Transcript
Charlotte de Jong Schouwenburg: Today I'm going to talk to you about scaling. My name is Charlotte de Jong Schouwenburg. I'm the co-founder of Bravely Amsterdam. What we do is we work with mostly tech companies. We're a learning partner, we do all kinds of things. We do leadership development, culture, feedback, anything you can think of, anything really that has to do with human behavior within a company. To give you an idea of some of our clients, very happy to be working with this crowd. That's what we do. I've gotten two statements that I got from clients when I talked to them about this talk. I wanted to know a bit about like, how did you experience your hyper-growth sometimes? How did you experience your growth, what happened? Actually, one of the things I got in various ways, and I summarized in this quote, we added teams across time zones and suddenly decisions took 10 times longer. You think, let's scale. We're growing so fast. We need more teams. We need to pump up the volume. We need to go faster.
Then, actually, everything takes way longer. You see a lot more drama. There's a lot more conflict coming. Another one is, we scaled from 10 to 100 engineers. Pipelines held up, systems worked, technical worked just fine, but people just didn't work together anymore, just didn't cooperate anymore.
LeanIX: Growing Up Without Growing Apart (Case Study)
What I'd like to have a look at with you is the case study of LeanIX. It's also a nice one to just look up. It's freely available. It's one of the very few research studies that's actually significantly researched. It's qualitative, but still. They looked into the development of the company LeanIX. What happened is LeanIX grew from 2012, so they had this really tight team of 12 employees. You can imagine the idea of everybody knows each other, very short lines, everything is super easy. Then they grew to 500, and now they're much bigger. During that time, a case study was done. What do you think would have been some of the challenges they faced?
Participant: Processes.
Charlotte de Jong Schouwenburg: What about them?
Participant: If you scale up this rapidly you most likely have some processes that [inaudible 00:05:32], to coordinate with 12 people, with 500, if you need to have anyone's opinion in the room, it would be quite tough.
Charlotte de Jong Schouwenburg: Yes, absolutely.
Participant: Decision making or responsibility for decision making. What decisions and where are decisions made [inaudible 00:05:48].
Charlotte de Jong Schouwenburg: Who decides?
Participant: Re-architecting everything, remote culture comes then during Corona, for that scale.
Participant: A group of 12, everybody knows everybody else. They know not only the skills and the strengths but also the personality, what's important for them. A group of 500 you can imagine is quite hard to find the perfect team setup.
Charlotte de Jong Schouwenburg: What we're going to talk about today is a little bit of the background of what kind of things are difficult when you scale on the human side. What are challenges? Which psychological processes underlie these, so you understand a bit more of why something actually happens or goes wrong. From there on, we're going to go into some concrete solutions.
How it went. Started as a tight team. They've grown. Then what you could see, coordination got tough. A lot of things broke down. People became unclear. Roles started to overlap. Decisions were slower. That's what happens. What they had to fix was role ambiguity, cross-function overlap.
Charlotte de Jong Schouwenburg: When it goes wrong.
Participant: It has a team nowadays already that is at least more aware of the importance of proactively thinking about communication.
Charlotte de Jong Schouwenburg: A little bit. Obviously, from my position, I'd love it to be more. Because when you knock on our door when it's 500 people and a mess, it's much harder to fix, and if you can do that beforehand. It has gotten better. Now we get quite a lot of requests at like 100, 150 people, simply because employee surveys start to come up. How satisfied are you? Burnout starts to rise. They are getting the signals off. It's not going well. It could be better, because you prevent a lot then.
Really, what they also saw was that communication was in silos. That means that you just do a lot of double work, unnecessary work. There's a lot of frustration between teams. The leadership style, that's a big thing we see again and again at companies. Usually, the founders have a certain style. They hire for that style. Then they experience that it takes more flexibility in leadership, because that just doesn't suit everyone. It's also cultural factors that play into this. Location factors, remote, face-to-face.
