of missed opportunities. So Dave, you've got experience with the missed opportunities — do you wanna share that with us?

Dave: Sure. So this is from a point earlier in my career when I became a leader, team lead. And I was in a consulting role where we were going to build a platform for a customer. And I had some experience with this and had done it before.

And so I got our fairly good-sized team together to figure out — kick this off and begin implementation, set the principles out — and I felt I'd communicated really clearly about how we should implement what we needed to do. And from that point forward, I was frustrated a lot. I felt with the fact that the team never seemed to be quite doing what I expected. And the program was successful, but it was a real slog to get to the end of it and get everything delivered. We kept seeming to go back and be dealing with miscommunication.

And fairly long after the fact I realized that I had been pretty forceful in saying what had to be done and how it should be done — and this very strong opinion came off as, "well, I am not going to take any challenge to that." And what it led to was the team not taking ownership and not challenging me on this design that I had in my head — that could have been improved upon. And this ultimately led to us delivering something that was maybe not as good as it could have been.

And so I lost the chance to get my ideas improved upon by others, because I kind of unintentionally made it clear that I wasn't interested in people's opinions.

Dave: So we've given you some examples of what can happen when you're lacking psychological safety in your environment. So let's talk about what you can do about it.

So first you need to actually recognize the problem. A lot of people think they have a safe environment, but maybe they don't actually. And there are plenty of warning signs that you could look for. So Erin, what are some of those warning signs?

Erin: So — is your team quiet? Are you hearing about questions, mistakes, errors, or failures? It could be that things are going smoothly and everything's fine. But it also could be that maybe people don't feel comfortable speaking up. Maybe someone's run into an unexpected problem or someone needs help and they just don't feel comfortable sharing — maybe they're afraid they're gonna be judged or punished for failing.

Dave: And failure is a good thing to bring back to Amy Edmondson, because when she did her original research, she was looking at high-performing teams in high-risk environments. So nursing teams that were doing things like giving patients medication doses — that if you screw up you have a failure, you could have life-threatening consequences. And they found, almost counterintuitively, that high-performance teams had higher failure rates. And to the point where they were scratching their heads — she went back, she kept going through the data, they had lots of statistical data, and it didn't make sense. And so they sent someone in to do some observation of these teams, and found high-performing teams were willing to admit that they'd had the failures. Low-performing teams were not. So the failures were the same for both kinds of teams, but the high-performing teams appeared to have higher failures. So one of the warning signs could be: you don't have enough failures in your environment. People might be having failures, but they're just not being reported.

So one of the other warning signs is perfectionism — that inability to ever be finished. And Erin, you have some experience with that?

Erin: Yeah, I struggled severely with perfectionism earlier on in my career. I really believed that others wouldn't see me as good enough if I wasn't perfect. I wanted people to think I had all the answers. I knew how to do everything, and I never made mistakes. So of course, I couldn't ask questions or ask for help when I needed it.

And so the problem with that — beyond the detrimental effect it had on my own mental health — is that I was creating an unhealthy environment around me. I was leading my team to believe that they had to meet this unfair and unreal standard that I had set. And so they stopped asking questions, they stopped asking for help. We all sort of suffered alone in silence, figuring things out on our own. And we learned much more slowly, and it took us a lot longer to get our work done.

Erin: So how does it happen?

Dave: Yeah. So we thought we had psychological safety. Turns out that maybe we don't have what we thought we did. How did we get here?

So part of it is subtle things. It could be things you do, like maybe that reflexive response to ideas — where you're maybe just automatically critical, instead of just listening. Or it could be things you don't do. Maybe you don't support ideas of others, even if you're not critical, or you don't speak up. And you might not invite discussion — that could bring people out and again put things out there for collaboration to happen.

Leaders don't necessarily create this intentionally. A lot of people think they have a good, safe environment. But we all have our blind side — the things that we communicate that we're not really aware of. So that's important to keep track of, because the people who are receiving these messages and these cues from us — they aren't necessarily being — don't think of themselves as not having psychological safety, or being unsafe or fear. But they might hesitate the next time they should contribute something. And that can kind of build over time — that subconscious, "I don't feel quite right here, I'm not sure why, but I'm not gonna bring this thing up."

So Erin, this is probably a good place to talk about the Swiss cheese model.

