Gene Kim (00:00:00): You're listening to The Idealcast with Gene Kim, brought to you by IT Revolution. Gene Kim (00:00:06): In the last episode, I interviewed Admiral John Richardson, who served as Chief of Naval Operations for the U.S. Navy for four years, which is that service's highest-ranking officer. He now serves on numerous corporate boards, helping leaders of leaders better serve their customers and win in the marketplace. If you haven't yet listened to that interview, I recommend that you listen to it first because this is a continuation of that first interview. Gene Kim (00:00:32): Today, we discuss his views on the importance of training leadership in the earliest stages of a sailor's career, and why leadership is so important, various tools and techniques for enabling radical delegation and why he views that capability being more needed than ever. Some very important characteristics of the different ways that integrated problem-solving occurs in organizations, one of which is fast and the other which is slow. The nature of the functional organization that is US naval reactors, comprehensively responsible for the safe and reliable operations of the US Navy Nuclear Propulsion Program and why it warrants being commanded by a four-star Admiral. And in complex organizations when accidents occur, what should leaders do in the ideal? Gene Kim (00:01:20): And like in the first interview, I am joined by my mentor, Dr. Steven Spear. Okay. Let's start the interview. Gene Kim (00:01:27): One of the things that really leapt out at me is your emphasis around leadership development. And you had once told Steve and me a story about the extent that you went to, to reinvigorate training, to toughen up the training. You had this incredible story about how that training came about. Admiral Richardson (00:01:42): Right, right. Gene Kim (00:01:43): Getting away from the post 1940 human conveyor belt mentality and the amazing outcome that resulted from it. Could you tell us that story? Admiral Richardson (00:01:50): Yeah. So, well it happens really at every level and it's a task that's still in progress to be honest, Gene. It was the product of a lot of really smart people who helped us just sort of pick away at this at every level. Admiral Richardson (00:02:04): So the major entry point for the United States Navy is up in the Great Lakes at our bootcamp, right up near Chicago. We process about 40,000 people per year, all over the country, and even the world, to come in and join the US Navy. They come from every imaginable situation. And so this really is the place where we start the integration. We take that vast, disparate, talented cadre of people who have raised their right hand and committed to joining the Navy and now we begin the process. Admiral Richardson (00:02:40): So this was motivated by me walking around and talking to junior sailors and them saying, "Hey, that was a lot easier. That bootcamp was a lot easier than I thought." And I was like, "What?" And it was fairly consistent. And you take that observation against the background that we were really trying make the Navy a little bit more capable, a little more tougher, right. We made toughness one of the core attributes of the Navy. And so I thought, "Well, these two things are just ... That's a contrast I'm not sure I'm comfortable with." Admiral Richardson (00:03:14): So we started to take a look at how we might increase the value, really, of bootcamp. And a lot of that would be to honestly, to make it harder, right, to make it tougher on the recruits. And so just from about every dimension, intellectually, physically, from a team building standpoint, we improved the curriculum and we did make it harder, right? The physical standards became more challenging. We really went back to fundamentals of watchstanding so that we could, no kidding, graduate a sailor who upon arrival in their crew was ready to contribute with some fundamentals, for sure. And there's a lot of education still to go, but we wanted to do that. Admiral Richardson (00:04:03): So we actually started backing it up even before they arrived at bootcamp. So typically, somebody with some period of time before they actually report to bootcamp, just logistics, et cetera. And we backed the training up into that time before they reported. Particularly in the physical, "Hey, let's get together and workout," and all of that stuff. So we cranked it up. The other thing that we did though, and this was where the thoughtfulness of the team really was just astounding, was that we also developed some mechanisms, maybe call them coping mechanisms, for each of the recruits to manage that increased level of difficulty. Kind of sounds like a bad joke sometimes, the beginning of a bad joke. But it was a SEAL cognitive psychologist and a chaplain who got together and put together these coping mechanisms. It was astounding, the things that they really taught these young recruits in terms of the ability to manage their way through and help their shipmates through difficult situations. Admiral Richardson (00:05:13): And so just as a visual example, one of the very first things that recruits have to do is get comfortable with the water, obviously. They have this platform dive that they have to jump off of, a pretty elevated platform, into a pool. And that could be a scary thing, right? Some of the recruits that just don't have any experience in that regard and if you've ever done that, it can be a little bit of a fearful thing. So you could literally see these young recruits go up there and they start to, this is like I said very early on, they start to hesitate and either their instructor or their fellow recruit would just use a code word, I think it was calibrate or something like that. And they would start these techniques to just center their mind, do some breathing or whatever and then they'd step right off. It was just remarkable and everything would be fine and everybody would cheer. It was very self-reinforcing. And so you just start with that. Admiral Richardson (00:06:12): And so I think, the point being that, yes, we had to make bootcamp more challenging so that the graduates, the sailors that graduated would be more relevant and contribute more to the Navy upon graduation. Making itself tougher but also it was insufficient just to do that. We had to give them some tools for them to get through this. Admiral Richardson (00:06:33): So some of the great stories of that were that the instructors and the sailors also started to use these tools in other parts of their lives. We would have people say, "Yeah, I actually at home in my marriage, I'm able to use these tools to get through difficult times with my spouse," or whatever it might be. And so we really thought that we were onto something there, I think, write an article or something about this. These are very sort of boutique tools. And so if you go to high-performing teams and you can find these sorts of mechanisms at play, but typically those teams are pretty small. And we're doing it at scale, 40,000 people a year able to use these tools and get through this indoctrination period. So, we really felt that we were onto something. Admiral Richardson (00:07:19): The key, actually, Gene was getting the instructors on board. They've got to be real believers of these techniques, which in many cases were new to them. I mean the previous approach was just sort of yell louder and so we really took a different approach to that. It was interesting that when we first started it, we went into this doing a number of pilot programs and those sorts of things, refining it, and then finally, doing all of bootcamp. We're actually in middle of spreading it around the Navy where it can be applied. You've got to go slowly and thoughtfully because of the training involved with the instructors. Admiral Richardson (00:08:03): Early on, we were taking a look at different pilot programs and experimental phase, and we found that one of the teams up at bootcamp just weren't getting it. They were performing in the same, et cetera. And so we dug into that and we found it was because the instructors really had not fully adopted the approach. You really got to be a believer when you do this. Admiral Richardson (00:08:27): Now, of course, when we cranked up the level of difficulty, a lot of people kept telling us, "Hey, this is going to be a disaster. This generation just isn't going to be able to handle it. You're going to see everybody leave bootcamp and drop out." And in fact, we saw exactly the opposite. Our graduation rate went up, our attrition went down and even the day to day, "Hey, I don't feel so well today, so I'm going to call in and call in sick." All of that went down. We really felt like we had captured something special by virtue of these techniques, because it just became so much more effective across the board. People seem to be really aligning with it. Admiral Richardson (00:09:12): I mean, the Navy is about as talented as it ever has been, by every measure right now, grades and fitness, et cetera. And these young people can graduate from school, high school, college, whatever it might be and write their check anywhere in the world. I mean, they really are that good. A lot of competition for that talent, and yet they chose the Navy. So we didn't want to shy away from that. We wanted to give them the challenge that they signed up for, and they really leaned into that and made everybody proud. Gene Kim (00:09:43): And you had this great story about the new sailors, their role in a firefighting scenario, before and after. Admiral Richardson (00:09:49): Yeah. I don't know about before, but we wanted to get some feedback from the fleet, the graduates, how were they doing? Was this really achieving the aims that we had set about? It's a little bit anecdotal. In general, we got very positive feedback, but there was this one story that really made me smile, which was the story of a Chief Petty Officer at sea talking to somebody back in the recruiting command saying, "Hey, what are you guys doing back there?" And it's, "Well, why do you ask?" "Well, we just had a brand