Job search advice has the same context problem as any other advice. Someone got hired because of a referral they had, a market that was hiring, a specific gap in their resume that happened to match a specific gap on a hiring manager’s whiteboard that week. They tell the story afterward as a method (five things I did to land my dream job), because a method is more shareable than a shrug, and because most people, in good faith, misattribute their own outcomes to the parts of the process they controlled rather than the parts they didn’t. That’s the ordinary version of the problem, and it was the whole subject of the first post.

The job search version has an extra layer, and the extra layer is money. Losing a job is one of the more acute forms of vulnerability a person can experience in professional life. Your income is gone or going. Your identity, if you’re anything like most people in tech, is tangled up with your title in ways you don’t notice until the title disappears. You are anxious, you are scrolling more than usual, and you are, for a window of time, unusually open to anyone who sounds certain about a way out. That is precisely the emotional state an entire content industry has learned to target, and it targets it with a specific, recursive product: financial freedom, sold to you in the form of a $47 PDF that teaches you how to sell $47 PDFs.

I want to be precise about what that actually is, because the packaging is designed to obscure it. It is not a skill. It is not an asset. It is a replica of the sales funnel itself, handed to you as though it were the destination, when it is actually just another copy of the machine you already stepped into by watching the video. You buy the map, and the only landmark on the map is the gift shop where they sell the maps. It is a beautifully drawn map. It is laminated. It has testimonials. It was made by someone who has, in a meaningful number of cases, never actually been anywhere except the gift shop.

This is not a new scheme. It is one of the oldest ones there is, wearing a hoodie and calling itself a personal brand. What’s new is the delivery mechanism: an algorithm that has learned, with some precision, exactly which faces to put in front of you the week your employment status changes on a professional network. It doesn’t need to know you lost your job. It only needs to know you started looking at job postings, and searching “how to negotiate severance,” and pausing a half-second longer than usual on a video with a rented car and the word “freedom” in the caption. The targeting is the product. The PDF is just the receipt.

I said something like this in a comment a while ago, on someone else’s post about being laid off and not looking: if I have to choose between interviewing and grifting, interviewing seems a little better. I still believe that, and I want to be honest that I also went and built something in the space between those two options. I started a consulting business, not because I had cracked a formula, but because I had a specific set of skills and a specific set of people who needed them, and that combination happened to be enough to keep the bottom from totally dropping out. I am aware of how close that description sits to the thing I am criticizing. The difference, and it is the whole difference, is that nobody is buying a story from me about how I did it. They are buying the work.

That is the test, and it is the same test from the first post, just aimed at a sharper target. Ask what the person is actually selling, not what they say they’re selling. If the answer is a service, or a skill, or a thing that exists independent of whether you go on to sell it to someone else, you are looking at ordinary commerce, and ordinary commerce is fine. If the answer is a copy of the pitch you just watched, dressed up as a shortcut to the freedom the pitch itself promised, you are not looking at advice, and you are not looking at a business. You are looking at a chain letter with a landing page.

During a job search, this type of non-advice can cost you more than the forty-seven dollars.

The first is part two of “Most Advice Is a Memoir” (this one’s about job searching, and the very specific grift of being sold financial freedom in exchange for a $47 PDF that teaches you how to sell $47 PDFs).

The second is about feedback from supervisors (specifically, the one question I ask at the end of every review that turns an ambush into a conversation, and what usually shows up once I ask it).

Back to a normal schedule next week (probably).

Advice is autobiographical. Every piece of it. The person giving it to you is telling you what happened to them, what they concluded from it, and what they would do again. They are not, and cannot be, telling you what will happen to you. That gap is not a flaw in the advice. It is the nature of advice. The problem is that it gets delivered in the second person, as if it were meant for you specifically, and received in the second person, as if the recipient’s job is to apply it directly. And almost nobody says out loud: this is a story about me, filtered through my conclusions, offered to you as a suggestion. Try it on. It might not fit.

There is a difference between taking advice and listening to it. Taking advice means importing someone else’s conclusion without their context. Listening to advice means trying to understand what they learned and why they learned it, and then asking separately whether any of it applies to your situation. The first is a shortcut. The second is the actual skill. And the reason most people are better at the first than the second is that it’s much faster, and the person giving the advice usually seems pretty confident, and confidence is contagious in the specific way that other people’s certainty about your situation can temporarily feel like your own.

