Every leader I know keeps a list: two or three names, usually, and it never gets shorter. Yours probably includes the partner who walked out with the client roster, the hire who interviewed like a saint and billed like a pirate, or the number two who negotiated his next job on the company phone. You don't write the list down, but you know it cold, and if I asked you about it over dinner you could give me the year, the dollar amount, and the exact moment you realized what was happening.
I have my own list, and I still know every name on it. What I want to talk about is what the list does to you afterward, because I've watched it happen to better leaders than me and I've caught it happening in myself. The burn heals. The lesson you take from it doesn't, and the lesson is almost always the wrong one.
The wrong lesson goes like this: I trusted too fast, so from now on people will earn it. Trust becomes a thing you ration, so new hires start at zero and climb. Every delegation gets a check-in, every check-in gets a follow-up, and somewhere around the fourth signature on the approval form you have built an organization that runs on suspicion and calls it prudence. There's nothing wrong with a season of caution after a betrayal. There's a great deal wrong with caution that never ends, because it stops being a response to one person and becomes the operating system for everyone. It feels like wisdom, but it is simply the last person who burned you, still running your company from wherever he ended up.
The Bill Comes Due Everywhere Else
Here's the part that took me too long to see: the list is short, and the people paying for it are many. Two names, maybe three, over a career, set against hundreds of people who showed up, did the work, told you the truth when it cost them something, and never once gave you a reason to check behind them. When you ration trust because of the two, the hundreds are the ones who feel it. They don't know about your list, only that the boss reads their emails twice.
And they learn from it, because trust is reciprocal in a way that most leadership advice skips right past. A team member who is handed distrust by default doesn't wait patiently to graduate out of it. She gives it back. She assumes your questions have motives and reads a scheduling change as a message. Small misunderstandings that a trusting relationship would absorb in a hallway conversation turn into three-week grudges, because nobody on either side is willing to assume good faith first. Leaders who trust slowly teach everyone around them to trust slowly, and then wonder why the building feels cold.
At B:Side, when I do my rounds of all-staff one-on-ones, the thing people describe wanting, once you listen between the lines, is predictability from the top. They want to know that the way I decided something yesterday is the way I'll decide it tomorrow. That is them working out whether they can trust me. At the same time, I'm working out whether I can trust them, and whoever moves first sets the terms for both. At ASU, I put students into teams in the opening weeks of the semester with people they met days before, and I've found the same math plays out in miniature. The teams that assume the best of each other move. The teams that wait for proof spend the term auditing one another's contributions. Nobody on those teams has been burned yet, by the way. They've simply been raised in a culture that treats suspicion as sophistication.
That culture is why this matters more now than it did ten years ago. We're living through a stretch in which trust in nearly every institution is thin: banks, universities, the press, the government, each other. Meanwhile, the teams themselves are getting smaller, and I wrote in February about why: AI is taking over the work junior hires used to do. One consequence of a leaner org chart is that every person on it carries more weight. A leader who has to verify everything a team of eight does will run out of hours before he runs out of things to verify. Trust has always been the cheapest coordination mechanism there is. In a small shop under pressure, it's the only one that scales.
Seneca's Order of Operations
Seneca wrote to his friend Lucilius about this two thousand years ago, and his third letter is the cleanest thing I've read on the subject. Lucilius had sent a letter by way of a mutual friend and then, in the same breath, warned Seneca not to tell that friend everything. Seneca called him on it. If you'd trust a man to carry your mail but you wouldn't trust him with your mind, he said in effect, then you're using the word "friend" the way people use it for everyone they've ever been introduced to.
The heart of the letter is a sequence, and the sequence is the whole point. Deliberate for a long time before you admit someone into your confidence. Then, once you've decided, take him in with your whole heart and speak to him as freely as you'd speak to yourself. Seneca's complaint was about people who run it backwards: they make a friend first and judge him afterward, which is exactly how you end up with a list of names. Judge slowly. Then trust completely. Both halves are instructions, and the second one is the half we skip.
He also saw the mechanism I described above, and he described it better. Some people, he wrote, fearing to be deceived, have taught others to deceive; by their suspicion they hand the other person the right to do wrong. Regard someone as loyal and you tend to make him loyal. Regard him as a suspect and you've told him what you expect. Suspicion works as a curriculum, and the students learn fast.
Seneca also named the failure on either side of the line. There's the man who tells everything to everyone he meets, and there's the man who tells nothing to his closest friends. Both are faults, he said, and then he made a distinction I keep coming back to: he called the first the more honest of the two mistakes, and the second the safer. Trusting everyone is a failure of judgment, while trusting no one is a failure of nerve that happens to look like experience.