new recruit and she shows up on the ship and typically we put them in a learning phase and we don't expect much of them. But just the other day we did a fire drill and I was watching this damage control locker, to just monitor performance, and this young sailor came up and she opened the damage control locker. She just went to town. She broke out the hose, flaked it out, charged it, grabbed the knife and started doing exactly what a seasoned firefighter might do." Admiral Richardson (00:10:54): And so basically the feedback was, "I've never seen that level of proficiency on arrival before, so whatever you're doing out there, keep doing it." So it was really a nice bit of feedback there. Gene Kim (00:11:07): I can imagine, maybe before you had those success stories, some of the reactions you would get from the instructors would be, "Holy Cow, the CNOs, trying to be awfully nice." Nice meaning maybe opposite of effective or tough. Could you tell us what maybe that discussion might have sounded like to refute that, not someone being nice. Admiral Richardson (00:11:27): Right. Well, it was really a complex discussion Gene, because first there was the discussion about toughness. Is toughness the right attribute. And really, I think, Steve, you might've been there for some of those sessions, just joining us there for the leaders meetings, et cetera. But they would say, "Well, we're really after resilient or toughness. We don't want to make it tough on ourselves. We want to train recruits to be more capable, but we certainly don't want any abuse of any kind going on, hazing or anything like that." We're very mindful to protect against that. And so people were, I think, thoughtful about, is that the right word? And at the end, we sort of stuck with it and with all of the attention that it deserved. So that was the first part of the discussion is, "Hey, is toughness really what we're after." Admiral Richardson (00:12:26): And so then it was, "Okay, so how do we go about building it, raising the standards." And everybody understood that. So instead of 10 pull-ups, something as simple as that, how do you get there? Mental techniques were a little bit of a new approach. So now that's where it's like, "Oh, wait a second. Why can't I just yell more?" We really wanted to build teammates who were ready to take on challenges, that can be quite scary, very challenging. And so how do we build them up? It was really that combination. I mean, US Navy SEALs, pretty tough crowd, no doubt about that. The psychologist just did a tremendous amount to add to that. And then, of course, the chaplain is monitoring and bringing all of that to bear. It was the balance of those body, mind, and spirit that created a set of techniques. Again, the proof was in the pudding. There was a lot of skepticism. Eventually we got the instructors to say, "Okay, well, we'll give this a try." The feedback ended up being positive and so it just kind of built on itself. Steven Spear (00:13:47): Quick reflection. One was a meeting of senior leadership. It's the flag officers and their civilian counterparts and CNO had thrown out the idea of resilience as a characteristic to develop. From an engineering perspective, it's the right term, but people didn't like it, it wasn't macho enough. So I think that's where the toughness, if I remember that's where that one came in. Gene Kim (00:14:06): Gene, here. I had mentioned in part one of this interview, how much I admired the two documents that Admiral Richardson mentioned, had shared with me. One had the title of Designing for Maritime Superiority. And the other was the Navy Leadership Development Framework. I loved both documents because they were both so clearly written. And it was such a treat listening to Admiral Richardson and Dr. Spear talk about some of the early meetings as they were establishing some of these important concepts that went into those documents. Gene Kim (00:14:37): In version 1.0 of Designing for Maritime superiority, that attribute of toughness is one of four attributes described. As I described last time, the document starts off with mission, the strategic environment or problem statements. The next session is core attributes. Then he describes the four lines of effort, which I read last time. And the document ends with desired outcomes and conclusion. Gene Kim (00:15:02): So toughness is one of those four core attributes. The first is integrity, our behaviors as individuals, and as an organization align with our values as a profession. We actively strengthen each other's resolve to act consistently with our values as individuals, as teams, and as a Navy, our conduct must always be upright and honorable both in public and when nobody's looking. Gene Kim (00:15:24): The second is accountability. We are a mission-oriented force. We achieve and maintain high standards. Our actions support our strategy. We clearly define the problem we're trying to solve and the proposed outcomes. In execution, we honestly assess our progress and adjust as required. We are our own toughest critic. On their own everybody strives to be the best they can be. We give 100% when on the job. Our leaders take ownership and act to the limit of their authorities. I love that. We foster a questioning attitude and look at new ideas with an open mind. Our most junior teammate may have the best idea. We must be open to capturing that idea. Gene Kim (00:16:03): And fourth is toughness. We can take a hit and keep going. Tapping all sources of strength and resilience, rigorous training for operations and combat the fighting spirit of our people and the steadfast support of our families. We don't give up the ship. Gene Kim (00:16:16): I want to share three observations. One, after thinking about this document for weeks, I find it super interesting to look at its structure. The way most strategic documents seem to be written is you have high-level vision and mission, and then the objectives and key results. But contrast that to this document. Gene Kim (00:16:35): First, its mission and the strategic environment or the problem, and then they zoom down to core attributes and then back up to the four lines of efforts or objectives. Isn't it interesting that the core attributes is between the problem and the solution? Gene Kim (00:16:52): I think it's because this further underscores the need for radical delegation. In fact, they clearly state they want quote "leaders to take ownership and act to the limits of their authorities" end quote. Until this document, I don't think I've ever seen that done like this before. Gene Kim (00:17:09): Two, I was curious about how much effort obviously went into this document, so I asked Admiral Richardson over an email, whether he used writing as a means to clarify his own thinking. And he did indeed confirm that is the case. He wrote about how important it is to write words down, as opposed to just writing PowerPoint slides. And his next step in the writing process is to keep reducing the size of the document to reduce the likelihood of misinterpretation. Gene Kim (00:17:36): Which leads to my third observation. All this reminds me of one of the famous practices of Jeff Bezos, founder and CEO of Amazon, which has been written about considerably. He would start the first 10 to 15 minutes of his senior leadership meetings by having everyone read a six page document that was prepared by the person presenting a problem, an idea, or a proposal. Bezos observed that in a traditional meeting where someone presents a slide deck executives would interrupt with questions. However, with a written document on page two you pose a question and on page six you present your answer. In a 2012 fortune magazine interview. He said that writing memos is an important skill to master quote, full sentences are hard to write. They have verbs, the paragraphs have topic sentences. There is no way to write a six page narratively structured memo and not have clear thinking. Many have remarked that this is a clarity of thinking you need in order to reason about complex business decisions. This was all very fun to read about because it demonstrates why writing is so valuable of a skill for senior leaders. I'll put links to some of my favorite articles. I found on this practice at Amazon in the show notes. Many have remarked that this is a clarity of thinking you need in order to reason about complex business decisions. This was all very fun to read about because it demonstrates why writing is so valuable of a skill for senior leaders. I'll put links to some of my favorite articles. I found on this practice at Amazon in the show notes. Okay, back to the interview. Steven Spear (00:17:48): Gene, this is an aside and less Navy related, but this idea of thinking through all work as a process, which has an intended outcome, an intended deliverable, all processes exist to generate an output and that output has to be delivered to somebody. And that's true, not just for manufactured product, but anything. You have someone coming to a class, the whole reason for that should be that you're preparing them for their next step in life. And we've seen it, actually very rare, that people responsible for educational developmental processes think about the next place the person will end up and designing for that next place. Steven Spear (00:18:23): So not surprisingly the exception to the rule is Toyota, which I spent a lot of time documenting in my book. I think Chapter Six, about working backwards from when the moment that person steps on the line, are they ready? And so that story about the young recruit coming out, and she's just ready for the fire party, that would have been a very Toyota story and pretty much nowhere else. Steven Spear (00:18:44): In contrast, we did a study of how medical students go through their process and it's without any context at all, as to what's coming up next. And so the predictable consequence of the first approach is it probably takes less time to get much more skill into people and they're much better ready for whatever's next. And the predictable consequence of the typical approach is it takes a lot