Gene Kim (00:00:00): This episode is brought to you by IT Revolution, whose mission it is to help technology leaders succeed through publishing and events. You're listening to the Idealcast with Gene Kim, brought to you by IT Revolution. The last two episodes were with David Silverman, CEO and founder of CrossLead and a co-author of one of my favorite books that I've read in the last decade, Team of Teams. As well as with Jessica Reif, director of RND at CrossLead, who like so many of us, comes from a software background. Gene Kim (00:00:38): The last episode was David Silverman's presentation that he did at the DevOps enterprise summit, London virtual in June. If you haven't listened to those two episodes yet, I'd recommend you listen to those first, because this is a continuation of the first interview. In this episode, we discuss how internal marketplaces are structures that can connect mid- level leaders to each other, helping allocate scarce resources to where they're needed most, which enables the further unlocking of capabilities. We compare and contrast the concept and terms found within the Agile and special operations communities, especially those words, that concretize concepts that we've thought of for years, but didn't actually have a word for. Gene Kim (00:01:22): We talk about the desire to bring down the cost of change, and what happens when the cost of change is intolerably high. We talk about the new skills that mid-level leaders need in order to survive and thrive in an era where just being functionally excellent in one's own functional silo, is not enough to help our organizations win. And we talk about where they think we go from here. Okay. Let's jump in. We left the first interview talking about the role of internal marketplaces within the Team of Teams story. I asked Jessica about parallels that she's seen in the technology domain. Jessica, have you seen examples where, in a non-military environment, you've seen kind of these internal marketplace, horse trading mechanisms work as well? Jessica Reif (00:02:11): Absolutely, Gene. I think that's a really good question. One example that we've seen is, where there are shared resources across teams and where those resources might be more valuable to another team at a specific moment in time. So let's say for example, that I'm a development team and I'm short on design and resources. One of the situations that I've seen, working in the technology space is, you're able... The managers of the teams or the team members themselves are able to work out amongst themselves, perhaps have somebody who has a particular, whether it's design expertise or they're a front end developer that is very talented and able to get up to speed quickly, will actually go support another team for two weeks to support with a particular effort and then disband and go back to her original team once that effort's completed. Gene Kim (00:02:56): Dave, when you hear that story, does that resonate with you in terms of yeah, that's exactly the same sort of phenomena and dynamics that you saw in theater, whether it was Iraq or Afghanistan? David Silverman (00:03:05): Yeah, exactly. For us, it was helicopters and collection platforms that could give us visibility on targets. And without those two things, you couldn't get there and you couldn't figure out where your target was and then monitor it as you're coming in. And so, the higher priority units had pretty much dedicated resources on a nightly basis. And so, if you were a lower priority unit, which is where I found myself on most nights, I was able to go to my counterpart and say, "Hey, can I borrow said asset for a period of time, where you potentially have a window in your system? And obviously you'll pull it back whenever you need it." But that extra window gave me the ability. And I would adapt my tactics and techniques, knowing that I only had a limited timeframe. So it might increase the risk profile a little bit of what we were doing, but ultimately, it unlocked capabilities that we previously were constrained to just the very, very high end, tier one units. Gene Kim (00:03:58): Jessica, when you hear Dave speak, he's obviously incorporated the language of Agile and DevOps terminology into the way he describes things, and I find that kind of intriguing just because it's my belief that so much of the post industrial age management methods, really are being mostly pioneered by the technology community. One, do you notice Dave doing that too? And two, to what extent do you think that the technology community actually has kind of developed a language to talk about these things in a way that's actually more direct than other domains? Jessica Reif (00:04:32): Yeah, absolutely. I think that the parallels are really uncanny. When I first started working with Dave, which was now seven or so years ago, and was working not only with Dave, but other former special operators, and I started to learn their language and having some familiarity with Agile myself and knowing the scrum methodology, knowing that daily stand-ups and regular retrospective meetings were sort of par for the course in that domain, I was very surprised in a good way, when I heard that those practices were also heavily used in Dave's old world. For example, Navy Seals. And if you've read the book Culture Code, they go into AARs in detail. Daniel Coyle goes into the detail in that book, but the AARs or after action reviews, are a core part of Navy Seal operations and improving as a team. Jessica Reif (00:05:24): And that is very similar to the retrospective that is held at the end of a sprint for an Agile software development team. Now, as far as the technology community doing a better job communicating these principles, absolutely. I think that pretty much any major tech company that you go to these days, they're going to be practicing some form of Agile methodology, whether it's scrum or something else. And they have words and a specific language that they're using to describe, whether it's inspect and adapt or retrospective, but they have words and known terminology that they're using to describe these processes of introspection, that help teams get better. And I think that having that institutionalized in such a pervasive way makes it easier to engage in those types of thoughtful conversations. Gene Kim (00:06:11): Dave, can you react to that? Maybe on a scale of one to 10, right? I mean, to what extent has the terminology, vocabulary of technology created a word for you that you didn't have before, that has been useful to you? David Silverman (00:06:23): It absolutely has. I mean, for me, sort of the epiphany was, when we got out, we were struggling to sort of quantify and articulate the experience we'd just gone through. We knew it was special, we knew it was unique. So much of what we had sort of learned and institutionalized over decades at the small team level, right? So that fire team, that squad, that troop, that assault group, effectively was more or less tried and true for the last, call it 15, 20 years. What was different about this was how it sort of connected up. And all of a sudden it felt like, what it did at a small level, at scale. And that was really something different for us, which was the impetus of the book. David Silverman (00:07:02): And so now fast forward, I'm sitting down with... I was super fortunate. One of my true heroes is a guy named Brad Smith, he's a chairman CEO of Intuit. Just one of the most phenomenal leaders I've ever come across in my life. And Brad was inspired by this story. I think he has a natural appreciation for service, specifically the military. I think there's some history there for him personally. And so, he liked the story and at the end of a couple of discussions, he said, "Hey, you should read this book." And it was Lean Startup. He goes, "We've been using this language, you're using different language, but it's the same thing. And what you're talking about is doing it at a scale that this book doesn't really account for, which is very interesting. I'd like to figure out how to explore with you guys, here at my firm." David Silverman (00:07:44): And I was sort of blown away because before then, I'd never heard of Agile. And so when you start looking through the terminologies, yeah, it was almost like... It was immediately able to sort of map every single thing we were doing, both at small level, and then at the larger level, I think it gave us a framework to start thinking about how to translate our messages more effectively to a larger audience, which the book Team of Teams was really all about. Gene Kim (00:08:08): Awesome. And by the way, I even retold the story about the global call, 3,500 people. Jessica, you highlighted the point about, it was so much about exposing and discovering dependencies, right? Both that were either unknown or to make explicit. I think on the Frontier Software right now, so much of our work is impeded by dependencies that we don't know about or dependencies that we're just coupled to, shackled to, and try to unshackle ourselves from that. Jessica Reif (00:08:32): Oh yeah. I was just going to say, absolutely. If you take two teams that are identical on paper and that have similar skill sets and ways of working, and you give one of them a Greenfield application and say, go build this. And another one has to build something that sits within a larger system, it's amazing how quickly the progress diverges between those two teams. Dependencies are a real barrier to getting things done. Gene Kim (00:08:55): So last time, it was actually before our interview, we were talking about the story that Steve Spear told me that... At that point, it was probably just a week and a half prior. And I had asked for your reaction. I was stunned that, that too also resonated with both of you. And so the story went, in the mid 90s he went to Japan with his mentor, [Dr. Kempoan 00:09:18] from the Harvard business school, and a VP of manufacturing from Big Three automotive manufacturing plant. Gene Kim (00:09:21): And so, one of the first things that they saw there was a plant manager describing how they were doing 60 line side store changes per day. And so the line side stores are basically for any given work center where you store your inputs, the raw materials, right, of which they would process and then become outputs. And then the VP of manufacturing from the Big Three plants said, "That's crap." Right? We tried six line side store changes, and we ended up shutting the whole plant down for three days because parts were not where they were supposed to be, certain work centers were starved and suddenly they couldn't actually manufacture completed cars at the end of the line and would actually take three days to unwind from all that chaos. Gene Kim (00:09:55): And I think my big learning from that was, that there are really kind of two types of plants. When I asked, kind of what distinguishes one where you can do 60 line sized store changes in a day, easily, quickly, smoothly, safely versus one that's so brittle, you can't even do one without potentially causing global impact. It really is, do you have a central mothership that knows everything? And if you try to change something, often you miss a detail and suddenly, you end up with a potentially catastrophic disruption versus one where you can make small changes all the time. And to me, it really is, to what extent can you really bring down the cost of change so that you can make them easily, quickly and so forth? Does that resonate with your own experience? You brought up the notion of, before you were able to unlock all this productivity, there was another effort where you tried to double the operational tempo and then that actually led to a lot of mishaps, accidents, injuries and so forth. Can you just talk to what that story that came from Spear means to you? David Silverman (00:10:51): Yeah. It takes me two very different directions, right? So in one sense, it sounds like, how do you remove handoffs or blinks in the system because there's risks that's inserted in when that happens. And the implied thing that Jess said, which I think is directionally correct, which is if you can reduce interdependencies and you can bring people into a common, let's say, value stream, and you can figure out mechanisms