Dan Rummel

@drummel

A Field Guide to Building Successful Teams: Part 2 Psychological Safety

Practical tools for developing psychological safety on your teams.

Image courtesy of Ricardo Gomez Angel

In part 1 of this series, I introduced the five dynamics of highly successful teams identified by a massive Google research project. There I discussed each and then the importance of starting a company wide conversation about these topics. In this part I will go deeper on the first of the five traits, ‘Psychological Safety.’ Researchers deemed it the most important dynamic of the five by far. They found that its absence made it impossible to form a truly effective team, after reading this I’m sure it will be clear why.

‘Psychological Safety’ is a group trait defined by Harvard Ph.D. and professor Amy Edmondson in 1999. A team with a high level of psychological safety means that its members feel comfortable taking interpersonal risks without the fear of negative professional or social consequences, (i.e., embarrassment or demotion of status). Achieving this dynamic has been shown to improve team learning and increase the efficacy of innovation of both product and process. I think many would claim that learning and innovating, together, form the core purpose of any startup.

In this post, I aim to paint a picture of an overarching framework that nurtures psychological safety within your team. Three major pillars stand out:

  1. Creating a voice
  2. Making room for failure
  3. Leading by example

For each, I’ll point you to key tools that can create the scaffolding for psychological safety. These tools are well discussed elsewhere, so I won’t go too deep on the implementation of each. As my friend Akshay says, “There will always be details.” I’ll connect the dots, and leave hammering out the details to you. From there you can hopefully get the hang of it and create an ethos of psychological safety in every corner of your organization.

Creating a Voice

Giving your team a voice is the first key ingredient to psychological safety. With groups of ambitious individuals, a strong voice will often manifest on its own. However, that doesn’t mean you shouldn’t intentionally build healthy channels and processes for that voice into your organization. If a voice is lacking or not providing much signal, these tools can absolutely help jumpstart the process.

Begin with high caliber one-on-ones to boost the voice of individuals. For those unfamiliar with one-on-ones, they are quite simple but can go a long way when done right. On the surface, they are simply regularly scheduled meetings with direct reports and peers with whom you have an important work relationship.

So, how do they help create a voice? First, they create a safe space for people to start talking. Second, they give you a chance to coach your team on participatory behaviors. A culture of effective one-on-ones is critical for a startup or team of any size. Even if your team is just a handful of people, I can still recommend making time for these.

Quality is vital. Ben Horowitz claims that building the communication architecture is probably the most important task of a leader, and good one-on-ones are essential to that. Remember communication is bidirectional, not just telling your story out. One-on-ones are the leading edge for information gathering, listening, and thereby a crucial part of your infrastructure. If you are not getting a good signal at the beginning of the funnel, the whole pipeline is pretty worthless. A prevalent phrase comes to mind, “garbage in, garbage out”. Teach yourself and your team to make them meaningful.

There is an abundance of methods describing how to have productive one-on-ones. Mark Rabkin even shows us how to have “awkward” one-on-ones to solicit consequential feedback. My advice is to embrace the diversity of techniques, find a format that jives with your style, but always have a few tricks up your sleeve. Different personality types on your team can often require different approaches. Some individuals might need more coaxing, and some individuals might provide information overload where you’ll need to distill the kernels of value.

Whatever your method, through the lens of building a voice, there are a few key things to keep in mind.

  • Keep them frequent and regular. Do everything you can to keep these commitments. This builds trust in the one-on-one as a valid communication channel. As issues or concerns arise throughout the work week, your team will know there is a time and place to discuss them. I like to have a Google Doc where we can add topics to the agenda at any moment. This allows both of us to safely put those issues down and out of mind, letting us focus on the work at hand for the day.
  • This time is primarily their time to talk. If you are doing the majority of the talking, then you are not doing very much listening. Ask the right questions and even force yourself to be silent to the point of awkwardness, (thanks Mark). Of course, it’s different with peer one-on-ones, time should be shared here, but still remember it’s about listening more than talking.
  • Reflect on the format and quality of the one-on-ones. Get meta. Talk to your team about the purpose of this time and get feedback on the effectiveness of this time. This will help instill ownership in the process and increase the effectiveness.
  • Don’t fix everything yourself. Let me say that again, don’t fix everything yourself. We all want to be superheroes for our team and solve problems as quickly as possible. However, the most impressive teams solve their own problems. Instead of coming to the rescue, this is where you put on your coaching hat. Look for any opportunity to encourage individuals to speak up and find solutions. It could be having a candid discussion with a peer, presenting an idea that would create tremendous value or bring up a feature that might be going sideways. Give them advice on how and when to communicate, affirm your support and help them shine when they do step up to the plate.

