Every manager has to deal with people leaving. When they’re valuable employees, you want to do everything you can to motivate them to stay. Here’s the paradox: if you’re a good manager, you’re going to be powerless to prevent their departure.

People leave teams and companies for a few different reasons:

• Overworked
• Underpaid
• Under-appreciated
• Dislike their coworkers
• Dislike their manager
• Dislike the environment
• Work is not engaging
• Disinterested in the product

Every single one of these is something that is usually in your power to, if not fix, then at least influence. Of course, the employee could be wrong about all of these things. If they are, then there’s a discrepancy between their perception and yours. You either educate them, educate yourself, or live with it. If you can’t or won’t resolve it, then maybe it’s not such a problem for you when they leave.

Then there are the difficult situations where you recognize these problems, want to fix them, but are powerless to do so. Can you blame someone for wanting to leave then? And if you were powerless before they announced their resignation, you’re not going to be any less so after.

But let’s suppose you are appropriately empowered. If you’re a good manager, you’re listening to your people and attending to their concerns, preventing and preempting problems. You monitor and adjust their projects and commitments to ensure a good work/life balance. You find money to give them a competitive pay package. You recognize and appreciate their accomplishments. You foster positive relationships and intervene when they’re going bad. You do your best to be fair, attentive, caring, trustworthy, and all the other things someone needs from their manager. You work with other leadership to establish and improve the environment. You steer people toward work that gives them the right mix of challenge and mastery, and you help them connect to the product and the customers.

In spite of all this, one day they tell you you’re leaving. You’ve been a good manager. You’ve done everything you realistically can do to shape their job to fit their needs, but it wasn’t enough. In that moment, what is there that you can do? If you’ve been a good manager, the answer is probably almost nothing.

On the other hand, suppose you can think of a number of counter-moves. You can move them to a different project, change their manager, give them a raise, or something else to address the problems they described when they explained their decision. Why do you have those options available only when the person tells you they’re quitting? Why weren’t you tending to their needs before? Why weren’t you working proactively to make this the best job you could? It might be that you’re, if not a bad manager, perhaps just not a good one.

That’s the distressing fact about dealing with unwanted departures. If you can do anything to convince them to stay, then you failed as a manager. If you did your best already, then there’s nothing more you can do.

It starts innocently: “Hey Fatima. We should have a regular meeting, just you, Esteban, and me. There’s enough stuff that we work on together that it would be more efficient to get through it all at once.”

Then it grows: “I think we should add Divya to the meeting. I’d like to get her perspective, and I think it would help her to be part of the discussions.”

People ask to be part of it: “Divya and I are the same level, so I think I should be in the meeting too.”

The agenda gets filled up with a grab bag of small items, and the meeting loses focus. There are twenty people in the room, but most of the time only three or four of them are paying attention.

The original attendees start skipping the meeting because it is an unproductive use of time. The original meeting, having lost its senior sponsorship, loses its value and eventually gets cancelled.

But those senior people still need to get the same stuff done: “Hey Fatima. We should have a regular meeting, just you, Esteban, and me. There’s enough stuff that we work on together that it would be more efficient to get through it all at once.” And so we begin again.

Why does this happen?

• The original meeting has a loosely or implicitly defined agenda and membership
• Attendees of the meeting make isolated (versus holistic) decisions to add more people
• People who hear about the meeting dislike being out of the loop and ask to join
• Too few people get rejected, and nobody gets uninvited from the meeting because it’s socially awkward to do so, especially since it’s not clear who should be there and who shouldn’t
• People can be uncomfortable with speaking for people not in the meeting

If you can solve each of the problems listed above, then you’ll break the cycle of death and rebirth and achieve meeting nirvana*.

There’s one last thing to add: People don’t understand meeting costs. This seems like an obvious one. Estimate each person’s hourly cost and multiply it by the number of people in the room. That’s absolutely true, and it’s not considered often enough. But it gets worse.

The more people there are in the meeting, the more diverse the agenda. The more diverse the agenda, the smaller the proportion of people who are interested in it. Also, only one person can talk at a time in a meeting. That means the bandwidth for discussion and decisions is fixed no matter how many people are present, but it’s a declining proportion of attendance.