like scrum or whatever else to sort of govern how that team operates, you can potentially become much more productive. The challenge for us was, there were dependencies that you couldn't eliminate. And those dependencies, in our case, were instituted into law, right? It was literally like, this organization can and will do this by these authorities [inaudible 00:11:36] And so that would naturally create these pretty significant cultures and siloed behaviors. David Silverman (00:11:42): And the result was, we had to figure out a system that could hack that and say, okay, well, where can we find common ground, i.e, dependencies. Well, we can say, look, we know we are connected and we know you guys operate independently and we're not going to improve, let's say the performance of the CIA or the FBI or Homeland security, and you're not going to probably do that with Delta Force or Seals or Rangers, but where we have to come together, we can try to improve that. And increasingly what happened was, when we created the vehicle or the venue for that to occur, people figured out common ground and then that's when productivity really jumped, because instead of having to go up through a hierarchal structure where there was a bureaucracy that managed information flow and learning, we now said, hey, lets... We're still structured that way and we can't adjust that, but we don't have to operate that way. So we can operate thinking of ourselves as one common value stream of no handoffs, where we're all in it working together, and that's all of a sudden when the productivity sort of went bananas. And the hard part was managing the bureaucrats in between, right? So at the very top, I think people got it. Usually, if you put a group of people in a war zone with a known problem where life and death is at stake, they'll figure out how to solve the problem together, right? They sort of forget some of their programs. It was everybody in between that became a problem. And so most of our systems were designed to sort of hack that middle layers, the permafrost, that inhibiting progress. David Silverman (00:13:02): So early on, when you say go faster, what you meant was, everybody's operating in their own silo, just spinning quicker. And the result was, you started making mistakes because the way that system was designed to work was already optimized for what we assume to be the most productive level. And so what we had to do is, just fundamentally shift and say, look, we're actually going to go slower to start, and we're going to bring these people together and then we're going to then operate in this space. Now all of a sudden, you don't have these information gaps, right? So an analyst that's been studying a target set in Afghanistan for 30 years, 20 years, 15 years, whatever, deep expert on said site. They're going to know something and do pattern recognition on some piece of intel or information we're extracting from the target, much faster than anybody, even our own intelligence communities inside the special operations community and certainly more than an operator who's on a three-month deployment or six month deployment, because they're kind of in and out, they're looking at a different target every night. David Silverman (00:13:53): They might be looking at 10 targets a night, right? They'll have some recognition, but they know there's probably somebody else whose sole job is to understand said personality and how they work. And so when we're able to connect with that person directly and bring them into the engagement, make them part of the actual operation, now all of a sudden we're bringing that expertise to bear, effectively in real time, and so the quality and the pace just unlocked for the organization. And that was what the game changer. We went from basically doing one every other night, to do four or five a night, as far as kinetic operations. And in our case, our casualty rate went down, our success rates went up. So it was a significant boost in productivity for us. And that wasn't the full solution, obviously. I mean, we were only a piece of movement, we're just creating stability and local governance, but as far as suppressing the immediate, most damaging threats, that was sort of our role in the larger thing. That process itself got a lot more effective. Gene Kim (00:14:49): So just that notion of cross-functional teams, I'm betting the expertise is kind of where it's needed. I have to imagine that deeply resonates with your own experience. Jessica Reif (00:14:56): Yeah, absolutely. And I think that one of the main points of both the story that you told from Steve Spear and the example that Dave just gave, is that empowerment on its own is not inherently a good thing. If teams are empowered without having a common purpose and they're empowered without having shared consciousness, because having a lot of empowered teams can cause a lot of chaos really, really quickly, if the infrastructure isn't in place for those teams to communicate and collaborate and ensure that the operations that they're running or the tasks that they're performing, the code that they're deploying, whatever their focus is, isn't in any way negatively impacting the other teams, with whom they have to collaborate to achieve a shared mission. So we see this a lot, is organizations want to move faster. They want their teams to feel empowered to do things really, really quickly. However, if those shared mental models for how a change here impacts the rest of the system, don't exist, they really have to be built before that empowerment can be productive. And I think that, that's the process that Dave just described. David Silverman (00:15:59): Yeah, and it shows up two ways. For companies that are well-established, that are high-performing traditionally, but they've been working pretty independently because they're like, hey, we want to empower our people, which is... Most people would assume that as a good thing, the result is, it just bogs the system down, right? If you have a company that's high-performing traditionally and they're very empowered individual teams, but those teams are siloed. And as a system becomes increasingly more complex, what you see is, the ability for that organization to adjust and shift becomes slow. It becomes really, really hard. It's almost like concrete, so it just can't move. And so then eventually, what starts happening is, smaller companies who aren't encumbered by that, start picking at its edges. And eventually over time, a thousand cuts adds up and the organization becomes pretty sick. David Silverman (00:16:45): And for companies that are already farther down, some type of pivot or change in their industry, this can become decisive, right? So if it's not quite as traditionally high-performing, then you can see it compromise and they end up losing deals because the sales organizations aren't synchronized in their approach to a customer, who's changing maybe their perspective on how they buy or how they quantify value. And you get over bid, you aren't learning as fast, and so the whole thing sort of starts to really become a significant risk factor. So most of the stuff that we talk about is, how do you beat that? The challenge, I think that we're finding with a lot of organizations, it isn't always intuitive that there are these interdependencies inside of these value streams, and you can sort of define value streams at pretty different levels, right? David Silverman (00:17:26): So they can be pretty macro where it's sort of hard to say yeah, okay. Directionally, in our case, we're trying to protect the Homeland. And say, okay yeah, I got it. But if you went down one layer, the way the FBI and state department would define success associated to the mission, was radically different, right? State department would say, hey, we need to preserve relationships with these governments so that we can promote US interest. And the CIA says, well, we need to collect information, in which [inaudible 00:17:49] lie, cheat and steal against their governments, to basically provide us content. Well, obviously those two things are an opposition to each other. And then you add in somebody like the Navy Seals or any other special operations unit, they come in and they're literally doing something kinetic, which means blowing something up. And I tell you, that's not good for diplomatic relations, it is not good for long-term network. David Silverman (00:18:10): So you can see, where you go one layer below what was the initial value stream, which was, hey, protect the Homeland. And you've got pretty entrenched cultures whose incentives and models are not just misaligned, but actually in opposition to each other. And so trying to figure out common ground there, that was a challenge. That was the challenge, and it was a lot of give and take. It was like, you're going to compromise in this operation. You guys want to do tonight because this source is more important to us in the longterm, or you're not going to do that because I'm not going to be able to explain that to the host nation, even if we don't think the host nation is the best partner in what we're trying to do, right? And other times we said, "Well, no, the risk or the profile is too high." David Silverman (00:18:50): So it was a lot of give and take, where traditionally, that would have been, maybe decisions be made at the highest levels of government, which means it's inherently slow. Now we were able to sort of disseminate that context down to a much more local level, so we could start to have a series of values and principles that could govern our local decision-making, which obviously improved, substantially, the speed and quality by which we were doing things. I mean, it's really, really hard. I mean, for us, we had the benefit of a real live burning platform, right? You can sort of almost morally guilt people into doing the right thing, where locally, I think people are naturally problem solvers. You put them together, they sort of figure it out, right? They'd kind of have to get over their own biases of whose in charge and what, and eventually you start figuring... I think there was a marshmallow experiment that was done at MIT or something, right Jess? Where they look at, kindergarteners consistently solve a problem faster than people with MBAs. David Silverman (00:19:39): And the reason why, is because they weren't spending all this time trying to figure out where they fit in the system. So you have to overcome, I think that friction, but once you do locally, my sense is, you start to get on with it. The harder points above that layer, is what I've seen. People are more entrenched and they can't really see the value. So trying to get them excited about compromising maybe their positional authority or whatever it is, for a greater good, becomes a real kind of constant pervasive challenge. Gene Kim (00:20:09): Jessica, how about you? Does that word, concept change, does that resonate with you as a concept? Jessica Reif (00:20:13): Yeah, absolutely. And I think anything that an organization could do to lower that cost of change, they're going to... It's going to unlock some innovation, but one difficult aspect is just identifying, specifically, what's getting in the way of change. What is preventing us from changing quickly? Because oftentimes it's not obvious, it'll require two or three layers of analysis before you find where those true barriers are. Gene Kim (00:20:37): Gene here. I want to concretize a couple of these concepts, because I think they are so important. First, I love what Jessica said, that empowerment alone is not necessarily a good thing. We need a shared consciousness between those teams with a clear set of shared goals. We need the ability to have a low enough cost of change that allows teams to be able to work independently without small changes having potentially catastrophic consequences. And that often requires reducing the coupling between teams, that enables teams to be able to work independently, to create value independently, without having to constantly prioritize, sequence, communicate, coordinate, de-conflict and marshal with potentially scores of other teams. But this is often difficult due to the increasingly interdependent nature of the work that we need to do, and increasing specialization