of time, a lot of money and people arrive at wherever they're next and they're frustrated they're not ready to get going. And the people who are now responsible for them, are now frustrated with the amount of babysitting required. So anyway, the story about reformatting the basic training, resonated very strongly. Admiral Richardson (00:19:24): I guess, the framework in which we were thinking about that was just view the Navy as, in many ways, a leadership factory, And so we really are, even as we bring people in and we begin this sailorization process is what we call it there at Great Lakes. And at all of the entry points, we really are thinking about creating these leaders. As soon as they arrive at their first command, our sailors are ... they no sooner learn their way to their own bunk on the ship, in a very short amount of time, they're being put in charge of some small team. They're going to be leading before they know it. I think all the services just provide a tremendous amount of responsibility pretty early on in a career. And boy, we have to start preparing for that pretty early, beginning of bootcamp, to be honest, Gene Kim (00:20:21): How important is that, that these young sailors are being put in leadership roles immediately? Why is that important? Admiral Richardson (00:20:28): Well, I think it's important because you're going to want to build a team that is as adaptable and responsive to fleeting opportunities and stimulus as quickly as possible. And even on something that is as contained as a ship, you're going to have different things happening in different parts of that ship. You're going to want somebody to be able to take charge of that situation and lead their way through it. Whether it's an opportunity to excel or whether it's some damage control scenario where you've got to fight a failure or a fire or some emergency. Admiral Richardson (00:21:05): Yeah, I mean, I think that the Navy does just about everything that we do in teams and we need people to lead those teams. It scales down pretty quickly. We have teams of all sorts. Administrative teams, divisions, departments, those sorts of things. We have watch teams, which are operational teams, and there's just a tremendous amount of leadership opportunity. Steven Spear (00:21:28): We met when you invited me to a perspective Commodores course down at the submarine command. Admiral Richardson (00:21:35): Norfolk. Yeah. Steven Spear (00:21:35): Yep. The reason I hook on that is the impression that, after basic training, every time someone takes on a new role, whether it's a rank or whatever else that there's a formal training built in on how to be the person in that role. One, I just want to confirm that's true. The other part, and again, this idea of the perspective Commodore is that someone should show up for their next command, ready to command, not learning on the job. PART 1 OF 4 ENDS [00:22:04] Steven Spear (00:22:03): Show up for their next command, ready to command, not learning on the job. So if that's true, why was that not true for the entry-level sailor, that they arrive ready to stand their first watch? Admiral Richardson (00:22:13): Well, it's easy I think, to your point, your earlier discussion that, hey, what is the actual purpose? What is the output that we're looking for of this organization? And one thing about schools, training departments, or whatever is that I think organizational leaders need to visit those places often to make sure that they stay aligned with the aims of the organization and that the graduates from that school are actually serving the purpose that the school was stood up to meet. Because it's easy to drift off. I mean, they're a little bit of an organization unto themselves. And so unless you just sort of visit them, make sure that they stay in the mainstream or the main current, and they don't form kind of an Eddy pool off on their own. Admiral Richardson (00:23:09): We just saw a lot of examples of that. Where the school was, I mean, it was functioning as a school. They were doing a lot of teaching and everything, but the graduate wasn't exactly aligned with what we needed at school to produce, which is team members, watchstanders. You got to go and graduate and be ready to be part of a crew. And so if you don't kind of keep that front and center, it's easy to do a lot of really good education, et cetera, but you just don't graduate a person that's meeting the expectations of the fleet. And so there's that dialogue between the school and what it produces and the fleet who's going to assimilate that graduate and really wants to have that person be as productive as possible on arrival so that you don't extend that education responsibility into the operational force. Admiral Richardson (00:24:07): They're going to do plenty of that anyway, but bootcamp's set up to do what it's supposed to do, and I don't want that kind of bleeding over into the fleet. So managing those expectations between graduate and recipient, if you will, I think is really important. Admiral Richardson (00:24:26): The other thing about this and, and it's pretty accurate what you say, Steve, is that there's some education for just about every step forward in a sailor's career. But I'll just speak, many of those steps are just, I would call them linear scaling of the last job you had. So you become a division officer, you're in charge of, let's say, 15 sailors, that's your division. And the next thing you do is you get promoted, you come back as a department head, and you're in charge of a bigger group of sailors. So that's just a linear expansion of your previous job. Admiral Richardson (00:25:13): But there are some jobs where it is non-linear. The difference between your last job and the one you're about to take is really non-linear. And there are very few of those as I thought about it in the Navy. One is, when you first enter the Navy, you change from a student to a leader. And that is different in nature. If you think about your entire life as a student, it's all your grades, your performance, et cetera. You're being graded. You need to graduate. Your success is really based on your individual performance. Now you're a leader and to a great degree, even as a junior leader, your success is based on the success of your team. So you've got to invest a lot of your energy putting your team first and making sure that they're performing. And that's new. That's different in nature. Admiral Richardson (00:26:13): And that the next non-linear step, I think, in the Navy anyway, is you move forward and you take different levels of responsibility in the organization until you get to command. And command is different in nature than anything that has come before it. And so, the idea that, hey... Well, first, the regulations have different words to talk about command in terms of the extent of the authority and accountability and responsibility that you have. I remember my first time getting underway on a submarine in command, and there really is, there's nobody to look over your shoulder and ask, "What do I do next?" It's really kind of on you to get that right. And the other thing is when things go wrong, you're standing there and every eye turns to you like, "Hey, what are we going to do now?" That's different in nature than anything that you've done before. Admiral Richardson (00:27:10): And then, Steve, what you witnessed was a perspective commodore course, which is also a non-linear jump because now a commodore is going to be the commander of commanders. So that's different than just commanding one team. You have to figure out now how do I bring these commanders along, giving them the opportunity to enjoy command just as you did as a commander, try not to be the super commanding officer of the entire squadron and training these commanding officers to be better. That's different in nature. And then you get into really the flag ranks before it changes in nature again. So yeah, that's a very important school, that prospective commander school, because you have to teach people to think a little bit different than they have because this commodore job, commanding other commanders, is different than anything they've done before. Gene Kim (00:28:11): So neat. I want to tell the story that Steve told me that has consumed, for me, my mental space for about six months, seven months, and I'd love to see what reaction you have out of it. So the story was about a trip that Steve made with his mentor, Dr. Kent Bowen, to visit a Toyota plant in the 1990s. And there was a VP of manufacturing from a big three automotive manufacturer there. And among the many things that they were shown was the fact that in this Toyota plant they're doing 16 line side store changes per day. The line side store being where the input for a work center is stored. And so the VP of manufacturing reacted with incredulity, or maybe even disbelief, saying that's crap, because I think it was because it was so preposterous. And he said, we did six line side store changes in one day and it shut the entire plant down for three days. And so certain parts wouldn't get to where they need to be, final assembly couldn't happen, so no cars come out for three days. Gene Kim (00:29:09): And so the story that evoked for me, the memory was, in 2009, John Allspaw, the VP of operations at Flickr and his counterpart, Paul Hammond in engineering and development, said we're doing 10 deploys a day every day. This is in 2009. And I think the whole technology community just threw up in the aisles. They're like we do 10 deploys in a year and we can barely handle that. What sort of maniac would you 10 deploys in a given day when they don't need to? So I think the big aha moment for me was that in both scenarios, they were able to reduce the cost of change so that each small change could be done easily, seamlessly, safely, securely without causing widespread chaos and disruption. Gene Kim (00:29:52): And so much of that, in the manufacturing world, is enabled by these Kanban cards where an envelope with a to and a from and what parts you need, you can change the routings, but no one really needs to