Monday, 2021 June 07

I switched jobs recently, moving from Indeed.com to Cobalt.io. In the process of doing this I rather inadvertently got myself promoted to "Staff Engineer", a role I've been flirting with but mostly avoiding for the last few years. This is more of a technical leadership role and, in an effort to not be completely incompetent, I've started reading a series of "management" and "leadership" books. This is actually the second such book I read, but the first one I'm actually going to review. There will be a steady stream of these reviews going forward (that's my plan, at least).

Summary

This is a review of L. David Marquet's Turn the Ship Around, which describes his experience taking over as the captain of a low-performing United States Navy submarine and completely transforming the ship's culture, in the process turning it into one of the highest-performing ships in the fleet. I borrowed the Kindle edition of the book from the local library.

The executive summary is that organizations should strive to be leader-leader (versus leader-follower); every single person in the organization should have the authority to make meaningful decisions at their level within the organization. Additionally, however, they should bear responsibility for those decisions, and decisions should be made in a way that furthers organizational goals. This necessarily means that the person will need to know sufficient context to not only know the goals, but be able to reason through seemingly contradictory goals to arrive at the correct decision - at least be able to explain why they made the decision that they did.

Marquet's argument is quite persuasive and he gives many examples demonstrating improvements to his crew after instituting this method (and many supporting methods as well). However, in my mind one major issue is ever really addressed - everything in this book takes place within the framework of the US Navy, which is a massive well-funded organization with an extremely strong culture, a proud history, a strong sense of purpose, and strong discipline. I don't feel like Marquet ever really discusses how to apply these lessons to corporations which may lack in any or all of the qualities I listed above.

Still, even if all of the lessons aren't universally applicable, a number of them could be applied in just about any situation. This book is well worth reading.

Notes

This was a fast read. In addition to the management aspect of things, there were a lot of details about the day-to-day operations of nuclear submarines. For a military nerd like me, this was a huge bonus.

The book is broken up into 4 parts. Each part has a one-page summary, then some number of chapters. Each chapter is short, with a list of questions to consider at the end. The intent of those questions is to make you think more deeply about the lesson of the chapter.

Introduction

A summary of the book. Highlights:

Part 1

This is mostly background, how Marquet came to reject the leader-follower dynamic.

Chapter 1

Attempts to empower subordinates within a leader-follower structure will fail. "One of the things that limits our learning is our belief that we already know something."

Chapter 2

Are you optimizing your organization for success during your tenure, or for the long-term success of the organization as a whole? "When the performance of a unit goes down after an officer leaves, it is taken as a sign that he was a good leader." But what it often means is that the officer was a micro-manager who refused to divest any control, leaving behind subordinates with no idea what to do by themselves. Marquet uses the term "induced numbness" in subordinates who don't have to make the hard decisions or be responsible/accountable. "I was just doing what I was told."

Chapter 3

A new command gave Marquet a chance to try out his ideas. Since he was directly responsible for the performance of the sub (the USS Santa Fe), it was a big personal risk. If his approach failed, he'd probably never command again.

Chapter 4

Marquet interviews his chiefs and officers, listening to their views on problems with the ship and its culture - but also things they believe are going well. Simply starting a conversation like this started to give Marquet some personal credibility, simply because it was so novel.

Chapter 5

The interviews continue. A key finding: even when officers wanted to improve things, it was hard to get approval from above. This led to frustration and resignation, an "overwhelming sense on the ship was that we needed to avoid problems". This attitude is not compatible with excellence.

Chapter 6

Conversations with the crew. Marquet: "What do you do?" The answer: "Whatever they tell me to do." Marquet's key takeaway was that everything about the current operation of the ship "reinforced the notion that the guys at the top were the leaders and the rest of the crew were the followers".

Chapter 7

Marquet takes command with the goal of achieving excellence, not just avoiding errors.

Part 2

This section is concerned with the process of beginning to divest and distribute control. "Don't move information to authority, move authority to the information."

Chapter 8

Marquet starts his change effort with the chiefs - there are enough of them (both by count and by authority) to significantly change the culture, while remaining a manageable group for him to interact with personally.

Their role could be considered one of privilege (significant authority without much responsibility as the chain of command was then structured). This meant that Marquet really need to achieve a certain amount of buy-in and move cautiously, as he was asking them to take on more true responsibility. One way of doing this is to rewrite the genetic code of the organization - change policies, etc to reinforce the cultural changes being made.