Erin: Yeah, let's pivot for a moment. Talk about the Swiss cheese model of accident causation, which is used heavily in risk and root-cause analysis.

By standing up a number of slices of cheese, we can visualize a scenario where each slice of cheese represents a layer of defense — maybe a process or mechanism that should prevent some negative outcome. And due to the randomly-sized and -positioned holes, we can see how it's possible for those holes to just perfectly align and allow some negative outcome to get through all of those layers of protection.

And so when we're doing root-cause analysis, we really need to iterate over asking why for every trigger in that chain of events that led to our outcome. And that's how we can identify what are those holes in our defense layers — how did this happen. And what we usually find is that it was multiple contributing factors that led to that negative event.

Dave: So it seems like psychological safety could really help us out here.

Erin: Yeah. So psychological safety provides us with a really successful multi-layered defense system against any number of detrimental outcomes in software and systems. And when we don't have psychological safety, it's what erodes these holes in our defense layers against negative outcomes.

Erin: So now, what can we do about this? What can we start doing differently?

Dave: First, we need to have a basis of vulnerability. If you can't be vulnerable, you can't really have psychological safety. But what is the definition of vulnerability? If you look it up, it says that it's "leaving yourself open to emotional or physical attack." And that sounds pretty extreme, and that's not what we're talking about.

Erin: Yeah. Vulnerability is really that openness to share your thoughts, ideas, feelings, experiences with others. It's to feel open being your true, authentic, imperfect-at-times self with others.

Dave: So what can we do about it? These are things you can take and be actionable with.

So first, be intentional. This happens because we're not thinking about it, or we make assumptions. So actually look for warning signs. See if people are too quiet. See if it seems like your outcomes don't seem to match what's actually happening — that you can see visually with failure and such.

Be self-aware. See how you interact with your team, and how you're coming across, what you're communicating. And if you take a step back, is it not really what you thought?

And then really importantly, reframe your thinking about failure. Failure is an opportunity to learn — as long as you're open and vulnerable about the possibility of failure.

Erin: We need to model. Humans learn by watching others, and leaders can really set the tone in their environment through their own actions. So we need to model and demonstrate the behavior we want to encourage in others, to normalize vulnerability. We need to be vulnerable. We need to demonstrate that if we feel comfortable sharing our thoughts, our ideas, our feelings — whatever — that others can feel safe and comfortable to share as well.

We need to use more inclusive language. We don't wanna speak with too much assertion, because we wanna leave space for others to contribute to the conversation. So using a little more softening language — like "I think," "I believe," "maybe," "might," "please correct me if I'm wrong" — really gives space for others to weigh in.

And finally, invite discourse. We need to collaborate. And so you may need to ask for that and invite it. Ask: "Are there any questions? Do you have feedback for me? What did I miss? What else is there?" Really ask for that. And when we collaborate on solutions with our different perspectives and experiences, we always end up with better outcomes.

Dave: And then we need to reinforce. So this is like everything about change — it's a continuous process. And you might benefit from training. Leaders might need training to become more aware of how they interact with their teams. And teams might need training to understand the value of this and the things that they might need to do and change when working with each other.

You need to gather feedback. You need to find out — part of finding out if there's a problem is just asking, and doing sensing and those sorts of things. And while you're going through the process, you can tell if you're accomplishing the goals that you believe you are. And then you can iterate. You implement an OODA loop — observe, orient, decide, act. Figure out if you're where you want to be with psychological safety, and figure out what to do next.

And so this is what we'd like you to take away from here: Psychological safety is a key to high performance and high-performing teams.

We've gotta reframe failure. We need to approach failure with positivity and openness. It's a learning tool. It's an opportunity. The sooner we can discover the failure, learn from it, the sooner we can get better and iterate.

Minor actions or inactions can harm psychological safety. It doesn't have to be intentional, it doesn't have to be toxic. This could be happening even though you think that you have a great environment.

And it's important that — we're not talking about making a place nice. We're talking about making a place comfortable for people to take risks.

And finally, leaders — we've got to model. We've got to go first. We've got to show that if we feel comfortable demonstrating a certain behavior, then others can be safe and comfortable doing so as well. So we need to prove to them with our own actions that we can all be comfortable being our true, authentic, imperfect human selves with each other.

Erin & Dave: And thank you. Thank you. You can leave.