know because they're for the people in the loading docks and where the parts need to go. And so that sort of decoupled and enabled the decentralization of the changes so they could be made more safely. Gene Kim (00:30:12): I think what's in common between the Toyota story and the technology story is that you're trying to enable these teams to work independently and quickly without causing chaos and disruption. So in that manufacturing example, each work center was able to do improvements, mostly independently, without having to let a central production control group know. And what happened in the American automotive plant was that that central MRP system or something, if you missed a detail, suddenly everything came collapsing down. Gene Kim (00:30:40): Lieutenant Commander Dave Silverman in the Team of Teams story, he was talking about how when they tried to push the operational tempo without decentralizing the planning and execution process, when they tried to go twice as fast, it ended up with higher numbers of accidents, injuries, and civilian casualties. So it seems like there's a structure where you couple of things together and the cost of change is unacceptably high. That small changes can have bad outcomes in a distant part of the system, or maybe even globally. And there are some structures that allow rapid innovation and changes at a more local level. Could you react to that and maybe talk about some structures that come to mind that are very brittle in that way versus ones that allow more independence? Admiral Richardson (00:31:21): Yeah, it's very interrelated with technology. So one dynamic that I think is fundamental to successful Naval operations, a successful Navy decentralizes as much as possible. Before it was necessary, this decentralization was unavoidable because you would get a ship and it's grew and it would leave port and it would go over the horizon and there was no way to reach it. And so whatever you told that captain and their crew before they left, they were on their own after that. They just had to respond. And the better crews of course were more creative, more independent, more responsive to these fleeting opportunities and came back having contributed a great deal to the central mission if you will. Admiral Richardson (00:32:07): So there's this kind of radical delegation that we talked about in our first interview. And there's a real art to crafting commander's guidance to give that commanding officer to say, "Hey, this is what we really need you to do. I'm not going to be able to talk to you for some time. So go off and do good. Here's what I define as good. And I'd really be grateful if you didn't do this stuff because this I define as bad." Admiral Richardson (00:32:35): And so at that builds into the DNA of the organization, this decentralized approach. Even in the cause of the greater good, if you will. Now of course, it's a different scenario. Through technology, you do have the ability to communicate a lot more. And so I think it's incumbent on the senior commanders now to just exercise some appetite suppression, to reach down in and want to just get constant updates or whatever it might be, which can hamstring that individual command from being as responsive and agile and innovative as we need them to be. Admiral Richardson (00:33:32): Particularly, as you said, as the tempo and intensity go up, that communications channel is going to degrade. I think it's not a mystery that command and control network will be one of the first targets so that you can just degrade that ability to communicate. And so we really need to continue to train our commanders to be very innovative, very creative, in the accomplishment of the mission. But there's that tension now with the information technology ability to kind of stay in constant video contact almost, no matter where they are in the world. Admiral Richardson (00:34:16): And so it's an interesting challenge for the Navy as we think about training commanders, but even as we think about training major commanders, the senior commanders, to allow everybody to reach their maximum creative potential, to stay as decentralized as possible. But yeah, so it's interesting. And then if you get down to the ship level, I kind of really... Gene Kim (00:34:43): Oh, no. This is great. Admiral Richardson (00:34:44): I'm talkative today. I don't know why. There's some things that can be decentralized, but one of the things that's really great also about service in the Navy is that you do have this idea of a crew. And boy, that ship, what happens to one happens to all. And so they do have to come together and integrate at that level just because it's essential to mission accomplishment and also survival. And so it's like an organism, Gene. You got to have a head, but you also have to have arms and legs and a body and hands and everything else. Gene Kim (00:35:29): I hope you're enjoying this conversation as much as I am. I absolutely love Admiral John Richardson's view of leadership. And it's been amazing to see how frequently his work has been studied and referenced within the technology community inside the US Department of Defense. We are well underway in creating our 2021 Virtual DevOps Enterprise Summit Europe event, with the goal of making our best programming ever. One of the announced speakers is Dr. Ron Westrum, whose work is so familiar to so many of us within the DevOps Enterprise community. This virtual conference will be held on May 18th through the 20th. Go to events.itrevolution.com/virtual and use the code idealcast to get $150 off your registration. Gene Kim (00:36:15): And you may have noticed that the DevOps Enterprise video library has been loaded with new content. Behind the scenes, we've been working to add all the talks from all our previous conferences. This is nearly a thousand talks dating all the way back to 2014. We're publishing them in batches, so check back in weekly to see what's new. To test it out, try searching for Dr. Andre Martin. His incredibly popular session from Las Vegas, 2019, hasn't been released to the public until now. Go to videolibrary.doesvirtual.com to access it. Gene Kim (00:36:46): I wonder if you can sort of paint a spectrum for me. So maybe if you were to take a look at whether it's a task force or something, some assembly of forces, can you paint kind of like a configuration where you would look at and say, "Oh, that's just doomed from the beginning." Just by looking at the communication paths or the way decisions are made versus one where you'd look at and your reaction would be like, "Ah, they are set up for success." Admiral Richardson (00:37:12): Yeah. Yeah. I would take a look at how they disseminate their orders, if you will, and then how those orders are then accomplished. So I would say that on the over controlled, micromanaging end of the spectrum, you would have a situation where not a lot of discussion upfront about why we are doing things the way we're doing, a deep discussion about what we're trying to accomplish. Instead, it would just be sort of a rapid fire series of do this. And then accompanying that would be little to no feedback from the subordinate commanders about, okay, this is how it went or something. It would just be kind of a one-way staccato of direction. Admiral Richardson (00:38:04): And at the other end of the spectrum, there would be a tremendous amount of what we call command and feedback. So spend a lot of time talking amongst commanders about, hey, this is the aim of the mission. This is what we're about. These are the assumptions that we're going into this with, but it's dynamic. And we're not an academic institution so we spent a fair amount of time assessing this, but we are where we are and it's time to move out. And so it might not be exactly as we describe it. If it's not, I need to hear from you. And so feed that back and here's your commander's guidance, and let's go out and execute. And beyond that, I'm just not going to give you much more direction. You've got everything you need to go off and command. Admiral Richardson (00:38:50): And somewhere in the middle, it would be kind of things like control by indication. Hey, do something. But just let me know what you're doing. And give me three Mississippi to say, "Hey, don't do that," if I have an objection. Admiral Richardson (00:39:04): And I think also it depends on the level of trust and confidence that you have. So there's not one answer that fits all. Personally, I just stay away from the staccato one way, burst of orders. I just think that is an unresponsive system that's perhaps doomed to failure in a complex environment. On the other end of the spectrum, boy, it takes a lot of trust and confidence. You have to know each other very well. And so teammates that have served together a long time, you mentioned Team of Teams and the Special Forces, they put a lot of value on that. Because they're going to be doing some pretty high end operations in a decentralized way. So they like to work with people that they have a lot of experience with and can trust. Admiral Richardson (00:39:48): In the middle, you're getting new commanders all the time. And so you want to start to train them to be that highly trusted, highly innovative agile commander, think on their own, be successful but you're also mindful that they're brand new at this. And so this kind of command and feedback, or control by negation is a nice way to make sure that it's their idea. And as long as things are going well, they're going to go well. You just let them go. But it also gives you an opportunity to do some training or adjustment if you need there. Gene Kim (00:40:21): Just to maybe go into sort of tools and techniques, it seems like one of the big challenges to overcome is, I think Steve in the last interview mentioned the notion of guard rails. How do you avoid the conditions where an immediate situation needs to be deconflicted, I'm just trying to think about what are the tools that enable this sort of this opposite of one person doing all the thinking and dispatching all the orders in one way. Things that come to mind are from