I have been given a lot of advice in my career. Most of it was well-intentioned. Some of it was genuinely good. A meaningful amount of it was, in retrospect, correct for the person who gave it and actively wrong for me, and I spent years trying to figure out why I couldn’t make it work before I understood that the advice had been generated from a different dataset than the one I was running on.

I am autistic. I have ADHD. I did not know either of those things for most of the time I was receiving the advice. The people giving it to me did not know either. So the advice was calibrated for a neurotypical person navigating a mostly neurotypical professional world, and I was applying it as if it had been calibrated for me, and then privately wondering what was wrong with me when it didn’t produce the expected results. “Be more visible in meetings” is solid advice if the reason you’re not visible is that you haven’t tried. It is not particularly useful advice if the reason you’re not visible is that you’re spending all of your available cognitive bandwidth processing the sensory environment and can’t simultaneously do that and perform spontaneous insight for an audience. Same advice. Completely different applicability.

This is the sharpest version of the problem, but it is not an unusual one. You do not have to be neurodivergent to receive advice that was generated from a different context than yours. The person giving it might have had resources you don’t have, or constraints you don’t have, or a network that made a particular move available to them that isn’t available to you. They might have operated in a different market, a different era, a different size of company, a different set of relationships with power. The advice might have worked for them because of something they are that you are not, and they may have no idea that’s why it worked, because most of us are not great at identifying which variables actually drove our outcomes.

Listening to advice, rather than taking it, means holding the conclusion loosely and asking about the conditions. Not in a dismissive way. Not in the way of someone who has already decided the advice doesn’t apply. More like: what was true for you when this worked? What would have to be true for me? Where does my situation match yours, and where does it diverge, and how much does that divergence matter? This is a richer conversation than the one most advice gives you access to, and it requires treating the advisor as someone whose experience is genuinely interesting rather than someone whose conclusions you are supposed to download.

The social dimension of this is complicated in a way I want to name. Advice often comes with an implicit expectation that it will be followed, and visibly not following it can feel like rejection. In mentor-mentee dynamics especially, in startup culture especially, there is a kind of performance of learning where you receive the advice, express gratitude, and then produce a future version of yourself that validates the advice. The alternative, which is receiving the advice, genuinely considering it, and then deciding it doesn’t apply to your situation, can come across as arrogance or stubbornness or ingratitude, even when it is none of those things. It is just an honest assessment of fit.

What I have tried to start doing, when advice is offered that I’m not sure applies, is saying something that is true: “That’s useful context. I want to think about how it translates to my situation.” This is not a polite brush-off. It is an accurate description of what I’m actually doing with it. I am listening. I am not automatically taking. And I have found that the advisors who are actually worth listening to tend to appreciate this, because they understand that their experience is one data point and not a prescription, and they were offering it as the former and hoping you wouldn’t mistake it for the latter.

The people who are least worth listening to are the ones who need you to follow the advice in order for the advice to matter. That is not advice. That is something else. The right trusted advisor won’t ask you to become smaller.

Not literally. (The lighting was fine.) But if you went back through my LinkedIn content from my startup years, you would find a person who was enthusiastic in a way that had no texture, supportive in a way that had no specificity, and aligned in a way that, now that I am outside of it and can see it clearly, was not alignment at all. It was performance. And I was not alone in this. The whole feed was full of it. We were all doing it, in the particular key that startup culture requires, which is somewhere between “true believer” and “very excited about our Series A.”

Here is the thing about that kind of content: nobody believes it, including the person writing it. Not fully. You cannot fully believe something you have been trained to say, any more than you can fully believe your own laugh track. The words come out right. The sentiment hits the expected marks. And yet there is something in it that even the most credulous reader clocks, at some level, as off. Not dishonest exactly. Just hollow. The kind of hollow that is the absence of something real, and that your nervous system recognizes even when your conscious mind gives it a pass.

Startup culture produces this because it has to. You are asking people to bet on something that does not fully exist yet, to commit their time and identity and professional reputation to a vision that is, by definition, not proven. The social contract of that environment includes a certain amount of performed certainty. We are all convinced. We are all here. We are all going in the same direction with the same energy. Expressing doubt, or individuality, or even a perspective that exceeds the four corners of the mission statement, can feel like breaking the spell. And breaking the spell, when the spell is what’s holding the thing together, feels dangerous.