required. Gene Kim (00:21:28): And this will likely require us to have working relationships and sanctioned interfaces between teams that likely go across vast portions of the organizational hierarchy, which enable them to quickly communicate, mobilize, marshal, horse trade and so forth. And these teams that span so many different parts of these organizations, may have to forge very specific, shared goals, which might be different than the more parochial goals of the organizations they are part of. This very much reminds me of how people in the DevOps community act and speak. Gene Kim (00:22:03): ... me of how people in the DevOps community act and speak. So this is very different than how classically the dev and ops organizations behave. So classically, dev says that in order to win in the marketplace, their job is to ship the most amount of features to the market quickly. And that means making lots of production changes. Conversely, ops says that in order to win the marketplace, they need to provide stable and secure services. And that means making no changes at all. Here is the root of that chronic conflict. You can't make more changes and fewer changes at the same time. That's why dev and ops are traditionally viewed as in opposition to each other, just like that dynamics that Dave described one level down between the state department, CIA and the Joint Special Operations Command. Gene Kim (00:22:49): So Dr. Eliyahu Goldratt had a very specific term for this in his theory of constraints called, "the core, chronic conflict." The theory of constraints was the underpinning theory for his famous book, "The Goal," which has written in 1984 about a manufacturing plant manager who had to fix his cost and due date issues in 90 days. Otherwise, they would shut the plant down. So this book has sold millions of copies over the decades and has been integrated into almost every mainstream MBA curriculum. It's often pointed to, as what helped Harold in the concepts of lean manufacturing into mainstream thinking. "The Goal," of course, is also the book that The Phoenix Project was inspired by and modeled after. Gene Kim (00:23:33): Over the decades, Dr. Goldratt extended his thinking beyond just manufacturing to project management, retail, supply chains, services, and much more. Around 2007, I was able to take three graduate courses on the theory of constraints and constraints management at the University of Washington. I took these courses with George Spafford, one of my Phoenix Project coauthors, and it was invaluable for preparing to write the book. One of the things I learned was this concept of core, chronic conflicts and how they appear at the organizational boundaries. Just like Dave mentioned, it is very common that in organizations, everyone shares a goal of surviving and winning in the marketplace. But when you go one level down, you suddenly find organizations that have diametrically opposed goals with each other. I mentioned the one for dev and ops, and that is at the core of The Phoenix Project. Gene Kim (00:24:26): Here are some other ones. In plant manufacturing, in order to achieve the plant goals, we need to protect sales, so that inventory is always available when a customer wants to buy it. But we also need to control costs, which means we need to reduce the amount of inventory. Again, two valid business needs resulting in two diametrically opposed outcomes; increased inventory versus decreased inventory. And here's the one for distribution channels. In order to have effective material control, we want to have materials available to the customer, which means you want to locate material near the customer. But also, we want to predict consumption effectively, which means that we want to locate material near the supplier. So again, two valid business needs leading to two diametrically opposed actions. Gene Kim (00:25:14): In the theory of constraints, the reason for doing what they call, evaporating cloud exercise, is to keep analyzing the problem to understand how to achieve the goals on both sides of the core, chronic conflict. For instance, we now know in dev ops, that you can actually deliver features more quickly and increase reliability and stability by making smaller changes more frequently, that it is indeed possible to break the conflict. In fact, in almost every domain, we can break the conflict by reducing the batch size, moving to a poll based systems, striving for single-piece flow, reducing reliance on inventory to buffer problems, focus on expanding capacity at the bottleneck. All this potentially requires us to do work very differently than what the dominant architecture dictates, where functional silos are actually at odds with each other by design; creating conflict between them and discouraging them from working together in the way that Dave and Jess describe. Gene Kim (00:26:14): I think so many of the concepts that Dave describe show up in the DevOps community. For those of you who are interested in theory of constraints, the best resource I'd recommend is the audio book called, "Beyond The Goal, "by Dr. Goldratt. And indeed, that audio book was the inspiration for Beyond The Phoenix Project, which I did with John Willis. I'll put a link to both of those in the show notes. Lastly, David mentioned The Marshmallow Challenge. The Marshmallow Challenge was created by Peter Skillman. In this exercise, teams of four were assembled with the goal of building the highest freestanding structure out of 20 sticks of dry spaghetti noodles, one yard of tape, one yard of string and a marshmallow, and the marshmallow had to be on top. It's an exercise that requires people to collaborate very quickly. Gene Kim (00:27:04): The big surprise in the study was that kindergarten students consistently outperformed almost every other population studied. They outperformed business school students, lawyers, and even CEOs. Luckily, the kindergarten students did not outperform professional architects and engineers. Two reasons are cited for why kindergarten students do so well. The first is rapid iteration. Kindergartners just jumped in and started experimenting. They didn't spend a lot of time philosophizing about the problem. Many business schools teams ended up with a zero score. They often fell into the trap of, "Let's get it right the first time," resulting in dismal outcomes. Gene Kim (00:27:47): The other factor that decided is that kindergarteners never jockeyed for power as Tom Wuteck his phenomenal TED Talk said, "They didn't spend time arguing to figure who was CEO of Spaghetti, Inc." One other note about this experiment is that adding a high stakes prize, such as cash, actually lowered performance. I'll put a link to two of the TED Talks; one by Tom Wuteck and the other one by Peter Skillman in the show notes. I think the challenges posed by The Marshmallow Challenge are particularly interesting, given how mid-level leaders need to overcome their traditional parochial goals and act in very different ways. Okay. Back to the interview. Gene Kim (00:28:27): Dave, in your talk, this isn't the first take and it didn't make it into you DevOps about by the London talk, but it really caught my attention because use used phrases that we may not have heard before, but just so beautifully evoked feelings that we've had. So you mentioned the fantastic feeling of a leader when you know what the goal is, you know where the kind of the boundaries are and you have a lot of autonomy and then the agility and nimbleness that can unlock. Then you described the opposite, when essentially, you use the phrase, "your decision space gets pulled up." Suddenly, you have people looking over people's shoulders, looking over your shoulders, and then suddenly, autonomy is now shrunk down, essentially depriving that leader on the ground of their autonomy and the ability to independently make decisions. Gene Kim (00:29:11): By the way, that actually reminds me of a book I read by Martin van Creveld on, "Command in War." He's describing the situation, often in Vietnam, because of the risk of any incident appearing in the evening news that you had captains on a hill with a major circling overhead in the helicopter with a Colonel and a helicopter above them, the general and the 707 with the hotline to a general in the Pentagon and all of them talking directly to that captain, essentially, depriving them autonomy. Can you maybe talk about both sides? What would cause you to actually start pulling decision space up from people that you oversee? And what does it feel like when you're on the other side, trying to pull the decision space back down? David Silverman (00:29:52): Yeah, it's an interesting phenomenon. I've probably lived and been on both sides of it on all the pros and cons. So I've probably had my own mixed record, but I think it comes down to risk, to me, fundamentally and where risk sits. So for leaders, you got to get comfortable with the unknown, the uncertainty, and at some point, you have to basically let go and sort of just embrace the fact that you're not going to be able to affect something. You have to trust somebody. So that's part of it. There's two ways that realization can occur. One is the pace is so fast that it's just impossible. So you get overcome by events, meaning you have a unit in a firefight. There's just no way at that point, for you to really do anything. You're sort of just waiting to see what happens. David Silverman (00:30:35): So then what you're really counting on is you've got a leader or a group of leaders on the ground that are able to sort of figure and sort it out. You're trusting their training and their experience and their time together to do that. That was really well personified on the Osama Bin Laden raid. When you look at that picture of the situation room and you're seeing the president, the secretary state, and the chairman of the joint chiefs, these people, and you can see their face when they were just startled when they saw the helicopter crashed and people were like, "Oh my God." At least the lore goes is that Al McRaven, who was... Can't remember if he's in the room or he's on a call, he was like, "Don't worry. These guys have planned for this contingency a thousand times. They'll be fine." David Silverman (00:31:17): Sure enough, within 20 seconds, you see the team assemble pivot on their [inaudible 00:00:31:23], get back on target and within a minute and a half, the mission is still going as if it was. Now, people are like, "Wow, that's amazing." Well, it's not really. These guys rehearsed that scenario a thousand times. So there was conviction all the way up to the, in this case, the commander of the unit saying, "No, no, no. This is what these guys do." So I think that you got to figure out how to find that comfort level. The challenge with technology today is your ability to actually go down in the weeds and micromanage has never been greater. David Silverman (00:31:51): Right? But it's a false reality because what comes with that is the pace and speed is also exponentially more than what it was before. So you've got to figure out where's that balance of eyes on, but hands-off. I can see things and if I need to step in something, I will to manage a risk, but I want to do that very surgically, only I pick my spots very, very carefully where I can minimize the risk based on a larger appreciation for what's the situation what's happening, whatever else. In peace times, you let people work through that and they learn. You sort of figure out those because failure, in this case, isn't life and death, but in combat, it's tougher because the consequences are so great. So you tend to initially kind of overreact and kind of step in, and then eventually over time, you get comfortable and you start to back out again. David Silverman (00:32:40): So I think it's a combination of sort of experience and trust that lends a leader to sort of let go. Then the responsibility on the other side of the frontline leader is to basically instill confidence. So to me, speed, it was a function of credibility. Credibility is a function of delivering, what you said you're going to do. Maybe not the way you said you were going to do it, but as long as you're able to produce the result and you do that enough, that gets you credibility. Credibly creates freedom of