Scale Correlates with Strain
With scaling just come a lot of challenges. We get communication overload. There's a loss of shared context. The trust is a problem. Let me go through all of them with you. You communicate well with 12. You know each other. It doesn't happen with 500. It's just not going to happen. Dunbar's number, is that it's a guess, an educated guess that we can keep if we are neurotypical. Let's also put that out there, because Dunbar is a little while ago. He measured it on neurotypical people. We can sustain about 150 social relationships. This includes your friends, and family. I hope it's not just colleagues. That's just 150. It sounds a lot. Maybe if you're married, think back to your guest list, 150 is not that many.
No wonder if you don't continuously maintain that and have processes to actually be connected to other teams, you just won't feel connected. We also get this communication thing, where so much communication is being shared. I like the Slack channel. That's what I hear a lot from people. There are so many Slack channels. Everybody is posting stuff, or the endless emails with 1000 CCs. At some point you have no idea anymore. It's just like this total chaos. Humans do have cognitive limits, and that really is one of the big differences and big changes that come about when you scale. Your systems can scale. Like the technical side, that can scale. You can't really scale humans that way. You have to take that into account. Yes, sure, AI is doing a lot and taking over a lot, but this will still always be true.
There's also a loss of shared context. At some point you just don't know anymore who does what. Like I said earlier, problems are double fixed. You reinvent the wheel again and again. You use different systems, different solutions for same problems. You're basically wasting time and efficiency. It massively grows frustration. Because if then after a month of work you find out that, Joypro already fixed it. Great. It's not as if we didn't have anything else to do. It becomes inefficient. I'm a psychologist. This to me is most important. We know that trust doesn't replicate automatically. Just because you have high trust in your existing teams doesn't matter that if you build another team just like it, it will automatically have high trust and high psychological safety. That has to grow. It has to build. I'm going to introduce two practice examples for you just to make that clear. Spotify has these squads. They're very autonomous teams, usually smaller than 10 people. Very quick to make decisions.
Then the company scaled, bigger and bigger. More than 1,000 engineers. Loads of squads. What did we get? There was a lot of reduced psychological safety within and between teams. They also saw that those teams, when they had to rely on each other, that didn't work. Because if you don't trust another team, you can't rely on them. That meant that there was a lot of micromanagement going on. A lot of indecision. Decision fatigue. A lot of zoning out. Also, in-group, out-group, "Yes, but that team. Yes, but that team". Sounds familiar maybe. They had this great realization that you can't replicate trust, even if you can replicate autonomy. You can make all of these new teams just as autonomous. You can copy paste the system. You can't copy paste trust.
Just to be clear about what trust means in this context, and what the difference is with psychological safety. Trust is the belief that others will act with integrity and reliability, even when you're not doing well. Even when you are having a rough time. Even when you're not sure what's the best solution for this. That means that you can give feedback openly and honestly. You can communicate with each other. You can also communicate about, I'm really struggling here, without fearing being excluded from the group. Psychological safety is the idea that you can take risks within that group. This is a bit more. It's like the funnel of trust. Psychological safety and trust are highly interrelated.
However, you can sometimes have a group with high trust but not that big psychological safety yet. That's usually due to processes or way of work, that says, you're not allowed to make mistakes. Psychological safety is all about growth mindset. We love it, because teams just become much better. Creativity goes up. Output goes up. They fix things faster in their own way. They use the autonomy that they've been given. You can give a team autonomy, but before they can harness that and reap the benefits, there needs to be trust and psych safety. You can copy the structure. You can go, let's just duplicate that. You have to rebuild how they interact with each other to get the results that you're looking for.
Another example is Google's reorganization anxiety. Maybe you're aware of Project Aristotle. I'm a very big fan because it gave us a lot of data, a lot of insights on how teams function, looking at social drivers. What we would see is that they would split up teams, teams that were working really well together. For some reason, reorganization, whatever, they would split them up. Even though everything around it stayed identical, trust would go down incredibly. They would really have to rebuild that. When you have a social system, a dynamic, and that changes significantly, you're going to have to fill up that trust tank or that safety tank again. You have to find new dynamics. Maybe you've known that when someone in your team left or someone new got added to the team. You have to reconfigure the social dynamics. That takes time.