I understand this. I genuinely do. I am not writing this to shame anyone who has posted that way, because I posted that way and I know exactly what it felt like and why.

What I want to name is what it costs. Not to the audience, though it costs them something too (their attention, and a little of their faith in authenticity). What it costs the writer. Because what you give up when you perform alignment in place of expression is not just a more interesting LinkedIn presence. You give up the thing that actually builds a career.

Your strengths do not live inside your company’s mission statement. They predate it. They will survive it. The things you are genuinely good at, the perspectives you have earned through actual experience, the problems you find interesting enough to think about at 11pm when you should be sleeping: those belong to you, not to the employer whose talking points you are currently reproducing. And the version of you that shows up in your content, either the performed-alignment version or the actual-human-who-knows-things version, is what people remember when you are not in the room.

This is not an argument for airing grievances about your employer on LinkedIn (please do not do that, for reasons that should be obvious). It is an argument for writing from the place where your knowledge actually lives, which is usually one or two levels deeper than the mission, and which does not require you to pretend the mission is perfect in order to engage with it honestly.

The people I find most worth following in any industry are not the ones who are most enthusiastic about their employer. They are the ones who are most clearly themselves: who have a perspective that you can actually disagree with, who notice things that most people in their position don’t notice, who occasionally say something that costs them something to say. Real alignment, the kind that actually means something, is not agreement. It is shared purpose pursued honestly. And you can write from that place while still being employed, still being a team player, still caring about whether the thing succeeds.

The mission statement will change. Almost every startup I have watched long enough has changed its mission statement, sometimes more than once, sometimes in ways that required quiet retrospective edits to old posts. The things you actually know and care about do not change like that. They accumulate. They compound. They are, in the long run, the only professional asset that is fully portable.

Write from there. The performed version is less interesting, less durable, and less true to whatever made you worth hiring in the first place.

The situation was not complicated. Someone had been leaving comments on my posts that were designed, with some precision, to make me look like I was out of my depth. Not hostile comments (those are actually easier to deal with, because the hostility is visible and people can see what’s happening). These were the more sophisticated kind: agreeable on the surface, subtly correcting in the subtext. The kind where if you screenshot them and send them to a friend you sound paranoid, but if you read them in context you can see the architecture of what’s being built. They were positioning me, comment by comment, as someone who meant well but didn’t quite know what they were talking about.

I knew what was happening. I’m autistic, and one of the things that comes with that, at least for me, is a very fine-grained sensitivity to patterns in social behavior. I might miss the obvious stuff that everyone else seems to catch instinctively, but I am genuinely good at noticing when a pattern of behavior has a structure to it. This had a structure.

And still I didn’t block them. For six weeks.

I wrote a few months ago about why LinkedIn makes you feel worse after you close it, and one of the things I didn’t get into in that post (because it deserved its own space) is what that platform does to your ability to protect yourself. The same mechanism that makes the scroll depressing also makes the block button feel dangerous. LinkedIn has convinced a meaningful portion of its professional user base that their reputation is a live, fragile thing that can be damaged by any visible conflict, and that the block button is a kind of conflict.

So there I was, a grown adult with two decades of professional experience, doing the math on whether blocking someone who was quietly working to undermine my credibility would hurt me more than letting them continue.

The math I was doing went something like this: this person has a reasonably large following. We have mutual connections. If I block them, they might notice. They might post about it (people do this). Mutual connections might see. I might come across as petty, or thin-skinned, or unable to handle professional disagreement. My “thought leadership” (a phrase I will never stop finding mildly embarrassing) might take a hit. My engagement numbers might dip if the algorithm interprets a block as a signal of some kind.

I am describing this in the tone of something absurd because, in retrospect, it is absurd. But I want to be clear that in the moment it felt like genuine risk management. That is how completely LinkedIn has colonized the part of my brain that is supposed to handle self-preservation.