action. Freedom of action is speed. So that was sort of everything for us. David Silverman (00:33:08): Look, I need to operate this pace because I'm closest to the problem. I know the pace is necessary. You got to give me that latitude. They're saying, "I'll give you the latitude when you can validate that you can consistently perform standards on a set of tasks," right? So there's a mutual benefit how to do it. Where I think junior officers or senior NCOs in the past, potentially could find themselves in trouble is if they were sort of dismissive of authority or the process, then all of a sudden, they would lose that credibility that would then tighten down the speed and then the whole thing sort of becomes a less than ideal situation. Gene Kim (00:33:40): Can you speak to maybe a situation where your decision space is being pulled up and what it takes to actually bring it back down? I grounded it in the general belief that you actually have the point of most apical information. David Silverman (00:33:50): Yeah. I won't use names, but when I got to Iraq in 2005, at that point, it was already very sensitive for US or coalition troops to be going into sensitive religious sites or targets. The enemy had adapted. They were moving weapons, potentially operations and people into these sites because they knew it was a political hot button. If you went into to a mosque and kicked down the door and started detaining people, it sends a really negative message, as you can imagine, to local community and therefore, where you may not had a problem in a community, now, you've got a problem. So the whole idea was like, "Okay. We got to figure out how to send local law enforcement or military units in, even if it's a sensitive target, vice the US." David Silverman (00:34:34): So I happen to be partnered with the premier Iraqi counter-terrorism force. They were our partners. So we had the ability to send Iraqis, Muslims, and specifically, the type of sect into the appropriate location. So if this was a Shia or Sunni, I could even tweak the people going into that demographic, but it was still we were involved and either way, if something bad goes on where, we're going to get some blame. I remember, I had been chopped from one commander who I basically trained up with to a different commander for operational purposes. I had gotten this mission approved. In this case, it had to go way above my immediate commanders for approval. In this case, I think it was the core commander, which is the guy in charge of all ground operations and the entire country, three-star general. David Silverman (00:35:20): My administrative boss at the time, who happens to be overseas operations, joined other people without necessarily me at the time says, "I'm not comfortable with this op. I want you to stop." And I said, " Okay. Well, if you want me to pull the Navy guys off of this objective instead of down, I can, but the army guys are going to go and so is the Iraqi unit because they've already got approval." He looked at me because he's like, "Look, I'm just not comfortable, what this says." I said, "Look, I can promise you we've rehearsed it. I've got the right team that's going to go in. We're not going to desecrate or do anything silly. Let me have this opportunity. If we get Viet wrong, I'll never ask for it again." And he was like... He begrudgingly said, "Okay," and the op went fine. David Silverman (00:36:05): It was totally fine. To be fair, most of the ops too, most of them if you do, if you do a good job, there isn't any really problems. The ones that go wrong is the ones you hear about. I remember he came back to me afterwards and he apologized. He's like, "Hey, you got. I was wrong. You've reinforced my trust." So it doesn't mean we didn't make mistakes and we made a ton of mistakes, but I think he had to have the conviction to competence. He was new to the scenario. We were new the scenario. So there's a lot. The risk was real. So I think it was a combination factors, but I chalk that up to two things; experience, both in this case, the theater, and then really being able to quantify and manage the risk. In that case, I think I had a plan for the second, even though we didn't really have the experience to validate maybe some of that hubris, but in our case, it happened to work out over time. David Silverman (00:36:51): I think we built confidence and more importantly, we started being leveraged by other units to do that. So we were getting flown all over the country, which was the whole purpose to basically solve, in this case, a British commander's problem set because he's like, "Hey, these guys have actual locals that can handle this problem about inciting a riot." There's a lot of parallels if you look at what's going on today in law enforcement in the US with around racial injustice and how you become a responsible community actor, and you see people who maybe don't necessarily spend a lot of time in that community coming in that community and potentially using something they don't understand and then people get stressed and someone makes a mistake. Then all of a sudden, it becomes a national situation. So some of that could be addressed by, I think, better communication and then leveraging people from that community to help solve some of their problems. Gene Kim (00:37:39): Gene here. Both Jessica and I thought it was pretty funny when David mentioned that micromanagement is something that the technology community is very good at. In The Unicorn Project, there was a phenomenon called, the square. Teams were often unable to make decisions to do what the customer needed and had to transit the square. In other words, to make a decision, you had to go up to over two and down two, and then the return path in order for two engineers from different silos to work together to do something on behalf of the customer. That's because the structure didn't have any official channels for those engineers to work together directly. When something went really wrong, like in the aftermath of the deployment failure of The Phoenix release, the square got even larger. Now, every decision had to go up three levels and then down three. In other words, the decision space got pulled up in this case, by Sarah, the SVP of retail operations, for exactly the reasons that David described; the lack of trust that the teams were actually doing the right thing and the fear that something would go wrong and impact her. I love the precision of which David is describing some of these scenarios that we like and don't like, the notion of decision space being pulled up, which is not desirable and on the opposite side, the concept of eyes-on, hands-off. These are both concepts that are used in the military that describe with surprising precision, the structure and dynamics of situations that we've all been in, whether desirable or not. I'll have more comments on why I think this is important later. Okay. Let's go back to the interview. I had asked Jessica about her thoughts on whether and in what forms micromanagement takes place in the technology space and what we can do about it. Jessica Reif (00:39:25): Yeah, absolutely. So I would say in terms of whether you are a micromanager and you're trying to see the light and managed differently or you are being micromanaged, I think that one of the first things that someone on either side of that equation can do is try and pull out what the reasons are for the situation. As an example, someone per Dave's example, could be micromanaging because they believe that there's much risk in having the individual who is executing, they believe that there's risk and the manager feels the need to swoop in to ensure that whatever that piece of work that the individual is doing to ensure that it goes smoothly. Whereas there are other cases where the manager's not even worried about risk. It's really just because they're a control freak and they like things done a certain way. Jessica Reif (00:40:10): The code could look like this or that, and it's going to run the same thing and the manager might have a preferred way that it's done. If it doesn't line up with that, the manager wants to see it changed. The solution to that problem is different than de-risking the operations on the edges. It's something that the manager needs to overcome. For lack of a better term, being a control freak. In other cases too, it's that the manager enjoys the job. The reason that they got promoted is because they were so good at the job that they used to do and there's part of them that says, "I still want to be cutting code. I still want to be in those spreadsheets." And instead of seeing the information reported back to them, they like to deep dive in the numbers because it's something that they genuinely enjoy. Jessica Reif (00:40:53): So these are just a few of the reasons why people micromanage. I think that getting to the root of why it's happening is effective, no matter which side you're on of the equation because it helps you work to a solution that works for both parties. Gene Kim (00:41:11): I'm so excited that IT Revolution has published a new book from one of my favorite authors, Mark Schwartz titled, "The Delicate Art of Bureaucracy." As always, you'll be able to appreciate his background in philosophy, history and a career that included being the CIO of the US Citizenship and Immigration Services. As a leader of a formidable bureaucracy that detained, he teaches everyone on how to create a new bureaucracy; one that is lean, learns and enables. Gene Kim (00:41:43): On November 10th, look for John Smart's debut book, "Sooner, Safer, Happier." In it, he reveals patterns and anti-patterns to show how business leaders from every industry can help their organizations deliver better value, sooner, safer and happier through high levels of engagement, inclusion and empowerment. You can pre-order the hardcover version now. Following up on a fantastically successful DevOps enterprise summit two weeks ago, we've finally launched our new video library. You can watch all of the incredible presentations from both the London and Vegas virtual DevOps enterprise summits, as well as purchased unlimited access. You can find both, Mark Schwartz's John Smart's presentations there. Go to doesvirtual.com/ virtuallibrary. Gene Kim (00:42:36): If I listed a second example where you have a leader who sort of misses the work and maybe just kind of gravitates towards the work and by doing so, kind of accidentally puts him or herself in the approval path of basically everything, what are the concrete pieces of advice that you might give that person? Jessica Reif (00:42:53): Yeah, absolutely. So I think that the one thing that they can do is they can make a clear differentiation between when they're wearing their brainstorming hat and when they're wearing their, "I'm the executive decision maker for this project, I'm your boss," hat. So as an example, if the three of us are a team and Gene, you're our leader and we're brainstorming on a project, if you express an opinion and you express that, that opinion is, "This is just my 2 cents, but we should integrate that with the other things that we've been talking about," that's much different than saying, "This is the way I think it should be done." Jessica Reif (00:43:24): So I think that managers can benefit a lot from differentiating that they're speaking as a team member, rather than as the leader of the team or that they're offering a suggestion rather than making a demand. I think that just differentiating between those two is really important because oftentimes, the person on the receiving end can't necessarily differentiate. So it's safest for them to assume that it was a demand and to just make the change, even if they perhaps have information that would suggest that what the leader is suggesting isn't the best approach. David Silverman (00:43:55): The other thing I would add is that a suggestion to leaders potentially struggling with this is I think it's important that they sit down and figure out what are their actual roles? David Silverman (00:44:03): I think it's important they sit down and figure out what are their actual roles and responsibilities? What are those things that they have to do to be successful in this job? What are those things they like to do? And what are those