Another way to look at it is this: every person who is added to a meeting makes it more costly and less efficient for everyone else in the meeting. It’s the opposite of a network effect. Usually people only look at the benefits of adding someone to a meeting, but it’s not enough that the person bring benefits. They have to bring benefits sufficiently great to exceed the total cost inflicted on everyone else by their presence. The default expectation is that the average person added to a meeting is net negative. If I’m in your meeting, you’d better be getting something significant from me being there.

The gaps between the rungs on the career ladder are not evenly spaced These gaps happen when your old way of understanding your job no longer works and must be replaced by a new paradigm.

One of the big gaps is somewhere between first level management and senior management. The early to middle stages of a career are all about doing more and doing better according to a well-defined job description. The next levels are defined very differently. You can no longer do more, because everyone’s capacity is finite. Not can you make as much progress on doing better, because there are so many different things you have to do. Instead, it’s about doing the most important things adequately. This switch is hard because the intuition on how to get ahead that has been trained and refined and reinforced for years stops being the right one. While the right strategy isn’t exactly the opposite, it is hard to let go of the previous strategy that worked so well.

The second difficulty is in knowing what those most important things are and what adequate means. More senior positions don’t usually have a limited and clearly defined set of activities. It’s more likely that the job is effectively “do the right thing in everything related to X.” Nobody tells you in advance what “the right thing” is. Nor do they tell you exactly what they mean by “everything” or “related.” You can’t even copy someone else, because nobody has a job similar enough to yours. You can’t even copy your predecessor, because she had different strengths and weaknesses and was leading through different circumstances. Though they won’t tell you the standard (because they can’t), you’re still accountable to meeting it. It’s just that now part of the job is figuring out what the job is.

This transition is a classic example of “what got you here won’t get you there.” No actionable strategy will serve you all the way through your career progression. The most you can hope for is durable meta strategies or components of strategy, like how to learn. If your strategy is ineffective, then you won’t progress. If it is effective, it will only be that way for a while. Eventually you’ll outgrow it, and you’ll either stop progressing or have to define a new one. As soon as you figure out the answers, they change the questions.

In discussions with senior leaders, I’ve often heard frustration or paralysis when dealing with complex problems. They could be about whether to fire someone, whether to take a particular stance on a politically volatile issue, or which team to shrink when budgets are limited.

The way the situation is often described is one where they can’t see a clear answer, where the only options are terrible ones, where you have to draw a line but can’t prove exactly where, where every major constituency thinks a different one is the right one, or where you could be setting a dangerous precedent for the future. It’s understandable for this to be stressful. It’s not acceptable for it to be a complaint, and it’s especially not acceptable for it to block making a decision.

Making simple decisions is other people’s job. If you’re a senior leader, you should never see a simple decision. You should have people and processes that handle them before they get to you. Many of them you shouldn’t even know about. A well-run organization will only bring difficult decisions to you. Perhaps sometimes the difficulty is in the stakes, but more often it’s in the uncertainty, ambiguity, and conflict. This, by design, is the responsibility of a senior leader.

That means you have to be willing to take a leap of faith into the fog. It means you have to decide to deliberately go down an ugly, painful path. It means you have to draw an arbitrary line between two nearly identical shades of gray. It means you have to frustrate multiple constituencies by choosing something they oppose. It means you have to make the decision that’s right for right now, and in the future, make a different decision in a seemingly identical situation. A decision is better than indecision, but it’s going to feel worse and be received worse. If this was easy, you wouldn’t deserve the fancy title and the big bucks.

We overvalue what is easily observed. When assessing someone’s performance, either during a review or when considering promotion, we will bias toward information that is simple and obvious. In my experience, that means overvaluing accomplishments that are greater in number or more diverse in nature. What we then undervalue is the quality of the accomplishment. For example, we will see more value in a manager who runs a large team adequately compared to one who runs a small team well. We’re more impressed by someone who does three different things than we are by the person who does one thing three different times.

This can combine negatively with the desire for novelty. The unpleasant fact is that mastery of any complex skill is not just difficult, it’s boring. It doesn’t matter whether it’s verifying accessibility compliance, managing to a budget, or writing a code generator. For many people, a new skill is interesting at the start, but it starts getting boring once they’re decently good at it. Being really, really, really good at something means pushing far beyond the point of boredom, either via necessity, obsession, or discipline.