reading a lot of Tom Clancy books, I guess. I guess the US Navy has come up with some very novel ways to sort of avoid that need for communication. Like identifying some areas of sea that only one submarine is allowed to be in, or what are those sort of other tools that allow for this absence of one node continually telling every other node what to do? Admiral Richardson (00:41:11): Right. So that's one way is you just sort of parse things out. And the Navy is moving to a concept of operation that is more and more like that for every type of fighting element, if you will. And so it's very distributed. Because the technology is just to the point where if you're massing forces together, that's a much more detectable type of an event. And the ability to reach out at long range and precision and target that is just increasing. So you want to make that targeting problem much, much harder for your enemy so you distribute. Admiral Richardson (00:41:55): But you still often have to coordinate effects, whether that's getting that distributed force now to operate in synchronicity. So these ideas where you're going to distribute geographically, distribute amongst a domain: So air, sea, under sea, space, land. They all have to coordinate and synchronize together. So some of that's going to be geographic. Some of that's going to be functional, but it's getting to the point, Gene, where this ability to really outstrip our adversaries in capability comes down to almost nanosecond timing. And so you've got to have kind of a clock that is moving everybody forward. But it goes to this guardrail because you just can't anticipate... They've got that phrase, "No plan survives contact with the enemy." Admiral Richardson (00:42:55): And so even within that, when things go wrong or they're unexpected, or this particular signal was just missing, it's been disrupted, by virtue of that really melding of the minds before execution, I think that's what helps define the guardrails. So, hey, this is your mission. Here's the entire landscape of risk that is yours now. And so it's a big landscape. It's a big part of the mission, but it's not everything. And so there will be areas where, hey, if you go outside that, we're actually going to start to conflict internally. The worst thing being some kind of a blue on blue type of engagement, where we're shooting at ourselves or something like that. Admiral Richardson (00:43:45): And so there will be some guardrails or boundaries that, hey, inside that, exercise all the creativity you can, and here's the tempo and the synchronous plan. We're going to move together. But if you lose that signal, hey... PART 2 OF 4 ENDS [00:44:04] Admiral Richardson (00:44:02): ... together, but if you lose that signal, right? Hey, try and contact me, get back, and if you're completely isolated for a while, while you're isolated, here's the commander's guidance that you can continue to operate within those boundaries and still move the mission forward. Gene Kim (00:44:15): When you were describing what kind of good and effective and maybe not so effective look like and evoked in my mind this phrase I heard a mentor of mine say, she said, "You can measure the intelligence of an organization based on the frequency and intensity of communications between kind of the nodes and the structure." Right? In one case, you have sort of one central nodes are broadcasting out one way orders, right? Nothing coming back, versus a lot of intense communication at the edges, with the edges. So the question is going to be about integration. It seems like a theme in the military for centuries have been pushing that point of integration further and further down the command structure, as opposed to just at the highest levels of command. And it seems like you were just suggesting there that in a certain theater of operations, there's going to be a tremendous amount of integration happening at the edges, whether it's air, sea, under sea, so forth. Gene Kim (00:45:04): Can you talk a little bit about that, why it's so necessary for integration to be happening further and further away from just the very tops of the organizations? Admiral Richardson (00:45:11): Yeah. I'll be happy to. Just one quick thing, even as I think about it more, there is a role, I think, for the sort of one way giving of an order, which would be maybe in extremist, right? Where you have an emergency and it's sort of, okay, I see this clearly. We need to turn left now or we're going to have a collision. So you really do have those cases where even that type of a leadership is appropriate. I think that you'd want to minimize those. So why integration further and further out from the center, if you will, that's really kind of where the action is happening right now. And there's this whole concept of jointness, right? Which I'm sure you've heard of, which at its most simple is the army and the Navy and the Air Force and space and everything is all going to operate in harmony. Admiral Richardson (00:46:15): But really what's happening though is it's becoming, I think, just too complicated to over centralize this, right? You've got to have that coordination at the edge because the complexity and tempo are such that to try and say, okay, I see this back to headquarters, what do you want me to do? Head headquarters only has an incomplete at best understanding of what you're actually seeing, the full context of it, but they're going to say, well, do this. And that may or may not make sense given what you're seeing now because the time has passed. So the tempo of it all and the dispersion of it all really, both geographically, functionally, technologically, it's just happening at a pace that if you kind of go back and centralize everything, you're going to fall more and more behind. This is not a choice, right? If you're going to succeed. Admiral Richardson (00:47:19): The other thing, and Steve and I, we talk a lot about football, what we find is it kind of goes all the way back to bootcamp because the ability to operate effectively at that pace out at the edges, it really mandates a command of the fundamentals in a way that's much more complete than before. Whereas when you're centralized, you could maybe absorb some inefficiencies or some parts of the place which were ineffective. But in this distributed scheme, everybody's got to have an unbelievable command of the fundamentals. To use the football analogy, you're not going to move into your no huddle offense and be able to do that effectively if you don't have the fundamentals of controlling the line of scrimmage just down pat, right? So you really have to spend ironically more time on the fundamentals because they're going to be the building blocks for creativity when you decentralize and move everything out to the edge. Gene Kim (00:48:32): One quick question before we go to Steve. One of the terms I picked up from your last interview was this notion of radical delegation. I mean, is there an overlap between the kind of this desire for radical delegation and what you're just talking about, the sort of need to keep pushing integration further from the center? Admiral Richardson (00:48:48): I think that they're one and the same, right? I mean, what you're doing is you're moving more and more responsibility, if you will, out to the edge, right? To achieve mission success. Where the trust and confidence in that commander out at the edge come in is that as the central commander, you can't ever escape responsibility and accountability for the whole mission, right? So you've really got to have a lot of confidence in that commander to achieve it because it all starts to kind of blend together pretty completely. You can't delegate that responsibility. So just to maybe go back and correct what I said, you're not going to radically delegate responsibility because that's just kind of something that can't be delegated, but you are going to delegate a lot of the mission success, right? A lot of the authority. Gene Kim (00:49:35): Gene here. Steve is going to start talking about the dynamics that result when you centralize and decentralize authority. And he's going to start referring to pictures that you won't be able to see because he drew them while Admiral Richardson and I were talking and use them to explain something that I think is super interesting and important. I will explain it all in about five minutes, so hang in there. Those diagrams were a huge aha moment for me because they point out some startling, common characteristics of the great dynamics that we heard from Dave Silverman, from Team of Teams, how engineers work together in DevOps and within the Toyota production system. And it also points out what goes so badly wrong when we require integrated problem solving from a vast number of specialties when they're not allowed to talk to each other directly. Okay, here's Steve, and I'll be back in five minutes. Steven Spear (00:50:26): Conversation we've been having about decentralized, centralized, et cetera, to a theme we've been developing as a structural issue and a dynamic issue. And I hand drew some pictures. So I'm disconnected to something like this, where each of us is doing our own thing. Perhaps we're doing our own thing with boundaries. Gene, this is your operating range, and just so long as you don't cross the street, anything you do on the other side of the street is fine so long as it's consistent with commander in tech, right? That's one. When you start having these higher degree interactions between the various pieces, then there's a tendency to want to go with a centralized approach, which Gene's in charge and everyone else has to await his instruction and then give constant updates. The problem with this approach is two; one structural, one dynamic. Steven Spear (00:51:23): The structural problem is it starts flowing too much information to you. So because of your inability to do all those computations, one of two things happens. Either you start asking for more and more rarefied information, which starts losing the idiosyncrasies of the locality, or you just get wicked slow in responding back, all