things that they should be empowering other people to do? And what you want to then do is you want to say, "Hey, you need to be allocating your time to those buckets in order of that I just said. What you have to do, and only you can do, because there's nobody else in that team is supposed to be doing that. So that might be probably managing expectations up or going external and dealing with customers and clients. And while you may like to spend time working on a piece of code or a piece of design or whatever else, if there's something that you have a really strong bias for, let's carve that out so people know." David Silverman (00:44:40): And then the last thing [inaudible 00:44:41] says, "You know what, if you're doing this, you're not doing your job. So who's doing your job if you're doing my job? And I think it's like one of these self-awareness tests that people should audit on a regular basis. I would advocate, with my senior executives, they do that quarterly. They get someone to say, "Show me how I spent time over the last quarter in these larger buckets. And is that appropriate? And did it produce the results you wanted?" If you think about the reason why it happens, it's less nefarious than it can potentially sound. Because to suggest this point, the likely reason why you got promoted, especially if you're a subject matter expertise in a certain field, is because you weren't really good at it before. David Silverman (00:45:21): And so then if you go up a level and you look down and they're doing it a different way than what you know, created success, it's going to be a disconnect. Because you're like, "Okay, well, hold on a second. I don't have the confidence or conviction that that's going to produce the same result. You got to gain credibility with me before I let you just do that." And so I think it is natural and it's hard to do. So I think some of that has to be little patience with the subordinate as well saying, "Hey, I'm not going to do it that way because that's not my style, but we need the same outcome obviously, or a better outcome. And so let's figure out how to get us both comfortable." David Silverman (00:45:54): And then once you get through that dance and do it a couple times, then you get like I said, that freedom of maneuver. You got credibility, so now you've got freedom of maneuver. But it's really on the leader, in my opinion, to initially set what's going to make us successful and what's my role in that, and where can I go focus my time on that. Because a lot of times it's easy to get distracted. And if you don't have some mechanism to hold you accountable, you run the risk of focusing down and just pissing off your team because you feel like you're doing their job for them. Gene Kim (00:46:24): Kind of doing a little self-assessment in my own time, I'm kind of like, "Uh oh, maybe there's a little bit too much in one bucket that I shouldn't be..." David Silverman (00:46:32): Yeah, it's a good drill. And it's really, really helpful. Gene Kim (00:46:35): Gene here. You could definitely hear me laughing when Dave mentioned auditing your time to make sure you're spending time doing the right things. And the reason why I laughed was because of this. I think anyone good at what they do has audited their time, but probably just to make sure they're working on the top project priorities. But I hadn't heard of someone auditing their time to make sure it was appropriate to their levels of leadership, or at least not in about 15 years since I worked with an organizational coach. I love that phrase, are you doing the things that only you can do? And if you aren't, you're doing someone else's job and that's a problem. And I also loved the acknowledgement that there are activities that you really love to do, and that one has to make allowances for that. After all, we want to love our work. Gene Kim (00:47:24): I think these concepts are especially important. Before, in a world where a leader's job is just to manage a silo and to be functionally excellent, in those days you could probably get away with doing the work that you've always enjoyed doing for a very long time. However, in a world where endeavors require the efforts from a much vaster surface area of expertise, from a much larger cross-functional set of teams, this behavior seems increasingly at odds with what the organization actually needs. Gene Kim (00:47:52): By the way, I have a sudden memory from 15 years ago, watching a chief information security officer insisting that they review every new rule for their intrusion prevention system. I suspect that many of you will find this just as curious as I found it way back when, and now I know why. That chief information security officer was probably doing someone else's job. Personally speaking, I will definitely be going through an exercise of making sure I write down what I think my job is and what I absolutely must be doing and what things I definitely should not be doing. Okay. Back to the interview. Gene Kim (00:48:27): So one of the things that came up in one of our previous conversations was how much innovation was happening at the edges during the [inaudible 00:48:36] teams era. So one of the notions in the Toyota production system is, problems really should be solved closer to where the problem actually is. And you had mentioned that you had these experiences where there was a dramatic expansion of capacity around certain things that could go on intelligence gathering platforms. And I think given the trend towards increasing specialization, the need for expertise at the edges, that's going to keep going up, not down. The example that you gave was such a dramatic example of things that could only be done at the edge. Can you talk a little bit more about that? Because I think it's very useful for the technology community, because I think it validates certain intuitions that we have. Kind of give us proof that these things are actually true. David Silverman (00:49:16): Yeah. I mean, in any large entity, if you're trying to produce a specific outcome, there's probably different components that lead to that outcome. And oftentimes they might be siloed, they might be geographically dislocated, whatever else. It's almost like this value chain that adds up to this piece. And for us as operators, we were going on objectives every night. And if something wasn't working, our tolerance for using it or continuing to execute that tactic was effectively zero, right? And so they could create the coolest damn cell phone on the planet. But if you couldn't get operators to carry it, it didn't matter. You weren't going to solve the problem. So there had to be a much tighter coupling of the persons that were doing some of the initial research and development, maybe some of the design and development work, and the end user. David Silverman (00:50:04): Now everybody talks about that in the military, right? And they'll go through these very long procurement cycles where they're going through some, and for very complicated systems that make sense. If you're trying to put a man on the moon for the first time, that's not something you can just iterate off of a couple of times, right? It just doesn't work. But in our case, because what we were really trying to figure out was, people were all now connected through digital means; how do we basically track and identify and locate the right people at the right time so that we could do something to disrupt it? David Silverman (00:50:34): And so in that case, the technology was advancing very, very quickly, similar to how I would say more commercial technology advances today. And that was very different from the way the military did stuff, right? So you'd have science experiments happening in somebody like DARPA. They might get picked up buy an operational unit and validated. And then five years later, you'd see it in the general purpose force. Well, now all of a sudden you've got people pushing out new devices on the commercial side every six months to 12 months, new iOS patches and code potentially every day, every week. And people creating a whole set of different apps for collaboration, communication that people can pivot across. David Silverman (00:51:08): So it was like, "Okay, this is like making your head spin." So we had to figure out a way to take our smart people and basically have them match the pace of what the consumer world was doing, and doing that through the lens of an end unit like myself who was operationalizing. So pretty quickly what you saw was, you'd go into a ready room, which is like the place where people come together to get ready for an operation. And instead of just seeing a bunch of people that sort of looked and felt like me, all of a sudden you'd have this Star Wars bar of personalities and backgrounds who are all, in this case, trying to figure out like, "Well, this will work, and that'll work," and you're like, "No." David Silverman (00:51:44): And what was interesting was before that would have been a very, very siloed conversation. It probably never would've happened. And the people, let's just call them the geeks, they would never have been in the room. The guys would have been like, "I don't have time for this, you go do whatever you're doing." Now, we were super interested. I remember having so many conversations where it was, "Tell me how the thing works and give me the science behind it. And yeah, I don't have a degree in what you're talking about and I'm not that smart anyways, but I'm trying to figure out how I'm going to be able to leverage that here and somewhere in this conversation, there's a solution." David Silverman (00:52:14): And sometimes what you did is said, "You know what, you're coming on the target with me, I'm bringing you on the objective with me." And they're like, "Wait, what?" You said, "Yeah, yeah, yeah. Your little device that you're the only guy knows how it works, you're you're coming with, you're going to hold it up." And then they were like, "Oh." Right, so it really closed that gap where before we never would've done something like that. I mean, I was bringing interpreters on targets, that people like me, "No, no, no. I sit in a room and listen. I don't go on objectives." David Silverman (00:52:38): And people are like, "Well, Dave, you're crazy." I'm like, "No, I'll figure out how to keep him safe. That's the easy part. No, relatively easy part. But we need to get that institutionalized." And then once the boy saw, "Hey, wow, that really improved our op." Now all of a sudden people were sort of breaking down some of those stereotypes and they started saying, "Well, I'll go spend more time with that person." And that's when you start to see the real creativity. David Silverman (00:52:56): So how do you get closer to the end user? We use data now, like when you look at software companies, they track all these things and they can basically use that as feedback. And I think that's a huge piece of it obviously, right? Because you can say that's unbiased and it solves a lot, but I think it's good to actually go spend time physically going through the user experience with someone who's got a problem. And I think the more you're able to do that, you're able to kind of close that loop. Gene Kim (00:53:20): But just two quick things. The worst moment in my professional career was actually watching, this was in 2006, watching someone use our product. It was two hours. And I felt like I was literally going to throw up. I mean, someone was basically apologizing for two hours saying, "Oh, sorry, I know there's probably a better way to do this," but the whole time I was thinking with the engineers like, "No, there's no better way to do that." We had no idea. But it gets really to the second point is what leadership support was required to have this, I would imagine as you said, unprecedented degree of integration between suppliers, designers, engineers, with the people ultimately using this in the field? David Silverman (00:54:02): Yeah. It's a delicate balance, right? Because you want the guys basically resting, lifting weights, eating, and then rehearsing. Right? And so that's what's you want them focused on. And then in my role it was kind of everything else, right? And my assumption is we would come together for the actual op, or in some cases I would just