Scaling at very high speed without any investment into this social system that has to grow, you're onboarding these teams. You're not only onboarding individuals, you're also onboarding the social systems of teams, and that takes time. Ideally, some guidance. You can add all the tools you like. Let's have another shared board where we can all put our ideas. Yes, you can. People will only make use of that if there is psychological safety. To summarize, scaling is a challenge for teams. There's communication overload. There's just so much going on that you almost can't keep an overview very often. There's a loss of shared context. You don't actually know what are the other teams doing. Are we duplicating? Are we actually working aligned? Trust doesn't replicate just because the other team was great at trust.
Spotting the Human Bottlenecks
We know that now. That sounds great. How do you actually see that? How do you know that this is happening? We like to call it a monitoring dashboard. It's just, you keep tabs on how are these things going. Let me talk you through them. Human latency. Basically, the question, how long does it take to do our job? How long does it take to move from a problem to a solution, to acting on it, to getting into motion? What we see, this is actually from a client of ours, an organization that we worked with, they measured this. They found that as they were scaling and they were getting many more teams, and teams were changing all the time, that actually it took them five times longer.
Think about the money that costs and the frustration. I hear that from so many coaches that I work with that they say, "It's great that my company is scaling, but everything is so slow now. With more people, you'd think we'd be more efficient". We also see it in error rates. Like I mentioned before, things that go wrong, miscommunication, rework. The idea of when is something done? When is something proposed? When is something rushed? What's a rush? What's a quick delivery? You could see it in decreased efficiency and engagement. That's usually a very interesting one that people underestimate because it's harder to measure. If you get this vibe of, this team isn't doing very well. People get cynical. They zone out. They complain a lot. What we see then is people, they zone out. Afterwards, when the deadline was missed, or you are checking in with someone in a one-on-one and they're super frustrated, and you're like, what's going on? What happened?
Then maybe they're like, yes, it's just not working, or, yes, I'm so done. Or, I didn't understand the brief, or the team didn't want to work with us, whatever. People, they zone out. Especially when you scale internationally, of course, but most of the companies we work with work internationally anyway, so more cultures usually get added into the mix. You get the cultural drift. You move away from, this is our identity as a company, and you move into, but this is how it's done over here and this is how it's done over there. Love the example where in the U.S. moving fast is like, let's just ship a lot. In Japan, the exact same team in the session where they were together, they discovered that their definition was just very different. It's just like, let's just not have delays.
Is there anything that you're already seeing at your company or in your team? What do you think, if you're going to scale, would be the biggest bottleneck?
Participant: From trust experience, two and a half years ago, it was decided to build a new product, and within the company the team was assembled, where people were picked from multiple different teams. I could really see what you talked about as development of some trust. Then, on top of this, after some time, it's safety, where you felt safe through directly addressing another person and say, "You introduced this bug here. What was happening?" This person now feeling safe to say, "Yes, true. I did this. Cool, let's see how we can solve it".
Charlotte de Jong Schouwenburg: Then it's not identities crushing. You can stay on the content, yes, let's fix it together.
Participant: People are just fine with saying, yes, I messed up there, let's just try this.
Participant: You see a lot around the culture, [inaudible 00:26:38].
Charlotte de Jong Schouwenburg: Isn't it so fascinating, at least to me, that's why I studied this stuff, why we assume so much. It's done. Excuse you? Why isn't it done? My idea of done is very different. Yes, but I thought we had some time to go over it together. No. That's such an interesting thing. In line with culture, just for all of you, if you're interested in, maybe you know the book, "The Culture Map". It's an easy read, and it's very interesting to look at how different cultures typically always with a grain of salt, but typically work.
Participant: Human latency, we experienced this, especially coming off in different teams. We touched it, more like, you were there or like operation or something like that. They usually take longer time to review the code, and you just wait. The reason I think, they have their own work, they're focusing on it. They're already off about that.