right? So the structure in that regard actually drives the dynamics of you getting very, very slow and unresponsive. I think this is why that plan spazzed on six changes a day, because the operating tempo down here is actually very fast. The granularity of information kind of like work size issue because of local idiosyncrasies is very, very high. But whether it's these bands of transmission or this processor, it can operate keeping pace here. So this is the failure mode of David Silverman's situation and Team of Teams at the start. Steven Spear (00:52:29): The situation the book has described is either something like this, where the civilian analyst and the ranger raiders and the seal teams are doing sort of their own thing or it's this, which is everything has to be discussed with everyone all the time and the computational pace and the computational detail is just inappropriate. So then you get this approach, which is a much simpler structure. It's just connection between the pieces that have to directly interact and know more. And the value of this is that these direct connections actually carry much less information than this model. There's far less processing required at each of these nodes. So you can have much richer conversation in a sense, communication about far fewer things and it's real time. And so this becomes a Toyota model. Steven Spear (00:53:24): And also I think in Team of Teams, this is what they become and this is why they go from that mission rate of one a day to 80 a week, or whatever those crazy multipliers were. And it's really fascinating from my perspective because there was no change in equipment and there was no change in staffing, plus or minus, right? From David's account. It was only a change in the speed, the accuracy of the communication and computation. So this ties back to the Toyota example of 60 changes a day. It didn't have to be a centralized set of decisions. Gene Kim (00:54:00): Gene here. Okay. I'd like to process what Steve was talking about because he just described what I think is one of the most important aspects of how structure affects the dynamics of how organizations actually work. So I'm starting to believe that you can predict whether an organization is a high performer or a low performer, just by looking at the communication paths of an organization, as well as their frequency and intensity. So in the not ideal case, you probably have communications that are dominated by vast escalations up and down the hierarchy, because to do any sort of integrated problem solving, it requires the efforts from different functional specialties, and therefore you have to escalate issues to a level high enough where the two silos meet. So in the unicorn project, this is called a square where in order to get anything meaningful done, you had to go up to, over two, and then down to in order for two engineers to actually work together. Gene Kim (00:54:56): So this was discussed in detail in my interview with Mike Nygard, SVP of platform engineering and enterprise architecture at Sabre. So on the other hand, in the ideal case, the majority of communications are happening within the team and between teams without any need to escalate at all because there are already sanctioned interfaces between those teams. And when escalations happen, they go up normally one level, not eight levels, showing that most problems can be solved locally. For example, when someone pulls the [inaudible 00:55:28] cord in a Toyota plant, most issues should be able to be solved locally by the team lead and if it can't be solved within say 55 seconds, then it escalates, choose a group area lead. And one of the things that make this possible is that the work is organized by value stream so that every team knows where their inputs are coming from and where the outputs go. Gene Kim (00:55:50): In other words, their internal customers are the next down stream step. This results in a very simple, what appears like a linear flow of work. And so simple really means more linear than not. Simple probably means more explicit and less tacit or nebulous. The result of this is that integrated problem solving can happen across a broad cross section of functional specialties without escalating. These sanctioned interfaces already exist between those teams that allow them to work directly with each other. So in the DevOps community, we might find that in product teams or in SRE or ops engineering teams where they are matrixed into the development or product teams. All right. So here's why I think this is so important. And this is what Steve drew on those pieces of paper. When integrated problem solving requires going up and down the organization that work is slow and so much information is lost. Gene Kim (00:56:44): For example, to escalate and get two managers to talk to each other, that could require a week to schedule. If you want to get two VPs to talk, that might take multiple weeks. And as Admiral Richardson said, those VPs may get a very incomplete view of what is actually happening and may consequently make poor decisions. So those interactions are slow. There's often not enough information and whatever information is there, some of it is being lost. And as Steve mentioned, it's easy to overwhelm the people being escalated to, slowing things down even further. On the other hand, when teammates talk to each other or where there are sanctioned ways for teams to work with each other with a shared goal, integrated problem solving is very fast. I can turn around and talk to the person next to me, or I can call up someone on another team and say I'm missing something or something doesn't look right. Gene Kim (00:57:36): We can adjust, we can read and negotiate what that interface is. In other words, what is the information that must flow between two teams? This is a better and faster way than opening up a ticket or escalating an issue. It's faster, there's less information lost. In fact, I can imagine a whole bunch of scenarios where we're actually creating information, not losing information, and we're co-creating a solution that improves flow or the ability to achieve the goals of the system or the parts of the value stream that we occupied together. In fact, Steve has a great example of how Beth, the chief chemist in a pharmaceutical project turned to her counterpart in biology and said, "Hey, don't just send us the data. Why don't you come and present your interpretation of the data to all of us in a meeting and we'll do that every week?" Gene Kim (00:58:21): This resulted in an incredible wealth of information, knowledge that was created that helped everybody better achieve the goals of creating effective pharmaceuticals with more precision and less effort. So if it's not obvious, the language I'm using is inspired by the book, Thinking, Fast and Slow , by Dr. Daniel Kahneman. This book describes so much of the work that he did in collaboration with Dr. Amos Tversky. I'm quoting the Wikipedia entry on this book. The main is that of a dichotomy between two modes of thought. System one is instinctive and emotional, whereas system two is slower, more deliberative and more logical. The book delineates rational, non-rational motivations, and triggers associated with each thinking process and how they compliment each other. In other words, system one is where our biases live. Gene Kim (00:59:11): We love using that part of the brain because it uses less energy, whereas it takes a lot of work to get into system two. It uses more energy, it's difficult, but this is the cognitive problem solving style that we need to activate for complex problems. And this is indeed what is activated when we do things like writing things down. So in the domain of the topics discussed in this episode and the last one with Admiral Richardson, I would submit that there are two domains of activities where we need to use the slow communication paths, where we require the leaders at the highest levels. Part of that is in planning, and to some extent, preparation or rehearsal. These are the activities where we need leaders to be very thoughtful about defining the goals and then defining responsibilities and the structures to support them. Gene Kim (01:00:01): So many leaders are probably familiar with roles and responsibilities, but there's a third R that we need to care about, the relationships between those components. In other words, the interfaces. We also need those leaders to decompose the big problem into small problems so that they can be worked upon independently. This is where we absolutely need those up and down communications, but there are domains of activity that must operate using only the fast communication paths. This is absolutely where operations resides, or what Admiral Richardson sometimes calls performing or operations and execution. And I'll make the claim that it's usually a mistake to operate or perform using the slow communication paths because it's rarely fast enough. And when leaders overreach down into operations or when critical decisions always need to get approvals four levels up, things don't go so well. Gene Kim (01:00:52): But the role of the slow communications come back when we assess and improve. This is when we ask, are we achieving our goal? And if not, how do we change it so we can achieve our goal? This is when General Stanley McChrystal asks in Team of Teams, "As good as we are winning tactically, are we achieving the strategic objectives?" And his emphatic answer was, "Absolutely not." These are what we do in agile retrospectives. This is what happens in football games during the huddles or during halftime. So in my previous interview with Dave Silverman and Jessica Rife, they talked extensively about how and why mid-level leaders need to act and think differently so that they can enable and foster these very different types of interactions across the entire organization. Personally, I am finding this mental model of slow and fast communication paths to be very helpful. Gene Kim (01:01:42): And I'll just make a last side note here before I wrap up is that, fast and slow is really collapsing a couple