send them on the op, right? So you got to figure out where the natural integration point is, because if you have too much of it, then it's distracting, right. If you don't do enough of it, then they won't use it or they won't bring the person or you won't get the value seen. So it is more art than science, but I would say that to me became the role of mid-level management. It wasn't the rehearsing and actually doing this stuff. David Silverman (00:54:39): I said, "Hey, look, I've got a leading petty officer that can run circles around me operationally. He'll figure out the best way to take down this target tonight." And once we get to the target, my job is everything else, right? Do we have the right close air support? Do we have the right contingency plans in place? Do we have the comm networks set up? Are we leveraging the right intelligence gathering apparatus? So that when the door was breached, everything after that was, the probability of success was going to be really high because I was very confident in their ability to do that. So it's on the seams and the margins, we had to figure out how to blend those, and I think that's became the focus. David Silverman (00:55:12): Going back to the earlier conversation, what are you focused on, I'd see a lot of officers like saying, "Well, I need to be on the objective and I need to be calling the shots." And I'm like, "Okay, well who's doing all the other stuff that I was just talking about?" And they're like, "I don't know." I'm like, "Well then, your chances of being successful is going to be pretty small. And do you really need to stand behind the point man and tell him which direction to go?" I'm like, "The dude's walked point for eight deployments, you've got two deployments. Like, what are you talking about?" David Silverman (00:55:35): So, "Well, he's never been a leader." I said, "Well, it depends how you define leader. Let's give him that until he proves us wrong." And so that's what usually worked out those kinks in the workup, before the deployment. And then that's when all of a sudden you saw people say, "Oh wow, I can take ownership and I've got it." And that's when again, productivity at the local level started to really unlock is when the smaller team leaders, you said, "Hey, you got this." Gene Kim (00:55:59): Okay. So Jessica, what do you make of all of this? Even that last story. I mean, can you just sort of describe your reaction because it sounds like a very different way that leaders show up and choose what is actually the most important. Can you maybe just react? Jessica Reif (00:56:16): Yeah, absolutely. So I think Dave brings up a great point. Those managers who say, "I need to be there. I need to make these decisions." That probably signifies that there's a problem with how they're allocating their time or perceiving their responsibilities. One of the things we see a lot as consultants is executives who feel like they have a lot of decisions that need to be made, a lot of time sensitive decisions that need to be made frequently, that they're constantly making time sensitive decisions. Jessica Reif (00:56:41): And that's usually a pretty good sign that they haven't necessarily gone through the drill of assigning out decisions based to other people, that they're not figuring out how to systematically empower somebody else to make those time sensitive decisions. Because no self-aware executive truly believes that he or she is in the best possible position to make all of the decisions. At least none that we've had the pleasure of working with. Jessica Reif (00:57:07): So it's really just about having that one, self-awareness to realize that you can't make all the decisions, and two, the processes and mechanisms in place to distribute that authority in a way that somebody does feel the responsibility to make those decisions other than the formal leader, whoever's in charge. David Silverman (00:57:25): Yeah. And increasingly I think the goal of the leader is to make as few decisions as possible. Right? It's when someone comes to you to say, the first question when they start talking is, do you see internally, is this something that I need to make a decision on? And then you might say that out loud like, "Are you asking me to make a decision on this? Or you just want some advice or some coaching? I'll help you think through it, and I'll put on that thinking hat and we can iterate, brainstorm, but then you go off and you make the right decision." And that also for someone who doesn't have the trust that maybe there isn't the credibility established, maybe they don't know if their teammate's the right teammate yet to be empowered with that, maybe that's the drill. The drill is, "Hey, let's do some scenario planning. Let's just talk through and talk about how we thinking about this." David Silverman (00:58:04): And once you think, "Okay, the way your logic processes is functional and making sense, you're likely going to get to an appropriate outcome," then you can say, "All right, well then off you go." And if you don't have confidence, you watch, right. It's sort of like watching your kid says, "Hey, dad, I want to cross the street today, and go down the market by myself." And I go, "Okay, you're five. You don't listen to anything. And there's like two major roads you're going across." I'm like, "Okay buddy. Let's just talk through what are you going to do here? What are you going to do there?" And he goes, "All right, sounds good." And then because he's five, you're watching from like 20 steps behind so that you can feel like you'll grab him. And then eventually you're like, "Yeah, yeah, go to the corner store. You got it." So I think you could follow a similar track to get to that place where you feel comfortable going. Gene Kim (00:58:48): What I found so interesting about your story about you as a middle manager, having to focus on these seams and integrations with boundaries that were so much vaster than before, those all seem very reliant on these other skill sets that you talk about. Connecting self-awareness, discipline, decision-making, effective communication, constant learning. Can you talk about that as a whole? Like why are those so important? David Silverman (00:59:09): For me, it was all really around winning. As a leader, you just wanted to win. The only thing you wanted more than to win was not to lose. So I don't know how to describe it any differently, but when you're part of an organization that's accustomed to a culture of success, you sort of value that over everything else. And so it allows you to be open to like, what's going to enable that to be the most effective. And so if you find yourself making local decisions a lot, then my guess is you're going to get overwhelmed pretty quickly. And in our training, that was the sole purpose of the cadre, when you were leading up to deployment, was we're going to keep upping the complexity of the scenario until you have cognitive overload and you can't react, and then something catastrophically fails. And then you learn from that. David Silverman (00:59:53): And really what that drill was around is figuring out how to let go, and to be able to rapidly prioritize where you're allocating your focus and energy so that you could have the most decisive effects. And so for me, that requires self-awareness. You have to be aware of how you're showing up, what's going on, how it's driving effects, or you're going to get overwhelmed pretty quickly. And more importantly, by the time you figure it out, which is the self-awareness part, the problem's probably already passed you. David Silverman (01:00:17): Communication and connecting was so important because for me in any complex problem, one of the hardest things to figure out, because if it's not a hard problem than somebody's probably already solved it. And if they've already solved it, there's probably a best practice, and you should just follow the book and the best practice. But nobody wins any medals or awards or gets big promotions for that, right. That's what's expected of you. So in this case, go tackle something difficult and hard, which again, most high performing people want to do. David Silverman (01:00:42): So in order to do that, usually what happened with the scenario is there's a perception or there's a concept that seems illogical or seems irrational. So you have an actor doing something and it doesn't make sense to me why they would do this. And somewhere in that is sometimes the root of the problem. And that can be as simple as actually trying to figure out how to take yourself out of your shoes and put yourselves in their shoes. And then all of a sudden, because I believe that most things happen for rational reasons with people. I believe that, even if it appears on the surface to be irrational. And then if you can get yourself in their shoes, all of a sudden you go like, "Oh, that's why they're doing that." David Silverman (01:01:17): And then you can start figuring out how to connect and solve problems. And so that's important motivating your own people, but also trying to figure how to solve a problem, assuming the problem involves other people, right? In this case, a consumer or a customer or whatever, you have to really figure out how to get in their shoes and understand their perspective. And that makes a huge impact on how you go about and tackle the problem. David Silverman (01:01:36): And the last one to me was just discipline. I mean, all the other ones are important, decision making, communication, and constant learning is a big one, how do you learn? But I'm assuming in this case, you already have a propensity to want to learn and get better. But discipline was so important because it was about taking habits that you knew worked and making those routines, so that it didn't require any effort, like almost muscle memory. Like you could do it in your subconscious. And that way, it gave you capacity for all the other stuff. David Silverman (01:02:03): And if you don't have that, if you don't have that rigor, that discipline, the fundamentals aren't in place, then your ability to deal with a changing condition is compromised because you're still trying to figure how to do the thing you're supposed to already know how to do well. And so that plays in two ways, one just yourself and how you're personally performing. But two, if you think about that, as far as how you're managing a team, if your team consistently says, "I don't know which Dave is showing up today," Then you're creating anxiety for them. And they're probably creating two contingency plans on how to deal with you, depending on what mood you're in. Or how you show up, if you're going to be late or not. David Silverman (01:02:32): And so that level of rigor and discipline creates consistency, and the consistency allows people to soon normalize. So the analogy in sports is like, there's no such thing as a bad referee that calls the same calls, right? So if someone has a strike zone, which you don't consider to be strikes, but they call it consistently, a good hitter will adjust. They'll be like, "I know this one out here is going to be a strike. I got to swing at that." But if they're calling it randomly, they call it a strike one time, then the same ball same place they call it a ball next time. That's when you're like, "Oh, geez, I don't know what I'm doing here. I got to swing at everything." David Silverman (01:03:03): And so if you think about that, that creates for a bad environment. It makes it tough for the employee base to adjust and for them to figure out their own routine. So I'm a big fan of routine because it lowers a cognitive load. And some people associate that with process, and process is bad because it creates rigidity. I think it's the opposite. I think discipline creates flexibility. I think it allows for capacity for innovation and creativity. And without it, you're basically just going to get lucky or not, in my opinion. Jessica Reif (01:03:29): Yeah, absolutely. So one of the skills that Dave hit on a lot is self-awareness, and at least in my opinion, that's probably the most important one when it comes to leading like a guard or in any senior leadership position, of course you have to be functionally excellent. If you want to be a respected CTO and you're not familiar with technology, you've