Tools for Behavioral Scalability
Charlotte de Jong Schouwenburg: We're now going to move on to what you can do about it. Because, after all, we all understand it's an issue, but what can you do about it? I'm going to introduce a few skills to you from bigger topics, and these are really about creating that behavioral scalability. They're not a magic trick. It's not like, solved overnight. I'd be writing books and be a millionaire if that were true. Unfortunately, it doesn't work that way. It still takes quite a lot of investment. What I'd like to invite you to do is while we're going through this, I'm going to throw a bunch of techniques at you because I'm always like, but then you can pick the one that fits you best. What I'd love for you to do is while I go through them, already mark for yourself, this is one that I could try, or this is one that would be really useful for our team, just so you look at it in a way that you can make it your own, straightaway. Otherwise, it's like, yes, it was a nice presentation. What did she say? I don't know, she was wearing blue. It's not going to help you. Try and focus on, this is one that I would like to use.
The first is communication architecture. We absolutely want for people to hear something again and again. You think, I said it once. It was mentioned in the meeting, should be clear. Remember that Dunbar's number and the information overload and our brain just reaching a capacity much more quickly when there's so much information. Repetition is gold. Use different times, different mediums, different ways of repeating the core message, of repeating the context, repeating goals. Why are we working on this with three different separate teams that are silos, but we're all working on this common goal, remember? That is really something that you want to be doing as a leader and as a company as a whole.
Keep reminding people of the bigger picture, of the things that you really want them to do or to remember or to work towards. It is a bit slow sometimes. It might feel a bit like, yes, discuss that, let's move on. It is a pretty surefire way to fill in the gaps of the people that you would otherwise lose because they just didn't get the message. Ways you can do that. Multiple communication channels. Have a summary. Put it in the Slack as well, put it in the email. Remind people during the all-hands. You can make these very standard processes. Doesn't have to be very different information, just being presented in different places repeatedly at different times, very consciously. We want to have more bridges between teams. When you make sure that more people in different teams cross-functionally that have to work together at some point, that they know each other personally, for example. Using an offsite for this is a beautiful one. Don't just have an offsite with the team separately and then go, but we'll have drinks together. What happens? Team A, team B, team C, we never see each other. We always talk about work. I'm not going to talk to the other people. If you want there to be more bridges, be more connections between teams, you have to create that. Like buddy programs, for example.
There's a lot of ways to do this, but you have to create bridges. Actually, with things like this, ask the people themselves. What do you think is the best way to connect more with the other team? What do you think you need to be more open and more chill about working together? Again, creating trust. We want the repeated narratives. This really is one where quite often we work with C-level and senior leadership to have very aligned narratives, very aligned messages, and how can they trickle through? How can you do that continuously and predictably? We want the backup rituals, things like retrospectives. What went wrong? Don't just point at the one team. Doesn't work that way. Let's have everyone who was involved or who could be involved being part of that. Frame it as, we learn. Sometimes we fail and we learn. Let's do it better next time.
Participant: How do you deal with people who see the team communication as a waste of time, and look at it differently.
Charlotte de Jong Schouwenburg: Explain to them what's the background. Explain to them why you do it. Just literally grab this and explain. The reason why we do it is X, Y, Z. Usually when people understand the ratio behind it, they are happy with the behavior that's happening. If they just constantly see the same thing pop up, they're like, yes, I've seen it, when frustration then goes up. Or ask them yourself. Ask like, this is the problem that we're presented with, what do you think is the best solution for it? Ask them. Give them ownership. People stop complaining once they have to fix it.
Engineering trust. We want people to have trust and psychological safety. We need them to feel free and comfortable and be rewarded for being honest and transparent. For saying, I agree to this deadline, or we agree to this deadline, and actually we're not going to make it. Let's find a solution together. Then ideally not the day of the deadline, maybe a week ahead. What we're usually quite good at is doing all this for the systems that we work with. Making sure there's a backup. Making sure that nothing can go wrong. We forget about the humans that actually have to do it. Creating a structure, a box around this where we remind people that it's good, it's rewarded to be honest, really helps. There we go. What can we do? Discussing together without shame, without, yes, but you did. Actually, whenever you start with, yes, but, not a good start. Make sure there's no finger pointing. Make sure you see post-mortems as an opportunity to grow and you continuously frame that. People look at leaders walking the talk. You can say whatever, if you do blame, people notice, they won't trust you anymore. Being very transparent in how decisions are made.