Here is what is actually true about the block button: it is a door with a lock. That is all it is. The person on the other side is not entitled to access to you, your work, or your comment sections. There is no professional obligation, no social contract, no industry norm that requires you to leave your door unlocked for someone who is using their access to cause you harm. The idea that exercising that lock is a reputation risk is something LinkedIn’s engagement-optimization machinery has implanted in your head, and it benefits from you believing it in the same way a landlord benefits from tenants who don’t know their rights.

The thing I kept worrying about, the social consequences of being visibly protective of myself, was not actually a risk in any meaningful sense. Nobody with a healthy relationship to professional self-respect looks at someone who has blocked a bad-faith actor and thinks less of them. The people who perform outrage at being blocked are overwhelmingly the people who were doing something that warranted the block. And the people watching, the mutual connections I was so anxious about, are mostly not watching at all. They have their own feeds, their own anxieties, their own six-week decisions they haven’t made yet.

What I lost in those six weeks was not nothing. The comments continued. A few people who didn’t know the context probably absorbed some of the framing being constructed. I spent cognitive energy I didn’t have to spare on managing my own response in a space where I should have been able to show up without that overhead. And I modeled, for myself, a pattern of tolerating something harmful because I was afraid of how it would look to stop tolerating it.

That last part is the one I keep coming back to. Because the fear of reputation damage, in this context, was not really about my professional reputation. It was about being seen as someone who couldn’t take it. Someone who needed protection. Someone who was, in some hard-to-articulate way, not tough enough for the game.

Which is its own kind of LinkedIn logic, and its own kind of trap.

The block button is not a career decision. It is a boundary. Those are not the same thing, and the fact that a professional networking platform has spent years blurring that distinction is one of the more successful pieces of psychological manipulation in recent tech history.

I blocked them. Nothing happened. My follower count did not move. No mutual connection sent a concerned DM. The algorithm did not punish me. The person did not post about it, at least not anywhere I could see.

What did happen is that I stopped spending six weeks doing math that shouldn’t have needed to be done.

That is the whole lesson. I wish I had a more interesting one.

There is a name for what I’m describing. The nirvana fallacy, sometimes called the perfect solution fallacy, is the habit of measuring a real option against an ideal one that doesn’t exist and concluding that because the real option fails the comparison, you’re better off doing nothing. The name comes from the Buddhist concept of nirvana, the unattainable perfect state, which is a little ironic given that the whole point of Buddhism is to be okay with impermanence. But the fallacy has taken the word and run with it, and it is running through your industry right now at a pace that would impress you if it weren’t costing you so much.

Here is what it sounds like in practice. “We can’t implement this feedback process until we can guarantee anonymity.” “We can’t hire until we’ve built out the onboarding program completely.” “We can’t address the team dynamics issue until we understand the root cause.” “We can’t launch the empathy initiative until leadership is fully aligned.” What all of these sentences have in common is that they are using a real concern (incomplete anonymity, incomplete onboarding, incomplete understanding, incomplete alignment) as cover for something that is not a concern at all: the imaginary version of the thing, which would be perfect, and which will never arrive.

I spent enough years in infrastructure to know what this looks like when the stakes are concrete. A monitoring system that catches seventy percent of incidents is not a good monitoring system in the ideal sense. But it is infinitely better than the perfect monitoring system you are still designing. A p95 SLO that you actually measure beats a p99.9 SLO that lives in a slide deck by any metric that matters. The engineers who understand this viscerally, who have been woken up at 3am enough times to have a healthy relationship with imperfect observability, are not confused about whether to ship the imperfect thing. They ship it. They improve it. They do this because they have learned, through direct experience, that the absence of imperfect coverage is not the same as perfect coverage. It is the same as no coverage at all.

The fallacy gets harder to see when you move from infrastructure into people. And I think this is worth sitting with, because the domain shift is not incidental. With systems, you can measure the gap between what you have and what you want. With people, with culture, with the soft and unglamorous work of building an environment where people can actually do their best work, the gap is harder to quantify, which means the perfect imaginary solution is easier to construct and harder to argue against. You can always conjure a more complete version of the thing. You can always find a reason why this particular feedback mechanism isn’t quite right yet, why this particular conversation is better had after the next planning cycle, why the team health check feels premature while the roadmap is still unsettled.