never written a line of code, it's going to be difficult to get respect from those you manage. But assuming that you have that functional excellence established, then really all the other skills become more important. And self-awareness is really what helps you toggle between the rest of the skills that you mentioned. It helps you understand if you need to be applying a process in any given moment, or maybe you need to be a lending an empathetic ear, to hear someone out, to hear a situation. Jessica Reif (01:04:15): I also think that self-awareness plays a unique role in helping us be innovative as people and see situations differently. One of the things that blocks us often from thinking creatively is that we have perspectives and those perspectives are something that exist well below the surface. They're not necessarily something that we can readily articulate, yet they block us from seeing solutions that could exist because we have a limited way of seeing the world. So a more self-aware individual is able to better articulate their perspective. And they're better able to take on the perspective of somebody else or someone who sees it a different way. And perhaps in that case, unlock some creativity and innovation as a result. Jessica Reif (01:04:54): So I'm with Dave on discipline. I think that discipline is incredibly important. And to Dave's point, it's not just about having processes and routines. It's about making the things that are routine, making them feel automatic so that you're freed up to think about other things and to focus on the situations that are really changing in our dynamics. So the example that we give a lot with clients when we're introducing this concept is if you think about the first time that you tied your shoe, or maybe you taught your child to tie his or her shoe, it's hard. They struggle. It's not easy to do that first time. And it really requires all of the mental energy that you have, but by the time you're an adult and you've tied your shoe a million times before, you can do that, you can think about other things, you can carry on a conversation. It just becomes automatic. Jessica Reif (01:05:39): So for those parts of our day and our jobs that are truly automatic things that we can do relatively easily and in a straightforward way that don't really require a lot of adaptability, it is appropriate to make those part of our routine so that we're able to free up our mental energy for that increasing part of our day that does require that adaptability and mental flexibility. David Silverman (01:06:01): Yeah. And just one other thing to build on that... Jessica Reif (01:06:03): Ability and mental flexibility. David Silverman (01:06:03): Yeah, and just one of thing that built on that, a lot of people talk about grit and resilience as a preeminent for success. A lot of books written on this, a lot of studies that say, "Hey, that's probably even more important than actually skill is this the ability to persevere." I associate that with discipline, right? So like when you're going through SEAL training, the whole goal of that training is not really teach you anything. It's only to make you super self-aware of where you mentally break. And they're forcing you to validate that you can mentally push past a physical barrier over and over and over again and still see it through it. Because the reality is you're capable of so much more than what you're personally feeling, like when you're physically are failing, your body is actually capable of like ... you're probably only 20 or 30% of the way there. There's so much more that you can actually accomplish if you mentally just figure out how to stay and stay focused. David Silverman (01:06:53): But to me, what enables that is having the ability to compartmentalize because you know that the fundamentals in the compartment are fine. you don't have to worry about that. You can kind of go back to those. And then that gives you the ability to keep just pushing through on what you're doing. A lot of sports and athletes have like, there's a lot of science behind how they do this, but ultimately if you take the physical out of it, it's pretty similar with just the mental. Is how do you make routine routine and become successful. Gene Kim (01:07:22): Gene here. Dave and Jessica talk about the specific skills that mid-level leaders need to have. I like the fact that they talk about these as skills. In other words, they can be taught and learned. As opposed to being attributes that you're either born with or you're born without. And this is actually a somewhat recent line of thinking in leadership theory. Gene Kim (01:07:44): In the late 1800s and early 1900s, the major schools of thought around leadership were called The Great Man Theory, Trait Theory, and in the 1950s Behavioral Theory. What most of these shared was the notion that great leadership was innate, meaning it was inborn or natural. In other words, you were either born with these skills, which is great for you. But if you weren't born with those skills, well, you'll never be a competent leader. However, in the 1970s there emerged other schools of thought, including servant, charismatic, and transactional leadership theories, and then transformational leadership theories, which state explicitly that these are teachable and learnable skills. Gene Kim (01:08:24): Both Dave and Jessica talk about these skills as teachable, that one can deliberately improve them with practice. I'm going to put a link to a presentation that Dr. Steven Mayner gave at the DevOps Enterprise Summit 2017 on the topic of the history of these leadership theories, as well as one that he did in 2018 about learning to be a transformational leader. Okay, back to the interview. Gene Kim (01:08:47): I have gushed to both of you just how much I admire the work that you've done, both in the book Team of Teams and the cross lead framework. Talking with Stephen Spear, we have the notion of structure and dynamics, but also dominant architecture. Dominant architecture is really kind of the way we've always done it. It's the primary way, that modality that we do things. So my friend [Mike Nygaard 00:03:06], he said, "In most organizations it's like geologic strata in reverse. The higher you go, the older the methods, the older the ways of thinking, the older the models." Gene Kim (01:09:16): And so there's no doubt that Team of Teams has opened a lot of eyes and certainly been read by some of the highest levels of leadership. On a scale of one to 10, to what extent do you think Team of Teams has affected the world outside of the special operations community? One is everyone's read it, but probably hasn't impacted the daily work of within the organizations. Or 10 is you see evidence of changes in all aspects of people's daily work, whether it's strategy, planning, operation, and so forth? How would you characterize to what extent these learnings have been incorporated? David Silverman (01:09:48): I would be remiss if I thought that the book was the reason why people are adapting these things. I think what's natural is this is like the new way of the world. And people are either going to evolve to this on their own accounts, or they can kind of maybe navigate a little faster by learning some lessons from like Team of Teams and other books like it. So I think the trend though is clearly in this direction. So I would say it's probably like a five on the scale of where I see the progress. Like five years ago, six, 10 years ago when we first started this, I would say it was like a one. The idea that culture and adaptability, being the preeminent thing that you look at and try to create and institutionalize in your organization was sort of fundamental. I think was like isolated to people that were solving specific product problems. And they were dealing with probably code or some product that was dealing with a consumer that was changing pretty quickly. David Silverman (01:10:43): Now, I think institutions are starting to realize that, "Wow, the strategy that we just paid all this money for externally or have come up with is no longer valid. Our assumptions are wrong." And they can say, "Well, it's a Black Swan event." But increasingly, like the hundred year flood is coming like every year. And so it sort of forces people to re-imagine where things are at. So I'd probably put it at a five in the evolution. And I think the pain is now ... I think the pain now is in people's conscious. Gene Kim (01:11:12): It's interesting. So used to be one, now at five, but heading towards eventually ... David Silverman (01:11:18): Yeah, I think it's adapt or die. I think people are going to figure this out or go away. That's my sense. Gene Kim (01:11:25): Interesting. Jess? Jessica Reif (01:11:27): Yeah I'm more bullish [crosstalk 01:11:28] I would, I'd say eight, maybe eight or nine, something like that. There are certainly elements from the book that don't necessarily translate perfectly to a civilian setting. As an example, we haven't seen a daily 90 minute meeting that takes place 365 days a year. We haven't seen that takeoff with any of our clients. But that example has caused many of our clients to rethink their existing operating pace, and in many cases, speed it up a bit. So it might've been meeting quarterly and they're meeting monthly now. Or they're meeting every two weeks or every one week. And in some cases, every day for a shorter period of time, every weekday. Jessica Reif (01:12:02): If we were being critical, I would say that Team of Teams is a really convincing before and after picture of what an organization can accomplish by transforming itself. But it's not necessarily an actionable roadmap for doing so. To your question on strategy and the areas where we've seen people making big changes, I would say with strategy and planning, we've worked with a number of companies who've really abandoned their old ways, their old traditional planning practices in favor of adapting to rolling plans where they're adopting a four, six, or eight quarter rolling plan in lieu of that typical annual plan. And I think that's a reflection of people recognizing that they can't see a year into the future, as well as they maybe could have 20 years ago. Jessica Reif (01:12:47): And similarly, I think we've also seen broader acknowledgment of the coordination challenges that happen when you are doing complex work in the context of a multi-team system. It's not enough to just have high performing teams. All of the connective tissue that makes those teams operate effectively together really needs to be in place and under-emphasizing that is going to lead to a collective failure. David Silverman (01:13:11): And let me give you another example, Gene, where this has come up recently for us. So after the DevOps Summit, we've connected with some of the leaders in your space like Scott Prugh being one of them. And talking to him, his perception on how you unlock some of this value more high-performance across these lines, was coming at it from a very different angle than how at least me personally have approached it, which was this concept of structure operating structure. Similar to what you said before with Spear, there's an optimal structure that will naturally unlock value. David Silverman (01:13:41): And our bias was that was like a TV set for somebody that lived in the jungle. It was like, well, we couldn't address structure. So I never thought about structure, truly thought about structure because it was off the table. So everything we were doing was trying to figure out optimize with that constraint in place. Where in the private sector, that's not the case at all. If you clearly see that you have known dependencies that exist, and they are spread across multiple different units, that is going to create tremendous inefficiency. So his removing the blinks or moving the handoffs was like, "We'll just organize them in one value stream. And then they'll naturally figure out how to pace themselves using agile principles and practices that will unlock consistent sustained value." David Silverman (01:14:28): And I'm like, "Hmm, maybe there's something is to structure." Because we were always saying, "Well, no, you can't change that. And so therefore we've got to figure out how you operate independent of your