of attributes. In control theory, there are really four attributes, not just fast and slow. There's frequency of communications, there's speed of the response, so are we acting on old information or are we acting on near real-time present conditions? Granularity in detail of the communication and accuracy or fidelity of the communication. And so when you hear me say, there are fast and slow paths, think of that as a proxy for all four; frequency, speed, granularity and accuracy. In operations, we tend to favor frequency and speed. For the deeply cognitive planning activities, we need granularity and accuracy. After all, we don't want to make decisions on information that is actually false. Gene Kim (01:02:36): Okay. Let's go back to the interview where I asked Admiral Richardson whether the ideas that Steve proposed, those mental models, do they resonate with him. One to 10, to what extent does that resonated with you? One is like, oh, politely acknowledge and we'll move on, 10 is like, oh, no, that exactly describes the dynamic that you painted, especially as you push the functional integration [crosstalk 01:03:00]. Admiral Richardson (01:03:00): Yeah. I think very high, a nine, 10, right? Because the other thing that is interesting is that in kind of an automobile manufacturing situation, you kind of have full control over the tempo, right? So whether you're operating fast like Toyota or you're operating slower like somebody else, you control all that and you just put out fewer changes or fewer improvements per time. In the Team of Teams or in an operational environment, of course, there's only so much that you do control, right? And so out at that edge, kind of going back to our point earlier, Gene, you either operate at that tempo or you don't. And the absence of a decision is a decision, right? I mean, things are going to move on and you've missed that opportunity. And I think there's an IT corollary now too, which is this processing at the edge, right? Admiral Richardson (01:03:56): So it's much more burdensome on the system to try and communicate a lot of raw data, just sensor data. But if you can make sense of it out at the edge and transmit the information, right? Then that's a much lower bandwidth type of a thing. So I think that there's going to be some virtue to awareness, some centralized awareness. That's going to have to arise in some way. But in terms of response, a lot of what Steve said is exactly what we're after, right? Teams moving out at the edge, and I leave it to you, Gene, to figure out how to get these pictures into the podcast. Gene Kim (01:04:42): Oh, we will. We'll figure it out. Steven Spear (01:04:44): If I could just, last interjection on this topic. So the point about, at what point does the center get involved? During the planning and during the configuration of the system. So you can have the center involved when you're doing your contemplative work, because this thing is not yet moving so the computational, the cognitive cycle times of this thing aren't an impediment. So when this thing can't be involved is when this piece is actually operating. And once this thing is actually operating, you have to take that slow processor out because the underlying process is moving quickly. So how do you reconcile the both of having to have system level conceptualization? So that's where flexibility comes in, which is we can do this quickly, operate quickly, stop, come back and do this quickly again, stop, come back and do this quickly again. Steven Spear (01:05:55): And this is where you get your truly agile systems, which is this ability to reconfigure and flip-flop very quickly between the- PART 3 OF 4 ENDS [01:06:04] Steven Spear (01:06:02): ... figure, and flip-flop very quickly between the, how do we call it? This is the conceptual, and this is the operational. Admiral Richardson (01:06:09): Yeah. Steven Spear (01:06:09): And one last thing I'm just going to throw out. So, just in defense of my Toyota friends, by the way, they work in a very hostile environment. It just happens to have cycle times different than team of teams. They have competitors tbat are constantly trying to eat their lunch, and there's weather. I think I told this to Gene, and then I will stop. The plant in which they were making these 60 line changes a day, that was one set of Kaizen, the other was narrowing the window. Steven Spear (01:06:39): Now, remember they're sourcing parts from all over Japan, and Japan's got crappy roads and weather, but they were trying to narrow the window in which trucks were arriving from a minute to 30 seconds. Because they were operating with so little inventory, and consequently such frequent delivery, that if Gene arrived at 12:01 at the unloading dock, but he was supposed to be there at 12:02, he'd be screwing up with my 12:01 delivery. And they were trying to narrow the variance down to a 30 second window from a one minute window. And the hostility of the environment there, is weather and traffic and everything else. Admiral Richardson (01:07:23): Yeah, the friction, right? The unavoidable friction of water and fog. So, you got to accommodate for that. And I think what Steve was just describing, is exactly what we were saying, is that there is a role for going back to the center. And so you do these iterative operational steps, that happens a lot out at the edge, but then you have to say, okay, how did that step go? And are we making progress towards our overall strategic objective? Particularly against, as Steve pointed out, whether it's in automobile industry or out in an operational environment. Hey, you've got a learning and adaptive competitor here as well. And so, only so much of your initial assumptions are going to hold for only so long. And so, you're going to have to come back and rethink things. Gene Kim (01:08:13): Your point about integration coming down, my impression, the army went through that a number of years ago, where the point of integration was at the division level, and they moved it down to the brigade level. And I'm willing to bet that some of their high operating tempo units actually integrate, have degrees of integration at the battalion level and the light infantry. But I know a guy we can talk to, who's in the nitty gritty of that. He's had various commands. Admiral Richardson (01:08:42): When you say integration though, how do you mean that? Because what we're talking about is integrating people into the organization. It sounds like you're talking about product integration or something like that. Gene Kim (01:08:56): So, what I mean specifically, and we all have our functional specialties. And at what point in our stove pipe, do we answer, how high in our stove pipe do we answer, before there's a cross-functional responsibility? Admiral Richardson (01:09:13): Okay, got it. Yeah, so happy to articulate on that, Gene. Our smallest unit of integration would be a ship's crew. And those come in various sizes, dependant upon the size of the ship. But you're always trained to learn the basics of the job of the person next to you, because in a casualty situation or something like that, you might have to pick up the responsibilities of that person, and also up and down. And then there's just some functions that everybody has to know. Everybody has to know firefighting on a submarine, because it's going to be an all-hands effort to get that thing out before all the oxygen goes away. And so- Steven Spear (01:09:59): That's someone else's job. Admiral Richardson (01:10:00): That's someone else's job, it gets very personal, very fast. Gene Kim (01:10:08): So I think by the book, there's the spectrum on one side to adapt quickly. You want to push decision-making to the edges of cross-functional teams. Then on the other extreme are these functional organizations now where you optimize maybe less for speed, but more for functional expertise. And one of the big surprises to me researching right before this interview was to look at the six, four star positions in the US Navy. And I was genuinely surprised to see that one of them was for NR, the nuclear reactor core. So it seems like we've been talking a lot about pushing expertise to the edges. It seems like the NR is one where it's actually very much in the core. Can you just talk about why that is and why NR is so important that it isn't just delegated away a little further down in the organization? Admiral Richardson (01:11:01): Right. I'd like to just pause here and just recognize the operational genius of Admiral Rickover when he set this organization up, right? And the foresight that he had to establish the culture and the business rules. So much of that continues today, even decades later. And the things that they did very early on have proven to be very wise, right? I mean, we're not cleaning up a lot of messes that were created by bad guesses back then, right? So it really was a remarkable beginning of a brand new technology, right? So, and Steve's [inaudible 01:11:44] shared on innovation, right? I mean, the speed at which that happened, Gene, where I think it was 1938, where the idea that you could get energy from splitting an atom was discussed, it was 1942 or 43 when Enrico Fermi underneath the stadium in Chicago actually did it. Admiral Richardson (01:12:06): It was literally 10 years later we were launching a nuclear submarine whose entire power plant was based on that theoretical principle, right? So, turning that all that theoretical idea into an engineering plant that propelled a submarine or an aircraft carrier, just remarkable speed. And you can imagine the amount of innovation and new ideas that had to be brought to bear very mindful that we're just talking about a tremendous amount of energy here. And so a miscalculation could be just disastrous, right? So that's the reason that Naval Reactors exists as a separate organization. Characterized often as a highly centralized organization, but it actually represents as an organism much more along the lines that we've been describing. And so, the center gets involved very much with the establishment of new policies, the maintenance of existing