What you are actually doing, in those moments, is comparing the cost of the imperfect action against the cost of nothing. And you are miscounting. The cost of nothing is not zero. The cost of nothing is everything that accumulates while you wait: the engineer who decides you don’t actually care because nobody asked, the team dynamic that calcifies because there was never a structured moment to name it, the trust that doesn’t get built because the conditions for trust never got created. The nirvana fallacy hides these costs because they don’t appear on any ledger. The imperfect solution you didn’t implement doesn’t generate a line item. It just generates a culture where people have learned not to expect much.

There is also a version of this problem where you are doing the imperfect thing, you have done the imperfect thing, and someone accuses you of not doing enough anyway. This happens more than people admit, and I want to address it directly because the nirvana fallacy does not only freeze leaders from the inside. It gets aimed at them from the outside too, and knowing how to hold your ground in that conversation is a different skill from knowing how to start the right project.

The single most useful question I know in that situation is: compared to what? Not as a deflection. As a genuine request for the comparison being made. When someone says “this isn’t enough,” there is always an imaginary “enough” hiding inside that sentence, and it is worth pulling it into the light, because about half the time it turns out to be incoherent, and the other half it is a real thing you can actually respond to. “Compared to what?” does not mean “what I’m doing is fine.” It means “show me the standard so we can evaluate it together, instead of against a feeling.”

The second thing that helps is making the imperfect action visible before the criticism arrives. This is partly about communication and partly about framing. If you implement an eighty-percent solution without naming it as such, the gap between it and one hundred percent looks like negligence. If you implement it while saying “here is what this does, here is what it doesn’t do yet, and here is when we revisit it,” the gap looks like a plan. The difference between those two things is not the action. It is whether the action has been given enough context to be evaluated honestly rather than just compared to an imaginary alternative.

The third thing, and the hardest, is to stop accepting the premise that imperfect action equals insufficient effort. This is the frame the nirvana fallacy needs to survive, and you do not have to grant it. Doing the sixty-percent version of the right thing is not evidence that you don’t care. It is often evidence that you are the only person in the room who is actually paying attention to what’s achievable. The people demanding the perfect version are not wrong to want it. But they are not the ones who have to answer to the ones on the outside watching you build it, which is a relevant asymmetry that sometimes needs to be named out loud.

I want to be clear that I am not arguing for thoughtless action. The answer to “this isn’t perfect” is not always “ship it anyway.” There are genuinely bad plans that should be improved before they go anywhere near a team. But there is a difference between a plan with real structural problems and a plan with the problem of being imperfect in a world where nothing is. Most of what I see stuck in organizational purgatory is the second kind. It is feedback mechanisms that would work at eighty percent. It is check-in cadences that aren’t perfectly designed but would be better than nothing. It is conversations that need to happen before someone is ready to have them flawlessly.

The one thing I keep coming back to, the thing that took me embarrassingly long to stop arguing against, is this: you are not choosing between the imperfect thing and the perfect thing. You are choosing between the imperfect thing and what exists right now. That reframe does not make every imperfect solution worth implementing. But it does make the question honest. And honest questions are the only ones that lead anywhere useful.

The perfect version of this post would have a cleaner ending.

This is the ending it has.

I happened to know someone else in that exact space, a man in my network who was becoming a recognized leader in virtual assistant services. So I did the networking thing. I recommended they meet. Collaborate, maybe. Share ideas, at minimum. And then I heard myself say it: “There’s a lot he could show you.”

I caught it the moment it left my mouth, the way you catch a bad config change the instant after you hit enter. What I meant was “I want to see you succeed.” What I said was “here is a man who can teach you about the business you already built.” She had the successful company. She had the track record. I had positioned her, reflexively and without a moment of conscious thought, as the student.

I have spent more than twenty years in technology, most of it in leadership roles of one kind or another, and I want to be precise about what that sentence was. It was not malice. It was not even a belief I hold, in the sense of something I would defend if you asked me directly. It was a default. It was the factory setting talking, the firmware I was shipped with, doing what defaults do, which is execute silently whenever nothing overrides them. I have been working on overriding them for years. The sentence was a status report on that work: incomplete.

This is the part of the conversation about privilege that I think the technology industry, my industry, is actually equipped to understand, because we already have the vocabulary. We know the difference between a one-time install and a service. Boxed software was something you bought once, ran until it broke, and felt no further obligation toward. The whole industry moved away from that model (to SaaS, to subscriptions, to continuous deployment) because we learned that anything worth running is worth maintaining, and that maintenance is not an event. It is a posture. It has uptime requirements. It gets patched on a schedule, not when the mood strikes.