Next up, nurture a constructive team voice with retrospectives. I believe retrospectives are the most impactful and versatile tool in your playbook. As with one-on-ones, there are many many forms of retrospectives. I tend to gravitate to the lightweight and light hearted methods like Mad/Sad/Glads, but feel free to peruse this exhaustive list and pick favorites. Hold them at the end of your work week, sprint cycles or whatever breaking points make the most sense for your teams. Here are things to focus on:

  • Keep them frequent and regular (sound familiar?). Again just like one-on-ones, keeping these consistent will make them a reliable venue for discussion and will keep your team focused. One trick I love is to increase the frequency of retrospectives for new teams or teams that seem to be struggling. This forces more communication and tightens the feedback loop, increasing the rate of improvements. When retrospectives start to feel like they are providing less value, stretch out the cadence. Your team has developed more shared principles, established trust and solved process problems and begun to norm. (For extra credit, reading check out Tuckman’s stages for group development).
  • Encourage everyone to contribute. A significant finding in Google’s research was that the most effective teams shared a very distinct trait. When observing meetings, everyone on the team had roughly equal talk time. This is an opportunity for you to coach a participatory team dynamic. If you notice someone not chiming in, ask for their input. Double down on the encouragement you’ve been seeding in one-on-ones. If an individual is dominating the meeting, turn the tables and find time to privately ask them for their help in encouraging everyone to participate.
  • Remember to celebrate wins. Big or small, this reflection time is a chance to get excited and be proud of the good work that’s been done. This is positive reinforcement for progress. Also keep in mind that doubling down on a newly discovered efficiencies can be more impactful that correcting minor impediments. Avoid the mentality of “everything is broken,” instead focus on the abundance of opportunities and the progress you’re making.
  • Make everyone a leader. Participation is not enough. One of my favorite memories is when I had our marketing intern lead a company wide retrospective of about 30 people, and he crushed it. Create small roles for each retrospective, moderating the discussion, taking notes and/or creating action items. The practice of leading will not only instill ownership, but also ensure that the mastery of retrospectives is distributed throughout the entire team. As you scale, you will likely be more and more removed from the details of the work. At this point, it might be best to get out of the way, (and the meeting). Let your team take over.
  • Distribute notes publicly. Hopefully, at minimum, you and your team leaders will be able to skim the notes from other teams. You never know when this cross-pollination might be just as useful for other teams. Transparency increases the volume of the voice you’ve established on these teams and creates social pressure to do them well.

Finally, create a venue for company wide discussions in town halls + Q&A sessions. Notice the larger pattern here? You’re establishing different channels for the different tiers of your company: individuals, teams, and company wide. As you move up each tier, the topics of conversation should become broader for the broader audience.

Townhalls can be pretty hard and take a lot of effort to get right. Some huge companies put in a ridiculous amount of work, however, positive effects of this type of communication are multiplied across the entire company. Each company will want to incorporate different things, data deep dives, show and tells, customer highlights, etc. There is a lot you can do in regards to creating a voice, here are some key ideas:

  • Collect anonymous questions in advance. Some individuals will never feel comfortable asking questions in such large forums. Also, it can give you some time to collect more information and prepare responses.
  • Create a two-way dialogue. This might seem obvious to veterans. Dialog is pretty much the definition of a town hall. Sometimes you can get too excited about sharing a barrage of information with the company you can lose sight of this. It is extremely important to be modeling open, candid dialog at all layers of the organization.
  • Close the loop. Show your team they have been heard in your other channels. You’ve already been carefully gathering the common topics from your team. Show the company that those lines of communications are open and working by regurgitating what they are saying/asking for. Hopefully, you can also have answers, but sometimes that takes time. Standing up and saying, “I hear you,” sooner rather than later can go a long way. Then be sure to follow up when you know more.
  • Share the stage. Ask other individuals to share their work and ideas. Don’t tap only the other leaders and managers on your team. Use this as a venue to level the playing field, giving a voice to those who don’t have the opportunities on a regular basis. It might seem counterintuitive, but this can actually be a safer environment to speak in front of peers. Having time to prepare and deliver, can be far easier for some than spontaneous moments in the spotlight.

Making Room for Failure

If there is one thing we discuss obsessively in Silicon Valley, it’s failing. Phrases like, fail fast, and success is built on failure, permeate the community. On a small team, you intrinsically have visibility into everyone's efforts as well as strong personal connections. This gives you a lot of built-in trust and context to fall back on when things go unexpectedly. As a team expands, both visibility and connections naturally fade and cannot be relied on. You must design ways to make room for failure and extract the learnings. When individuals know a process exists to catch them, they will feel more comfortable taking appropriate risks to better your product and organization.