Chapter 9

Divesting responsibility to the chiefs improved behavior almost immediately - the increased authority of the chiefs meant that crew was more respectful. Previously mouthing off had little change of being punished because the chiefs didn't have authority to do so - but now they did.

However, there was also a fear of being different - some chiefs were very uncomfortable with being asked to deviate from the "standard" navy culture, as they felt they were possibly risking their career for unknown benefits. Marquet had to address this by his own actions - after all, he was ultimately responsible for the entire boat - so he cultivated an attitude of "caring but not caring" - caring deeply about his subordinates and the overall organization but not caring about consequences to himself/his career.

This chapter concludes with a discussion of achieving cultural change through changes in behavior. Often this is the only way of doing it - because it's all well and good to try to change culture and wait for the resulting changes in behavior, but it's hard to tell if your efforts are working as the changes happen so slowly. So instead Marquet recommends "acting your way to new thinking" - if there's not time to change the way his officers thought, just change the way they act instead. This in turn should filter down to the lower ranks.

Morale was low because the crew believed they were "not proactive movers but only passive reactors to external events".

Chapter 10

This chapter leads off with an example, specifically how long and involved the charting process is. It requires many parties; the sailors who start the work may have imperfect knowledge of all of the goals/constraints, meaning that sometimes work must be thrown away and redone. Due to differences in process, sometimes the end result is inconsistent (such as color coding), possibly resulting in confusion at crucial times.

A discussion of this resulted in a new process. Instead of a game of telephone down the chain of command, the sailors were encouraged to seek clarification from the captain up front if necessary. The point is made that on a sub the captain's attention is a valuable, limited resource - but at the same time 30 seconds of the captain's attention early on can sometimes save literal hours of effort on the part of the rest of the crew. On balance, sometimes it's worth distracting the captain for the greater good. "Short early conversations make efficient work." (There is also a brief discussion of the benefits of conventions and consistency, but this is very much a secondary point.)

The larger point is that in a leader-leader system, supervisors must accept that they will be asked for early feedback/clarifications on the work of a subordinate. More importantly, they must realize that they will often see preliminary, incomplete, or even just messy work in early stages of development. They must be able to give useful feedback on assumptions, process, and the end goal while not being critical of the details of how the product evolves - being allowed to see such work is a sign of trust from the subordinate, and is definitely not a failing for anyone involved.

Another critical point is that seeking early feedback is definitely not a matter of a subordinate not knowing how to do their work - instead it's a matter of the supervisor having a broader perspective, being aware of additional restrictions or context, etc. A supervisor reviewing an early draft and responding with (for example) "your assumptions are wrong because you didn't know this piece of information, you'll need to redo this" does not indicate a lack of trust on the part of the supervisor.

Chapter 11

This chapter starts with Marquet issuing an order that was impossible to comply with on the Santa Fe - it was valid on earlier subs he was on, but not this one. Nevertheless a subordinate relayed his command, despite knowing it could not be carried out.

This leads into Marquet re-introducing the "I intend to" concept from the Sunfish. This is a way to turn passive followers into active leaders - by having them propose what needs to be done next, they are essentially operating at the next higher level.

Expanding on this, supervisors should avoid asking clarifying questions - the subordinate should be encouraged to provide all of the information they used to make the decision that they are proposing to act upon. Ideally the only response from the supervisor should be "very well".

Chapter 12

This chapter's example is one in which Marquet issued an order but did not explain the reasoning behind it. He then went to sleep, but woke up to discover that the situation was not quite what he had expected. This was because his subordinates, lacking context, made a series of short-term decisions that they believed would achieve the desired goal, but in fact did not.

The solution here is to provide more context - which on the surface seems similar to the lesson of chapter 10, but in this chapter is focused more on the supervisor pro-actively providing all of the context necessary to make a good decision, rather than asking subordinates to seek clarification.

In another exercise later, after applying this method, Marquet was happy to be able to watch his subordinates consult with one another and arrive at a good decision without him intervening - because they were now fully aware of the end goal and the reasoning behind it.

A secondary lesson here is that almost never must a decision be made immediately. Even in times of war there often are seconds, or even minutes, available in which to think through a decision, consult with others, and come up with a better decision than simply doing the first thing that comes to mind.

Chapter 13

This chapter discusses the (general) uselessness, even counter-productiveness, of top-down monitoring systems. The goal should be for people at each level of responsibility to have the ability to monitor their own domains and the ability to make corrections/changes when necessary.

It's strongly noted that this is different from top-down collection of data, definition of processes, and defining metrics. These can and must be done without judgement.