policies that have proven their value over time. Admiral Richardson (01:13:16): And so that type of commander's guidance, if you will, has maintained at the center, but the execution is all done by field offices, right? And so they're existing in all of those places where we have nuclear power warships around the country, the labs that support those, it's actually very decentralized in execution. And so there's a lot of conversations, command and feedback, right? When I was at the director, it was every week I was talking to the head representative in each of those field offices. And it was really just a listening exercise. How is it going out there? Do we need to change anything? If so, let's do it, but otherwise the center's going to stay out of your business. And I just want to understand, are we achieving the aims that we set out to achieve? It's actually within those bounds, those guard rails, which keep things safe, a tremendous amount of creativity happens within the Naval Nuclear Propulsion Program. Gene Kim (01:14:14): And ss there an easy answer of why the NR has a seat at that highest level table? Admiral Richardson (01:14:21): Well, I think one, the consequences of a miscalculation or a mistake in that realm, just the responsibility that comes with that attention at that level. The other thing about Naval Reactors is that they are a cradle to grave organization, right? And so, from the design of this technology, building the fuel, the nuclear fuel, bringing in the people and training the people and then moving all of that through to putting the propulsion plants together, testing them, and then regulating that. There's regulatory authority involved. And then, at the end of life, how do we then take the old fuel off or even dispose of the, I mean, it's a huge, huge, huge enterprise. And so it's worthy of that level of seniority. Gene Kim (01:15:16): Wonderful. And now that you explain it that way, it suddenly becomes very clear what the gravity of all that. Steven Spear (01:15:24): To interject on that point, that gets back to, and I know we discussed earlier, at what point is value created, and captured. And so everything that was just described as of Naval Reactors is actually presenting to the rest of the world as this module, which is you turn to Naval Reactors for everything related to propulsion, including the equipment, the fuel, the people, et cetera, et cetera. Internally, in order for that to work, there has to be a lot of integration of the parts, right? So that argues for that being encapsulated. Gene Kim (01:16:02): And some of those obligations span generations as this is the timescale of those commitments is just breathtaking. Admiral Richardson (01:16:10): Yeah. I mean, we are talking about a design that has to operate for 50 years. The entire complex blending of technology, the human element to everything, that's a 50 year lifetime, easy. Gene Kim (01:16:28): Gene here. I want to give some background behind the next question. I wanted to ask Admiral Richardson, what should happen when complex systems go wrong? In his ideal, what really should happen? Here are some examples that I gave where there was something deeply unsatisfactory about the outcomes. In the Volkswagen emission scandal, one engineer was held responsible for the firmware changes to the engine management systems that allowed Volkswagen to cheat on US Federal Emission Standards in the wake of the Equifax data breach, one of the largest data breaches of Us citizens. Then former CEO, Richard Smith stated that the breach was a result of human error and a technical failure. In this case, a security engineer who missed something and a technical failure being a software license not being renewed. Gene Kim (01:17:19): In the Deepwater Horizon disaster that led to the 11 deaths when a drilling rig exploded in the Gulf of Mexico causing the nation's worst oil spill, 134 million gallons, the person who faced federal prosecution were two supervisors for quote, botching and disregarding two safety tests. In each of these cases, there's something deeply unsatisfactory about those that sure seems like accountability and responsibility should go higher up the chain. So I asked Admiral Richardson, what does he think about that? Something unsatisfactory about each of those, that seems like responsibilities should go a little bit higher up the chain. So as a leader, how do you think we should really think about these events, where there are systemic, accidents happen in these systemic large complex systems? Admiral Richardson (01:18:09): Yeah, well, I think when you're talking about accountability, first step I think, is to do a thorough investigation, right? And so it might be tempting to take action. And thorough does not necessarily mean long, right? It doesn't have to last forever, but I think the decision-maker, who is going to think about accountability needs to have all of the facts and also a little bit of time. I think there's always an emotional response to these things. And so a little bit of time will allow emotions maybe to get to a more appropriate level. And so that would be the first thing. Then a lot of times what I would try and communicate is that particularly as you get higher and higher in the organization, the leaders are accountable for success, right? Admiral Richardson (01:19:12): And so I think a lot of times we conflate the word accountability with, "Hey, you're fired," right? And sometimes that's appropriate, but oftentimes it's really a matter of, "Hey, I think you just needed to be trained a little bit better," or there are some other methods by which we can address what happens, right? Now I would say we talked a lot in the first interview about the role of values and the ethical approach. Some of the examples you cited, I think are more in that realm, right? And if the values have been clearly communicated, if this leader has been assigned and there's an understanding that they get that, right? When somebody asks contrary to the values of the organization, I think that undermines the trust and confidence that is so important to this radical delegation, that can be, I think, quickly resolved and this person may not have a role on the team. Admiral Richardson (01:20:25): And then there are some things that are just so egregious, particularly in the Navy, there's this value or this sense of accountability of the commanding officer, the accountability of command at sea, and that can be stark. And so we put a lot of value in that. And so, oftentimes some of the measures for accountability result in the relief of that officer, but you're also right. That if somehow it's always the more junior people that are dismissed and you've got, I think that is really the result of either an incomplete investigation that really hasn't asked all the hard questions. You got to ask why five times before you start to get insight. Or it's a little bit of the, maybe the fox guarding the henhouse, right? That the people that are doing the investigation and meeting out accountability, they've just got too much interest in the outcome, right? So you need a bit of a detachment or separation that happens there. Gene Kim (01:21:43): And one of the things that I've heard you say is the more senior you get, the more simply you need to speak. Is that, can you just say a little bit about that and why you think that's true? Admiral Richardson (01:21:54): This is just, I guess my experience, I think it goes back to the understanding that everybody's going to look at what they see through the lens that they've got. Perhaps there's that old saying, "Hey, I didn't have time to write you a short letter. So I wrote you a long one." Right? And so I think it's incumbent on more and more senior leaders to be more and more thoughtful about what they're saying, the guidance they're providing so that they can provide that in as sparse and terse terms as possible that are more and more immune to misinterpretation, right? And so that's, I think, what I'm trying to capture there. And I'm a terrible example of that during this podcast, because I've been so proposed to shooting all over. But I think it goes back to this commander's guidance, right? Hey, say what needs to be said, and don't say anything else, right? Don't get extraneous because all of that is going to be seen through a multiplicity of perspectives and could be more and more subject to misinterpretation. So keep it as sparse and as terse as possible. Gene Kim (01:23:11): And maybe just to respond to that, never had it crossed my mind in the last two hours of you talking too much. Everything has been amazing. Admiral Richardson, thank you so much for all the time you spent with me and Steve. I know that everything that you've shared will be so helpful for technology leaders. Can you tell us how people can reach you and what types of outreach you would enjoy the most? Admiral Richardson (01:23:36): In act two of by professional life here, I'm on my own here. I do look forward to continuing this dialogue and whether I can be of any further help also, I'm learning, my learning curve is absolutely vertical right now. So every time I talk to someone, I learned a ton. And so if you want to get ahold of me, please just email Gene and his team and they'll reach out and touch me and we'll get right back to you. Gene Kim (01:24:02): I hope you enjoyed that interview just as much as I did. Coming up next is my interview with Dr. Ron Westrum, who created the famous Westrum Organizational Typology Model, which is featured so prominently in the state of DevOps research that I did with Dr. Nicole Forsgren and Jez Humble. So many of us in the DevOps community have read his work, but few of us have heard or seen him speak. Well, that's about to change. I'm so delighted that we'll be bringing you two interviews of his work, which is so relevant to everyone, not just in technology leadership, but all leadership. And don't forget to join me and all the amazing speakers at IT revolutions upcoming DevOps Enterprise Summit Virtual, Europe from May 18th to the 20th. See you then. PART 4 OF 4 ENDS [01:24:54]