One tool that we’ve seen become almost ubiquitous among engineering teams is the blameless post-mortem. I believe this mindset works for a broad range of teams; in fact the core ideas have been lifted from manufacturing and healthcare. Blameless post-mortems are a great place to start, simply because failure happens, even when we don’t want it to. I’ll leave the details of implementation to others more versed than me, here is a good intro to the topic from the Etsy team.

It’s pretty obvious how this makes room for failure, here are some key themes that relate to building a culture of psychological safety:

  • Focus on the process over people. Start by assuming good intentions from all parties involved. Recognize that people are imperfect. Decide on a process that drives to the root causes and prioritizes from a perspective of risk management.
  • Let them be boring. Apply your process rigorously and thoroughly. A sound process will remove the emotion. This is a good thing; the attention can remain on the analysis and betterments. Stay focused, don’t try to iterate on your post-mortem process during the post-mortem. The less excitement, the better. Take notes and discuss in a retrospective.
  • Allow the individuals involved in the incident to drive. Once you’ve established a process, empower those who are closest to the process to work through the template with your team. This instills ownership and, once you get good at it, can be a catharsis for your team members who very likely have a lot of internal stress about the failure.
  • Carry the values outside of the post-mortem. Post mortems are a place to practice the perspective and skills needed to build what is often referred to as “just culture.” (Read more about how Etsy has seen phenomenal throughput by embracing this idea). Bringing those ideas into every aspect of the organization will pay dividends.

The vast majority of issues in your organization do not require a formal post-mortem, this is where retrospectives should come into play again. They provide a place to bring attention to these small learnings and, as a team, commit to getting better. The efficiencies gained from small improvements compound to massive gains in productivity and happiness. Here are key dynamics to think about with respect to making room for failure via retrospectives:

  • Maintain the blameless mindset. Carry the same mindset from your post mortem practice. Coach your team to use a lot of “we” language. There is one team. Fail together, solve problems together and then win together. Coaching this mindset here will pay dividends in day to day teamwork.
  • Document and prioritize. This is good housekeeping, but, also removes the burden of storing an issue in the back of their mind. Bringing things into the light helps source ideas for betterments. Make sure action items find a second life in the same task tracking tool you use for product work.
  • Hold the betterments at equal priority to product work. This can be particularly hard. We’re constantly trying to deliver value to our customers at lightning speed. However, if process improvement work is seen as second class work, it will be tough to get done. Increased velocity and quality is a win for everyone in the long term. Manage your balance between the two.
  • Build a shared vision for risk tolerance. Don’t fix everything and talk about why. Progress is about mastering the art of the work not done, (see the principles of the agile manifesto). With any project, you can bury yourself with busy work that isn’t relevant to the goal or isn’t relevant right now. Talking about the various failure modes, prioritizing and committing to what fires are critical and which aren’t will give your team a unified intuition as to what appropriate risks are and are not.

If you want to up your game even further, try out pre-mortems. These are helpful as high risk, high investment projects begin to appear. They require a great deal of receptiveness, the ability to embrace potential challenges. This takes a greater extent of humility and trust within your team.

Lead by Example

Image courtesy of Filipe Dos Santos

Gene Klann wrote that leaders are always being watched and that this should be used to their advantage. If you’ve been able to implement half of everything above, you’re probably doing a great job setting the tone already. However, there are a few things that, as a leader, can be hard and even scary to model for your team.

  • Encourage your vision to be challenged. Let nothing be sacred. You got where you are because you are pretty smart, however, we all have blind spots. The whole point of having a team is to bring different experience and perspectives to the table. Let ideas collide and then pull the best together to form something new.
  • Admit when you don’t have all the answers. Own it when you just don’t know, but commit to finding out more. This is the behavior we’d like to see in everyone on the team. Gaps can only be filled after they have been identified. Create a safe space for this type of candidness. This will allow your team to be comfortable with what they don’t know and figure out their paths to growth.
  • Throw yourself under the bus. Publicly own the mistakes of the past. Thankfully, as an early stage engineering leader, you probably have lots of fodder for this game. In the early stage of a company, you cut a lot of corners and made plenty of mistakes. Lay those out, praise when others find flaws and find improvements. Show your team how the path of progress is paved with humility, teamwork and a little bit of jest.

The overarching theme here is to set an example of healthy vulnerability. The pinnacle of psychological safety certainly exists in a space where vulnerability is revered. I like a quote from Brené Brown: “Vulnerability is the birthplace of innovation, creativity, and change.”

Humans are as brilliant as they are fallible. That brilliance cannot be unleashed if individuals are trapped in the confines of their own mind. One of our many challenges is to implement structure that accounts for our fallibility and, at the same time, nurtures a culture where the voice of brilliance can ring true.

More by Dan Rummel

Topics of interest

More Related Stories