Most of what passes for allyship in leadership circles is boxed software. The one diversity panel. The single mentee. The statement issued the week everyone else issued one. You install it, you point to it, and you consider the matter handled, the way we used to consider a server handled because someone racked it in 2009. And like that server, it is quietly accumulating vulnerabilities the whole time you are not looking at it.

What I have started calling Stepping-Aside-As-A-Service is my attempt to take the other model seriously. The premise is simple. If you are someone like me (white, male, a couple of decades into a career where the doors opened so smoothly I rarely registered that they were doors), then the most useful thing you can do with your position is not to occupy it harder. It is to treat the redistribution of access as a standing service you run, with the same expectations you would put on any production system. Always on. Actively maintained. Monitored for regressions.

In practice it looks unglamorous, which is how you know it might be real. It is recommending the woman as the expert, not the recipient of expertise. It is hearing about a conference slot or a podcast invitation and asking, before you say yes, who would be better than you, and meaning the question. It is saying her idea was the load-bearing one in the meeting where the credit is being distributed, specifically in the meeting, where the record gets written, rather than privately afterward where it costs nothing. It is making the introduction and then getting out of its way, because an introduction you stay in the middle of is not a bridge. It is a tollbooth.

And it is the monitoring, which is the part my networking story is really about. Because the failure mode of any long-running service is not usually the dramatic outage. It is the silent regression. The thing that used to work, that you believed still worked, drifting out of spec while the dashboard stayed green. I believed my deprogramming was further along than it was. I had read the things, examined the assumptions, done the deliberate work. And then the default fired anyway, in a half-second of small talk, in a sentence I did not plan, which is exactly where defaults live. You do not find out what your system actually does under load by reading its documentation. You find out in production.

What I did next was the only part of the story I am at all proud of. I patched it live. I went back to the sentence and reframed the introduction the way I had actually meant it: two successful operators in the same space who might sharpen each other, with as much for him to learn from her as the reverse (and, given whose business was further along in the ways that counted, possibly more). A small correction. She may not have even registered the original sentence; the people on the receiving end of these defaults have usually heard them so many times that one more barely produces a signal. But I registered it. The patch was for the system that produced the bug, not just for the output.

That, ultimately, is why the “as-a-service” framing matters to me, and why I am suspicious of any version of this work that comes with a completion date. A service does not graduate. There is no release where you declare the maintenance finished and walk away; the moment you do, the drift begins. The defaults I was shipped with were installed over decades, by a culture that was very thorough about it, and they do not uninstall because I have decided I am one of the good ones. Deciding I am one of the good ones is, in fact, the most reliable way to stop checking the logs.

So I keep the service running. I step aside, on purpose, on a schedule, and I watch for the regressions, and when one fires in the middle of a perfectly pleasant conversation over networking event coffee, I treat it the way I would treat any incident: fix it now, figure out the root cause later, and do not waste time pretending the system is something other than what the logs say it is.

There was a lot she could have shown me. That was the sentence. Next time, I intend to say it on the first try.

Picture a leader who is, by any reasonable measure, the real thing. Not the LinkedIn version of empathic (which is to say: someone who posts about vulnerability and gives TED-style talks about psychological safety while running a culture where people are afraid to have a bad quarter). I mean the actual version: someone who listens without performing listening, who can tell the difference between someone being difficult and someone being in pain, who builds teams where people feel safe enough to say true things. Someone who has done the internal work and whose care for the people around them is not a strategy. It is just who they are.

This person is real. I have met them. I have worked for some version of them. And I am here to tell you that around this person, if you are not careful, something very strange can happen.

The team starts routing everything through them. Not because the leader demands it (they don’t, and would be uncomfortable if you pointed it out). But because the leader is so reliably good at holding the complexity of a situation, so genuinely trustworthy in a way that most professional relationships are not, that people stop trusting their own read as much as they trust the leader’s. Disagreement starts to feel disloyal, not because anyone said it was, but because the culture has started treating alignment with this person as a proxy for good judgment. The team’s collective ability to reason about hard things begins to run through a single node.