Having someone who is world class in an important skill can be incredibly valuable. Getting them can be hard, because sometimes the major payoff doesn’t come when they’re merely good. If you also reward dilettantism, then you’ll create a disincentive to develop that expertise. Find a way to observe and make evident when someone is amazing at something.

I have the gift and curse of an largeish, internationally distributed organization. There are members of my team in North America and Asia plus one person in Europe. To do my job well, I need to spend a lot of time with people all over the world.

It would be easy for me to schedule meetings based on what works best for me. I’m the boss, so they have to do it. Also, my time is more expensive. While I could do that, and even justify it, it’s the wrong wrong wrong approach. I need to fight the power dynamic. For the people on my team to trust me and believe I have their best interests at heart, I need to show it.

One way to do that is for me to be the one to make sacrifices rather than asking them to sacrifice (where possible). That means I meet my Asia team in my evenings and nights because that’s during their business day. It means going to them physically when there isn’t a global pandemic going on.

I have authority, I have power, and I will use both of them, but I’ll use them only when I need to. To use them to set up meetings when it’s convenient for my life is a waste, and it doesn’t demonstrate the respect and consideration that my team needs from me.

I used to believe that executive level positions were about flying at a steady 25,000 feet in the air. Now that I have one, I realize it’s actually about flying 50,000 feet in the air most of the time. But sometimes you see something down on the ground, maybe something sparkles, stops moving, or even emits smoke. When that happens, you dive, not just down to 25,000 feet, but maybe to 10,000 feet, maybe to 1000 feet, maybe even sometimes all the way to the ground. You reach for an opportunity, you solve a problem, you put out a fire, and then you take off again.

The key is not to be distant nor to be up close but to do both depending on the circumstances. The difficult part comes in knowing how to read the ground, knowing when to fly high and when to swoop down, being able to do both fly and walk adequately, and then leaving as soon as you’re done.

Engineering people often have a simple, binary view of truth. A fact is fact, regardless of who believes it, how many of them there are, or how often they state it. Either MongoDB is ACID-compliant or it’s not*. It doesn’t become more factual if a hundred prominent authorities are proclaiming it every day. This is true, and it’s also useless when it comes to working in an organization.

Organizations are made of irrational people who respond to social cues. It does matter who believes in something. It does matter how many believers there are. It does matter how often the belief is asserted. Part of this is because we are half-evolved social apes. But it also has to do with ambiguity and uncertainty. Only simple truths are cut and dried. MongoDB can be ACID compliant, depending on what you mean by it and how you’re using it. We can’t be certain we understand both the question and the answer. Also, achieving that understanding is costly. Those social cues help us resolve the ambiguity and uncertainty with less mental cost. It doesn’t get us the perfectly right answer every time, but rarely are perfectly right answers necessary.

I often hear enthusiastic co-workers say, “we should build a …” What they want to build varies, though there are recurring themes, for instance an online learning platform. Whenever this happens, I always ask two questions, which they have often not considered.

The first question is “how can we do this better than everyone else?” The idea is almost never something completely new. There’s usually someone, often multiple someones, building that product in that space already. If there isn’t something about us or our strategy that will make our product better, what’s the point? It’s easy to use market position to push a mediocre, also-ran product. That doesn’t mean it’s a good investment. Why do I want to steer someone to my lesser product instead of something superior that already exists? It’s not good for the customer, and it’s not how I want to spend my life.

The second question is “what do we stop working on to build this?” Ignoring the first question, usually the problem is not finding useful things to build. People are generally smart enough to mostly avoid things that are a complete waste of time. What they’re not as naturally good at is working on the things that are most valuable to work on. We have finite time, energy, and money, and we almost never have people sitting around doing nothing. Building one thing means not building something else. What is that something else? It’s easy to advocate for something when you don’t perceive a cost.

These aren’t fun questions to ask. In the negative case, they’re an unpleasant reality check that shuts down the idea. In the positive case, there are good answers to these questions that justify following through on the idea. Usually it’s somewhere in the middle: the idea dies, but the questions prompt thoughtful reflection and maybe even some learning. That’ll do.