The more people can follow your ratio, and that comes in line with your point, Stephan, the more people understand why things are done the way they're done, the more trust they have and the less frustration there will be. Once they understand, they're quite often ok with it. Be predictable in your rituals. It's something we have done as humanity since cave days. We like rituals. We like repetition. We like predictability. It calms our brain down. By definition, it connects us, makes us able to be a lot more trusting, more confrontations that end well, more feedback. All of that goes hand in hand. You also really want to use that vulnerability as a leader. Like I said, walk the talk. Stand by your mistakes. I messed up here. I'm learning from it. That's when you're showing, you're demonstrating behavior to people that you want to see from them as well.
Also, if you see that there's conflict in the group, say it. If you see behavior that's not ok, flag it. Do something about it. Show the way that you want people to interact with each other. Also having open metrics, giving people insight into, how are things going? What are other teams doing? It just becomes more clear. I know this is quite difficult to do, but it's like slowing down to speed up. Sometimes we need to slow down to build in some of these systems so that then the teams can use their autonomy and move quicker from there. We want to distribute cohesion. What does that mean? We want a very strong sense of connection. Shared purpose. That cohesion that normally we find in one team, we want to create a way to get that across teams. Distribute it. Distribute that connection. Like I mentioned earlier with an offsite, for example, find ways to bring teams closer together.
Things you can do are retros, virtual coffees if it's hard to get together physically. All-hands, have some storytelling once again. Frame it in a, what went wrong? How did we fix it? What did we learn from it? Really that growth mindset. I like the third one. We do that with quite a lot of companies that we work with. If you rotate facilitators, once you give people ownership where they are visible, they usually take it and step up. That's really nice because then they all have that shared responsibility and they can talk to each other about it. Of course, any kind of way to bring cross-team people together, you're building bridges.
Leadership really is very important in this. It's not about controlling the system, but you need to make sure that that social system is reliable when stress hits. That's trust. Enough trust is there. The team is coherent enough. You can measure things like, how long is decision making taking? Are the teams speaking openly? Is there cultural drift? Talk to people in your one-on-ones. Check this with them. What do you experience? What do you see? They're in the midst of it, they know. I think this one is maybe my most important one that I see so much going wrong. You want to walk the talk. When there are challenges within your team or cross-teams, just staying on the content, let's fix it within time, let's do this, X, Y, Z, is not going to fix it. It's a symptom, a missed deadline. Somebody not giving you a red flag, not informing you. Teams not collaborating, cooperating. They're all just symptoms. You know now what the drivers are beneath it. Go and look. Go and find out. That is where the solution lies.
Otherwise, you'll be repeating the same thing over and again. That's what we see. Loads of fires. When loads of fires are burning and you're constantly trying to clean everything up and fix everything, that's when you're only looking at symptoms and not the underlying cause. You want to be preventative as well. When things are going well, great. Let's install structures to keep that so. The last thing, yes, I know we have great systems and AI and all that shebang is amazing, but humans make mistakes, very simply so. Creating psychological safety and trust also happens by counting that in, knowing that that's just going to happen.
Then it not being the downfall and the drama, but that there's a backup plan. Even if the backup plan is, people, we've been in a tough situation like this before, let's fix it together. Great. If people see that happening again and again, got a ritual, got safety. As you scale as a leader, of course the content is still important. That's not really what you should be working on most of the time. You should make sure that your teams can work as best as they can. You should facilitate them to do the best job they can. For that, we really need to look at that human system.
Questions and Answers
Participant: I'm a big fan of circulating teams. I believe everything you said about trusting and time, and the same teams just tend to be better, and how just removing one person or adding one person may disrupt that system. Just accepting the reality that this will happen, can we make the humans better at team formation? Can we all get better at intentionally rebuilding that trust way more quickly, being quicker about it, and all that?
Charlotte de Jong Schouwenburg: Absolutely. Teams become more resilient. The higher your psych safety and your trust is, these rituals, stuff like that. We like rituals. If you know as a team, this is the way we always do it, then there's predictability. If someone leaves the herd, dynamics are changing, but at least the system is the same. At least the routine is still there.
Participant: When teams need to form quickly for something, for an intended purpose, can we make the individuals in those teams better trained at, ok, new team. I know what's going on. Let's work on building trust.