With top-down monitoring, the goal often becomes to avoid errors in process, and any such errors are addressed via more supervision/monitoring. This is institutional dead weight that slows down everything.

The overall theme is that there should be a match between the authority of a given person and their responsibility, and how to ensure that this is the case.

Chapter 14

Continuing the theme of sharing information, this chapter is about thinking out loud. This helps everyone nearby understand the reasoning behind your decisions - and challenge your assumptions or provide additional information if necessary or appropriate.

Chapter 15

External oversight should be embraced, not greeted with dread. Marquet specifically gives an example of meeting an external inspection committee with a list of known issues and deficiencies with the Santa Fe - but all of those items had proven difficult or impossible to address at ship level. Marquet's goal was to raise visibility through the committee, which had more power to address such high-level issues than he did.

Another value provided by external overseers is best practices - treat them as a valuable resource. If your group is struggling with something, it may be problematic for other groups as well and the overseers may have suggestions for better practices, or at least can provide you with ideas of what has not helped on other groups.

Part 3

This section concerns building competence, resulting in confidence and the ability for subordinates to understand the larger picture and make appropriate decisions.

Chapter 16

A sailor makes an honest mistake; the resulting inquiry reveals that he knew the correct procedures and simply didn't follow them because he wasn't really paying attention. Marquet decides that there is really no need to punish the sailor - he was honest about what happened - but instead considers how to ensure that procedures are more likely to be followed without adding additional process. The result was an effort to make every action deliberate - when following a checklist, for example, sailors should state the action they were about to take out loud, pause, and then perform it. Even when nobody was there, this would force them to pay at least a little attention and consider their actions. When others were present, everyone could act as a safety check.

One objection was that this would slow things down in an actual high-pressure situation - but those are the situations in which mistakes are most likely to be made and when actions should be most deliberate.

Later on, after scoring particularly well in an inspection, the inspector said the Santa Fe's crew "made the same mistakes - no, [they] tried to make the same number of mistakes - as everyone else. But the mistakes never happened because of deliberate action. Either they were corrected by the operator himself or by a teammate."

Chapter 17

This starts out with another account of an accident on the Santa Fe, this one in the torpedo room. Unlike the previous example, in this situation it was clear that nobody involved really understood the proper procedures - a mistake was almost inevitable. How could this be avoided in the future? The answer on the ship was "continual learning" to help people learn to perform at the next level, to match their authority.

This chapter lays the groundwork for "we learn", which become the (semi-official) motto of the crew of the Santa Fe. The gist of it is that everyone should be encouraged and expected to learn all of the time, with the goal of becoming more competent at their job.

Chapter 18

Marquet decides the crew needs to practice diving - it was taking way too long. On one of the drills he throws a simulated gauge failure at the crew to see how they handle it.

The existing procedure was for the dive officer to "brief the dive", which basically involved him reading off the procedure out loud. In theory this was to refresh everyone's memory - but in reality nobody was paying attention, and confusion during the action process was common.

They decided to turn this process on its head - instead of the dive officer reading off the procedure, all of the sailors involved in a process would recite their part of the process (in overall order) and the dive officer would "certify" that everyone knew what they were doing before beginning to dive. (This process was applied generally within the boat, not just for diving.)

This revealed a lot of confusion and lack of knowledge/preparation - as expected. There were many reasons, but most of them could be addressed via "we learn". Once everyone knew that they were expected to know the proper steps for a given procedure, even if they might be stationed at one of several locations, they realized that they actually did need to deeply understand how it all worked.

Chapter 19

A sailor goes AWOL - in a relatively low-key fashion, but still a major issue in the military. Marquet throws all precedent to the wind and takes a compassionate approach, going so far as to pursue the sailor himself to have a chance to talk things over.

There is a lot of military process in this chapter, but the gist of the problem was that:

  1. The sailor was heavily overworked due to a lack of qualified people willing to stand watch.
  2. It turned out that there were additional sailors on board who were technically qualified, but they were of sufficient rank that they had excused themselves - e.g. they had turned their rank into privilege.

Once Marquet became aware of this one example, he dug into the duty rosters and discovered this was not uncommon - despite him stressing early that rank did not equal privilege, this lesson had only been taken to heart by a fraction of the crew. He was not happy, but he resisted the urge to take back any direct supervision. Instead he instituted a new rule: supervisors were not allowed to have an easier watch schedule than any of their subordinates. His chiefs were not happy, and he was a little upset that he had to exercise his authority to reinforce this lesson.