This is a cult of personality. It does not require a charismatic egomaniac at the center of it. It does not require anyone to have bad intentions. It can form entirely around someone who is trying to do the right thing, and who would be mortified to know what had been built in their name.

I think about this a lot in the context of what I’m writing about in this book, because the thing I keep coming back to is that empathy is a capacity, not a personality trait. The goal of empathic leadership is not to be someone your team can always turn to. It is to build a team that does not need you in that way. Those are very different projects, and they require different things from you.

The version that builds dependency is not always obvious from the inside, because it feels like trust. When your team looks to you for guidance on hard judgment calls, it can feel like evidence that you have created something safe. And you have. But safety that is contingent on a specific person’s presence is not the same as safety that is woven into how a team operates. One of those is a gift. The other is a dependency. They can look identical from the outside, and sometimes from the inside, for a long time.

What tends to reveal the difference is a transition. The beloved empathic leader takes a new role, or leaves, or is on leave for a month, and the team either holds its shape or it doesn’t. The teams that hold their shape are the ones where the leader spent years doing something that felt, in the moment, a lot less rewarding than being turned to: they pushed decisions back. They said “what do you think?” when they had a perfectly good answer of their own. They deliberately made themselves less necessary, because they understood that the point was not to be a good leader forever but to grow people who could do things they couldn’t have done before.

The teams that don’t hold their shape, the ones where everything quietly stalls or fractures or just gets visibly worse the moment the leader is not in the room, reveal something harder to sit with: the leader had become the team’s capacity, rather than building it.

I want to pause here and acknowledge the obvious irony in what I’m about to say.

There is a clip that circulates online of Steve Jobs answering a question about the most important thing he learned at Apple. Jobs is, by most accounts, the canonical example of a tech cult of personality. His name is on a religion. People still argue about whether he was a genius or a monster, as if those were mutually exclusive, and that argument is itself a kind of devotion. He is not the person you expect to quote on this subject.

And yet what he says is this: when he sees something not being done correctly, his immediate instinct is to fix it himself. But over time, he learned to fight that instinct. Because his job, he realized, was not to fix the thing. His job was to build a team that could do great things over the next decade, not just the next year. Which meant that when someone struggled, his question had to shift from “how do I solve this?” to “how do I help this person learn from it?” He admits it is painful. He says he still fights the instinct every time.

I find this genuinely interesting, and not because it makes Jobs a hero of humble leadership (it doesn’t, and there is substantial evidence to the contrary). I find it interesting because even someone who spent decades being the center of gravity in a room full of exceptionally capable people eventually understood, at least intellectually, that the center of gravity was the problem. That the instinct to fix, to step in, to be the one who makes the call, is exactly the instinct that prevents a team from becoming something that can survive you.

There is also something worth naming about what happens to the leader in this dynamic, because I don’t think we talk about it honestly enough. Even someone who genuinely does not want to build a cult can start to find comfort in being depended on. The team’s trust feels good. Being the person everyone turns to feels good. Being the person without whom the hard things can’t get resolved is a particular kind of significance, and significance is not nothing. The risk is not that the good leader becomes corrupt exactly. It is that they stop noticing when their presence has become structural, when the thing they have built requires them specifically to function, and when quietly withdrawing from that center might be the most empathic thing they could do.

I do not think this is a simple problem, and I am not going to hand you a simple solution.

What I will say is that the question “does this team work without me?” is one of the most important questions an empathic leader can sit with, and that the answer you arrive at through wishful thinking is not the same as the answer you would get from a real test. The cult does not announce itself. It accumulates, quietly, in the gap between how much people trust you and how much they have learned to trust themselves.

The kindest thing a good leader can do, sometimes, is be less available. Not absent. Not cold. Not performatively hands-off as a management philosophy. But genuinely willing to let the team carry more than is comfortable, to let them make calls you could have made faster, to let them sit with uncertainty you could have resolved, because the point was never to resolve the uncertainty. The point was to help them get better at living inside it.

That is not a cult. That is a team.

Read more

Read more

But the boardrooms and the “huddle spaces” of the startup world …

Read more

We talk a lot about …

Read more

Read more

Read more

Before the era of sleek smartphone inte…

Read more

Read more

Read more

Read more

Read more

Read more