Charlotte de Jong Schouwenburg: For this, ideally, when you're creating a new team, I like to call it the relationship bank account. Whenever we create a relationship with someone, when we're making relationship, we have like a bank account. Can mean the plus or it can mean the minus. Any relationship we have with someone. Any social connection. Investing into a relationship bank account is done through anything but work. Talking about, where are you from? What makes you tick? What kind of person you are. Because that is what we need as humans to build trust. Then you're investing in the relationship bank account. If that is in a plus, you can deduct from that plus by, for example, giving someone feedback. That is clearly a team with a positive green bank account. We need the positive bank account.
If you form a team and you straight away go, deadline is done, full into project, they are all going to end up in very negative bank accounts because they have to give each other feedback and they have to be critical towards each other. They never build up that plus. You know there will be conflict. There will be frustration. The team will not feel connected. No, of course not, because there's no plus to deduct from. In every relationship that we have, we need to invest to first fill up the plus.
First, invest in the relationship, then invest in the content. That is my biggest tip for any new team that's being built. It's not that hard. There's quite a lot of techniques you can use for this, but it takes some time. If you invest in it, like when we have sessions with new teams, and you can facilitate yourself as a leader. Give them half a day or a day, but give them the rule that they cannot talk about work. Like you're not allowed. Let them be in couples. Where are you from? Who are you now? Where are you going? Then you let them translate or introduce each other to the rest of the group. Stuff like that, it builds a lot of trust. It doesn't take that much time, but you need to begin with that first.
Participant: How do you handle when teams have different preferences? One particular example, some people on my team, they prefer having webcams on. Some people experience Zoom fatigue. You want to acknowledge both the groups.
Charlotte de Jong Schouwenburg: Leading with vulnerability. Saying that out loud. Being the person that says, "Ooh". It's a little bit challenging, when half of us have the cameras on and half of us don't. It's a bit strange, maybe. How do you experience that? Have the team talk about it. Anything that is an elephant in the room, say it out loud. People won't feel as awkward about it. It becomes a conversation. Then you can come to a solution together. Maybe everybody's fine with it. Once they understand each other, maybe they're fine. Or maybe you can agree to have the cameras on for the first 15 minutes and then turn them off, whatever. Let them figure it out. Quite often as a leader, you need to be the one giving the safe space to actually discuss it. Any elephants in the room, call them out.
Participant: In this environment of like, at least in the tech industry, for constant layoffs, psychological safety takes a big hit because now people are afraid of admitting mistakes because they'll be on the chopping block. How do you install psychological safety in that environment?
Charlotte de Jong Schouwenburg: Say this out loud as well. Make that a topic of conversation. Then stay away from the content. That is the biggest challenge here because that is fear. Fear is not about, yes, but we're working on project X, Y, Z. That's not how you're going to get there. This is about fear of your livelihood, of the future, of your ego, of your identity. Have a conversation about that, "I hear your worries, tell me about it". Don't have to fix it. You can't fix it. You can provide a safe space. That is really what trust and psychological safety is, that you won't be judged. That it's ok to have a talk about it. You won't fix their worries, but at least they will be able to regulate themselves so they can function again.
Participant: For rebuilding trust, you talked about the transparent decision logs, where leadership is transparent about why they're making a certain decision. Typically, what I've seen is, when it's small, the leadership is pretty transparent. They tell everything as it is. As you scale, that becomes more dense. It's a very common pattern. Even though they know that this is what will build trust, they're not doing it. Are they optimizing for something else, or is there a fear that if you're too transparent, it'll cause some other issues?
Charlotte de Jong Schouwenburg: Of course, it's clever to think about, what is my audience and what information do they need? What information do they need in order to feel comfortable about the decisions that are being made? I wouldn't just straight on send everything that C-Level is talking about. Here you go, people, because then again, you get the information overload. It does make sense to narrow it down to what is truly relevant for them. Let's be fair here that usually also when a company scales, they're so busy with the new project and the scaling and funding, that this takes a backseat. It is worth investing in because otherwise you'll see that you lose the people that are actually having to do the work.
See more presentations with transcripts