The lesson Marquet took away was that he had not done a good job of reinforcing his lessons, and he was perhaps a little inconsistent in how he presented things. In the future he'd focus on more consistency, but in the lesson being delivered as well as the frequency of the lesson.

Chapter 20

One month before deployment, the Santa Fe seems to be performing well - it's going to meet the readiness deadline, and Marquet feels that the crew is working together as a unit. However, some things still concern him.

In particular, fire drills. Fire is incredibly dangerous on subs, but the procedures around them didn't necessarily seem to be the most efficient. (To be fair, the procedures had evolved over the years and there were reasons for every part of them - but that didn't mean that the end result was the best possible approach.) It was a clear case of the procedure taking precedent over the end result.

There are a lot of details given about the new process, but it basically boiled down to "you all know how to fight a fire, so it's up to you to organize yourselves and do what's necessary". Additionally, thinking out loud was encouraged to better help the self-organization. Drills were changed to make them more realistic. The ship eventually would win awards for the speed with which they addressed fire drills.

A second example is given, about internally-generated noises. Self-reporting was encouraged, with no recrimination, and this resulted in a much quieter ship overall - suddenly everyone felt responsible for the noises they made.

The overall lesson is "specify goals, not methods".

Part 4

This section, taking place as the ship is being deployed, concerns "clarity" - everyone at all levels of an organization should understand the goals and purpose of the organization. This clarity will help them to make the correct decisions even when lacking key information - because they will know what the end goal is.

Chapter 21

The Santa Fe is under way towards its deployment position. Marquet decides that everyone should have specific goals they are working towards to keep them engaged and to take up any slack they may be feeling after the rush towards deployment readiness.

Promotions for sailors on the ship had not gone well. Marquet set about fixing this by making sure that they all understood what exactly was necessary for a promotion as well as tying preparation into their goals.

This is part of the larger lesson of this chapter - "taking care of your people extends beyond their work lives". Some examples are given, and at the end of the chapter it's revealed promotions started coming through - as did re-enlistments. The Santa Fe was becoming a desirable posting.

Chapter 22

This chapter concerns honoring organizational traditions and legacies and using them for inspiration. The US submarine force has a very rich history, and there are many little ways in which modern submariners can and do honor it.

This chapter actually starts out with a mostly-unrelated example of a subordinate helping out a confused supervisor during a tense drill by anticipating the next action and placing his hand accordingly, allowing the supervisor to regain their place in the procedure and carry on without embarrassment. This is also a valuable lesson.

Chapter 23

Define guiding principles for your organization and use those to help make decisions. I feel like this lesson has been stressed multiple times.

Chapter 24

Immediately recognize and reward actions that go "above and beyond". Marquet is careful to stress that these should be measured against objective criteria; if the goal is to put out a fire in under 1 minute, reward every team that accomplishes the goal. On the other hand, if every team takes more than three minutes to extinguish a fire, there's no point in rewarding the least-slow team - because in real life everyone would have died anyway.

Care should also be taken to avoid pitting people directly against one another - at the end of the day everyone is still a collaborator. This is another reason that directly measurable criteria are superior for deciding rewards rather than head-to-head competition.

Chapter 25

Marquet had set a goal of being ready for deployment two weeks ahead of the actual deadline. This is common practice - but he actually meant it, because he wanted his crew to have some down time before their six-month deployment began. However, external parties often ignored ship goals in favor of the hard deadline, leading to little or no leave for crews despite their own best efforts.

"Begin with the end in mind." Marquet applied this concept both to figuring out deadlines for when things had to happen to achieve his readiness goal as well as to his personnel. Everyone would write their review for the next year (with measurable goals), then their supervisor would work with them to set incremental goals to reach the desired end state.

This becomes a larger discussion of mentoring. Marquet, analogous to leader-leader, stresses that this should be a mentor-mentor relationship, as both parties should be learning from the process.

Chapter 26

This chapter is a series of stories demonstrating the improvement of the Santa Fe crew, and illustrating how well they had adopted many of the concepts described earlier. It culminates in a subordinate informing Marquet that a rushed order was incorrect, then explaining why.

Chapter 27

This chapter is basically a summary of the key concepts from earlier chapters.

Chapter 28

A final example, demonstrating both the overall effectiveness of the crew as well as their ability to interact with external parties in a non-standard but highly effective fashion.

Chapter 29

Looking back, Marquet discusses the overall impact the Santa Fe's crew has had on the navy.

Friday, 2021 June 18 Sunday, 2021 June 06