structure." Almost- Gene Kim (01:14:41): Shadow structure. David Silverman (01:14:43): Yeah, exactly. And so, most of what we were talking about was like, "Look, I don't really care what your org chart looks like, I just care how you operate." That's what I say to clients all the time. And they're like, "Well, the org structure defines how you operate." I'm like, "Well, it doesn't have to." It didn't for us. I mean, that was the big aha. This didn't have to define us. The fact that you work over here and I work here doesn't mean we can't collaborate on an issue where we're going to find mutual value. David Silverman (01:15:06): And inevitably, coming on the heels of other firms that come in before us, a lot of times what they're doing is trying to drive efficiencies. And so they're trying to take cost out. And so they'll do a lot of structural redesign work. And then what's left is this organization says, "Okay, I'm optimized for this new structure. Theoretically, it makes sense in this scenario." Well, the problem is this scenario tends to change. In our case, like we've said, changes pretty quickly. So now the structure you just reorganized to, which I will tell you, is typically debilitating for the organization. It's really tough. Anybody who's been through a massive org structure change. David Silverman (01:15:37): And so then you're like, "We need some consistency." And then they're reorganizing a year and a half later for some new contingency. Eventually, you get in this cycle of well the only person making money here are lawyers and consultants. Everybody else is just in the hurt locker. So, we've always said, "Hey, don't worry about that as much." Like, yes, if you can address your structure and optimize it, that's great. David Silverman (01:15:57): But our thesis is your value streams are probably going to move around because your customer's moving around how they value stuff. And so you got to think about how do you have resilient systems that you can put in place that hack any given structure that allows you to aggregate and self-organize around a mission-critical item as quickly as possible. And then when that item is no longer valued, it can dis-aggregate without having to go through some massive org structure change. That's been our bias. David Silverman (01:16:22): But there's probably a middle ground. And especially when you start thinking about how you allocate capital, which [inaudible 01:16:29] in terms of prioritization, certainly we get to a higher level. We see a need to clarify some of those structures that drive some natural synergies. Because if you get to a certain scale, you can't possibly be self-prioritizing across multiple value streams effectively as a senior entity. It's too hard. Jessica Reif (01:16:45): Just to pile onto that. So we do a lot of organizational network analysis, which is the process of identifying the off the org chart interactions. And it's really amazing because sometimes the off the org chart interactions really closely near on the org chart interactions. That's what tends to happen in the more hierarchical organizations that we've worked with. We see information flows up and it flows down and those are really the only two directions. There's not too much happening from a lateral perspective. Whereas, we've seen other organizations where the networks that form are very, very different than the actual org chart itself. And in that case, it doesn't necessarily matter what the structure is because they're operating differently. Their culture is that they connect horizontally. And whether they're formally on the same team, they're formerly on different teams, they're in different departments, those people are still going to interact. They're still going to learn from each other and be able to collaborate. Jessica Reif (01:17:42): So, so to Dave's point on structure versus dynamics, our strong bias tends to be towards the dynamics. Not just because structure is harder to change and because you'll likely have to bring in consultants and lawyers to do so, but because what really matters is how the people are operating, which tends to fall more into dynamics. It's the same situation whereas you see organizations that have values that are written on their wall, which are very different from the values that they may live in their interactions with one another and their customers. The dynamics from my perspective at least, is how are they living their structure? What is it actually like to interact with other people? What are the dynamics governing a relationship? Gene Kim (01:18:27): Thank you so much. I've learned so much through every interaction that we've had. So tell us, how do people reach you and what sort of problems do you love helping people with? Jessica Reif (01:18:37): You can reach us at crosslead.com. I'm jessica.reif, R-E-I-F @crosslead.com. And that's C-R-O-S-S-L-E-A-D. And Dave is [email protected]. The types of problems that we enjoy working with the most are really the ones that we talked about in this episode. How do you operate more effectively as a network of teams and how do you address some of the challenges that come with working on complex work in the context of a complex system? David Silverman (01:19:06): Yeah, I think Jess just nailed it. I'm passionate, we're passionate about multi-team systems, specifically how organizations communicate and make decisions in environments that necessitate flexibility and adaptability. That's always been sort of my passion. I have a bias towards high-performing organizations that are committed to continuous improvement, because I think without that, it's sort of tough. And so helping instilling that culture and then driving the mechanisms that reinforce those behaviors is what I think we spend most of our time with customers and clients spent talking about. Gene Kim (01:19:40): Wow, that was such a cool interview. And I learned so much from it. I think the reason for this is that it was such a boundary spanning adventure for me. There's a saying that goes, "There are no new solutions under the sun, only old solutions to new problems applied in a different domain." For those of you who love the history of science, Dr. Thomas Kuhn describes something very similar in his famous book, The Structure of Scientific Revolution, published in 1962. He introduced the famous terms, paradigm shift and inflection point, popularized by Dr. Andy Grove, the third CEO of Intel, credited for helping them become the largest semiconductor company in the world. And is considered the father of OKRs, objectives and key results. Gene Kim (01:20:27): What Dr. Thomas Kuhn studied was scientific revolutions, such as the Copernican Revolution to Newtonian to Einsteinian. And his observation was that scientific revolutions from afar looked like the work of one person. In other words, the revolution can be credited towards Copernicus, Newton, and Einstein. But during each period, there was a cohort of scientists also trying to explain similar anomalies. Dr. Kuhn describes that moment, that revolution, a sublimation event, or a phase shift where air suddenly becomes a solid, the moment at which scientific revolution occurs from say Copernican to Newtonian to Einsteinian. And I think there's something very similar happening around the post-industrial methods of management that has been happening over the last several decades. Gene Kim (01:21:20): After this interview, I find it suddenly super interesting and curious that the birth of Agile in 2001 takes place very near the timeline of stories that show up in the book, Team of Teams. Agile and special operations were all about unlocking the potential of small teams for doing complex work. And Team of Teams talks about how to truly unlock the team's potentials when there are hundreds or even thousands of these great teams who must work together across a vast enterprise, fighting against an agile, networked, and decentralized adversary. Gene Kim (01:21:53): And upon re-listening to these interviews, David Silverman and Jessica Reif struck me as incredible boundary spanners. In other words, they span the boundaries of many different disciplines. The book, Lean Startup by Eric Ries influenced so many of us in the technology community, especially in DevOps. I found it somewhat astonishing that it similarly influenced David Silverman through Brad Smith, the chairman of Intuit. So there is a link between Lean Startup, Brad Smith, Team of Teams, Agile, and DevOps. Gene Kim (01:22:24): So many of these discoveries are happening at the same time. There's a strange familiarity and commonality in those journeys. And I'm starting to appreciate just how mutually complimentary those bodies of knowledge are. For me, so many of the stories told in Team of Teams validates so many feelings and intuitions we've had in technology. And it concretized some of those concepts. And they even have a language for things that we have all experienced and vice-versa. Several times during these two interviews, I've had that feeling of finally learning about a word that describes this feeling or thought I've had for years, but never had a word for it until now. Gene Kim (01:23:01): I think we'll find many more concepts in terms from the military that will help us define the magic methods of the age of software and data. I genuinely believe that in 10 to 20 years, everyone will consider these topics obvious and self-evident. But in the meantime, it will be these communities who are creating this body of knowledge. Gene Kim (01:23:21): One last note on this, I've been talking about Kuhnsian moments, and one of them was the Newtonian Revolution. I think it was about two decades ago that I read the book, Isaac Newton by James Gleick. And there was this one passage that I loved, specifically around the creation of Newton's three laws of motion. So in Isaac Newton's time, there were no concrete terms for critical concepts, such as force, acceleration, mass, and inertia. And in Gleick's book, it showed how difficult it was for Newton to frame the three laws of motion without these concepts. Gene Kim (01:23:58): So I'm going to read to you an older draft of the three laws of motion, and that reads like this, "If a quantity once moved, it will never rest until hindered by some external cause. A quantity will always move on in the same straight line, not changing the determination nor solarity of its motion, unless some external cause diverts it. There is exactly so much required and no more force to reduce a body to rest as there was to put it in motion." And there were some three axioms. And you can sort of sense that Isaac Newton was grasping for these concepts that had no concrete definitions yet. And only after he concretized those terms, did the final three laws of motion spring into being. Gene Kim (01:24:45): One, "Every object in motion remains in that motion unless an external force is applied to it." Two, "The relationship between an object mass, m is its acceleration, a and applied force, F. Specifically, acceleration and force are vectors." By the way, I just learned this, apparently Aristotle said force equals MV where V is velocity. In other words, it did not have a direction associated with it. And third, "For every action there is an equal and opposite reaction." Gene Kim (01:25:17): So in summary, this is why I'm so excited by Dr. Stephen Spear's concept of structure and dynamics. I think it's an extremely parsimonious and thus potentially very powerful way to view how and why organizations act the way they do. Gene Kim (01:25:35): Thank you so much for listening to this episode of The Ideal Cast. In the next episode of The Ideal Cast, I will have on Jeffrey Fredrick, co-author or the fantastic book, Agile Conversations. I was so delighted that I've gotten a chance to know him better because I've known of Jeffrey Fredrick's work for over a decade. First, through his groundbreaking work that he did with cruise control back in 2005, which was my first experience with something resembling continuous builds and continuous integration. And he helped pioneer the use of continuous integration as a co-organizer of CITCON, Continuous Integration and Testing Conference. Join us as he helped me concretize some of these concepts that you've heard over the past couple of months. See you then.