There's a bitter footnote to the letter, and I don't think you can read Seneca honestly without it. He wrote these letters in his last years, after more than a decade at the elbow of an emperor who trusted no one, and not long after finishing them Nero ordered him to open his veins. Seneca knew what a court without trust looked like from the inside. He described the cure precisely and lived inside the disease. I've written before about the gap between what Seneca preached and what he practiced, and I don't think the gap disqualifies the advice. It tells you what the advice costs when the man at the top refuses it. A court where the ruler trusts nobody ends up with nobody worth trusting. The only people who stay are the ones who've learned to survive suspicion, and those are precisely the people you should never have let in.
Marshall Handed Over the War
Seneca gave the principle, and George Marshall ran it at scale, which is why I find him the more useful teacher: he had a payroll. I wrote a week's worth of essays about Marshall two years ago, so I'll keep this to the one habit that matters here.
Marshall never commanded troops in battle, so his whole war, from the fall of 1939 to the fall of 1945, was a sequence of decisions about which men to trust with things he couldn't do himself. He did the judging half of Seneca's sequence with a ruthlessness that still makes people uncomfortable. Early in his tenure he stood up a board, with his own predecessor running it, to clear out the senior officers who couldn't keep pace with the war that was coming. Several hundred of them went, and the only criterion that counted was how a man was performing today. The legend says he kept a little black book of the officers he'd been watching since his years at Fort Benning. The book, it turns out, never existed (he kept the list in his head, which is worse for the officers and better for the story).
What he did once the judging was done is the part worth studying. In December 1941, a week after Pearl Harbor, he called a brigadier general he barely knew into his office. Dwight Eisenhower had spent most of the previous twenty years as a staff officer and had never commanded in combat. Marshall laid out the situation in the Pacific, asked him what the country's general line of action ought to be, and waited. Eisenhower asked for a few hours and came back with a recommendation: do everything possible to support the Philippines while accepting that they probably couldn't be saved, and build Australia into the base for whatever came next. Marshall agreed and told him to do his best. That was the whole handoff.
Around the same time, Marshall told him something Eisenhower put in his memoir years later. The War Department, Marshall said in substance, was full of capable men who analyzed their problems well and then felt compelled to carry them to his desk for the final decision. He needed assistants who would solve their own problems and tell him afterward what they'd done. Six months later he sent Eisenhower to London to command every American soldier in Europe. Eisenhower had been a lieutenant colonel eighteen months before.
The judging was slow, years of it, conducted from memory across peacetime posts. The trust, once extended, was nearly total, and Marshall didn't fly to London to look over Eisenhower's shoulder. When Roosevelt later tried to hand Marshall the command of the invasion of France, the job every soldier of his generation wanted, Marshall refused to ask for it. The President, he said, should decide on the country's interest rather than on his. Then he wrote out the cable announcing Eisenhower's appointment by hand and sent the draft to Ike as a keepsake. You cannot build anything that outlives you without handing the important things to people who might drop them. Marshall knew that, and he handed over the war.
The Math of Suspicion
The first cost of withheld trust is friction that never gets billed to the right account. Every layer of verification you add because of a name on your list costs hours from people who aren't on it. Think of the report that gets reviewed twice, the decision that waits for your return from a trip, and the contract a vice president could sign but sends up anyway. She's learned that signing it herself earns a question and sending it up earns nothing. Multiply that across a year and you've paid for your suspicion several times over in salary, and you've paid a second time in the people who left because they were tired of being audited.
Lending gave me a sharper way to see this, because a loan is trust with paperwork attached. Before B:Side puts money out the door, we do a great deal of judging: cash flow, collateral, the character of the owner, the history of the business. That's the slow half, and it should be slow. However, once the loan closes, we don't call the borrower every Tuesday to ask whether she's still in business. We ask for financials on a schedule, we watch the payments, and we let her run her company. The verification is proportional to the stakes and it's attached to the process, never to a running suspicion of the person. A lender who couldn't extend trust after underwriting would be a lender with no loans.
I'll grant the strongest objection here, because I've made it myself. Some of us work in businesses where a single person acting in bad faith can do real and irreversible damage, and "trust first" sounds like an invitation. That's true, and it's why Seneca put judgment before trust rather than instead of it. It's also why controls belong on transactions and never on individuals. Two signatures on a wire transfer protect everyone in the building, including the person signing; it's a control on a step where a mistake can't be undone. A second set of eyes on a person you chose says something different. It says you didn't really choose. Keep the controls where the consequences are irreversible, and take them off everywhere else.
The second cost is quieter, and it lands on people who did nothing to deserve it. My students haven't been burned yet, and they still walk into their first jobs braced for a manager who checks everything, because that's what they've been told professionalism looks like. The ones who land with a leader who hands them a real problem in the first month light up in a way I can see from across a room. The ones who spend their first year being verified learn to do exactly what's asked and nothing more. A person's worth is settled long before you finish deciding about them, but their growth waits on the decision. Nobody becomes capable inside a relationship where capability is presumed absent until proven otherwise, which means the guarded leader is doing more than protecting himself: he's stunting the people he's responsible for.
Start With the People You Didn't Choose
The hardest version of this problem arrives when you inherit a team. I took over B:Side in September 2020, in the middle of the pandemic, with a staff I hadn't hired, in a year when meeting anyone in a room was complicated. Seneca's sequence assumes you get to do the judging first. I didn't. I could withhold trust from everyone until I'd formed my own opinion, which would have taken a year I didn't have. Or I could extend it on the strength of my predecessor's judgment and correct as I went. Only one of those keeps the doors open. I took it, I got some of it wrong, and it was still the right call. After all, the organization ran while I learned it. If I'd waited to trust until I was certain, the waiting would have been the message, and the message would have been received.
If you've been burned recently, the temptation is to treat the wound as data. It is data about one person, and the mistake is letting it write policy for everyone else. The betrayer is the exception, and a great deal of leadership comes down to refusing to let the exception govern the rule. Everyone else on your team is counting on you to remember that they aren't him.
If you're early in your career, understand that the feeling of being trusted is one of the great rewards of working life, and that the fastest route to it is unglamorous. Make small promises and keep them with precision: tell someone you'll have it by Thursday at ten, and have it by Thursday at ten. People who can't be counted on for the small things are assumed, correctly, to be unreliable on the large ones. People who nail the small things get handed the large ones sooner than they expect. Consistency is the only currency that buys it, and it compounds.
Make the First Move
Trust is something you do before it's something you feel, and the leader does it first.
Finish the judging, then stop. Do the reference calls, watch how a person behaves when the deal is going sideways, ask the questions you're afraid to ask, and then decide. A hire you keep re-evaluating is a hire you never made.
Say the trust out loud. Tell the person what you're handing them and that you won't be behind them checking. "This is yours; tell me what you did" changes how someone works in a way a job description never will. Marshall said it to Eisenhower in nearly those words, and it's the sentence Eisenhower remembered.
Multiply the reps. Trust grows on repeated contact, so in a new relationship, compress the calendar. More conversations, more decisions made together, more problems worked through side by side in the first ninety days than the normal cadence would produce in a year. Density substitutes for time.
Put controls on transactions, never on people. Dual approval on a wire, a second review on a loan above a threshold, a checklist before anything irreversible. Those protect the person doing the work, while a standing second opinion on someone you've already chosen does the opposite.
Keep the other list. You already know the names of the ones who burned you. Count the others: the people who handled something hard without telling you about it, who returned a favor you'd forgotten you did, who stayed through a year that gave them every reason to leave. That list is longer, and it's the one that should set your defaults.
Deal with the exception fully and fast. When someone does break the trust, act. Confront it directly, end it if it has to end, and don't drag it out because you'd rather not be the one to say it. Speed here is a kindness to everyone else, because a betrayal that lingers unaddressed teaches the team that trust carries no consequences in either direction.
The List Stays Short
I want to be honest about what this doesn't fix. Will you get burned again? Almost certainly. If you lead people long enough and trust them the way I'm describing, the math guarantees it, and I don't have a version of this argument that makes the next one hurt less. What I have is the comparison: two or three names over a career, against the hundreds who deserved a leader who assumed the best of them and got one. The guarded leader keeps his list short by trusting no one, and he pays for it every day in everything the untrusted never bring him. The other kind keeps his list short because most people, judged carefully and trusted fully, turn out to be worth it.
Seneca wrote the sequence down for a friend, and Marshall ran it on a war. The rest is a decision you make every morning about the people who walk in the door and haven't done anything wrong.
Judge slowly. Then hand over the keys.
My latest book, Honor Under Pressure, is out now. It's about making the hard call before the pressure makes it for you. You can find it on Amazon.