6 min read

You did not inherit a team. You inherited a turnover number, and the difference is going to define your first ninety days more than any onboarding plan, any 1-on-1 cadence, any team offsite on the calendar.

The people still in their chairs aren’t a stable group with a few unfortunate gaps. They’re a self-selected sample of people who had reasons to stay through whatever happened. Some of those reasons are good (they like the work, they have outside options that haven’t moved, they’re close to a vesting cliff). Some are bad (no other options, exhaustion, already mentally gone). You don’t know which is which yet, and the cost of finding out the slow way is the next round of resignations.

Most advice on inheriting a team with high turnover starts with “do a listening tour.” Michael Watkins covered the assess-reshape-accelerate version of this in his classic HBR piece on inheriting teams, and it’s correct as far as it goes. It’s also incomplete for the high-turnover case, because it assumes the people listening to you are evaluating you neutrally. They aren’t.

The test the survivors are running

Inside the first thirty to sixty days, the people still on this team are doing one thing in every interaction with you. They’re checking whether you’re a copy of the conditions that drove their colleagues out.

It isn’t conscious most of the time. They’re pattern-matching your behaviors against a list of red flags they built up over the last eighteen months. The moment a few of those flags come up, the calculation flips quietly from stay-and-see to start-looking. Nobody comes into your 1-on-1 and says “I noticed you do the thing the last manager did, I’m updating my resume tonight.” They smile, they nod, they tell you the team is great. They start applying. You find out forty-five days later when the next resignation lands and you’re starting at a worse number than the one you walked into.

This is the Survivor Read. You’re inside the window from your first day until somewhere between day 30 and day 60. Treat every meeting, every Slack reply, every public decision as data the team is using to decide whether the dynamic that just blew up is back, wearing your face.

What the previous manager probably did

You almost certainly won’t be told the real story for at least three months. The survivors have learned that telling the new boss the truth about the old one is a risk: maybe you’re friends with that person, maybe HR will hear about it, maybe you’ll read it as them being difficult. So they’ll give you the sanitized version. “Things were busy.” “We had some attrition.” “It was a hard year.”

You can work without the real story. You build a working theory from three sources, none of which require anyone to inform on the previous manager directly.

The first source is what people bring up unprompted in your first round of 1-on-1s. Not what they answer when you ask “how is the team doing.” What they volunteer when you ask “what’s your week looking like.” If three out of four people unprompted name the same process or the same meeting or the same reporting requirement as a friction point, that’s your highest-probability source of the previous failure mode. The signal is in the unprompted mention, not the response to the direct question.

The second source is what’s missing from the team’s documentation. A team that lost five people in a year usually has gaps where roles, decisions, or histories used to live in someone’s head. The gaps tell you what the previous manager didn’t invest in (training, documentation, cross-coverage) and what kind of fragility this team is currently running on.

The third source is exit interviews, if you can get access. HR will sometimes share themes if not specifics. If your skip-level is willing to share what they heard from the leavers, ask. Patterns from the people who voted with their feet are the closest thing you’ll get to ground truth without anyone currently on the team having to risk anything by telling you.

You’re not building this theory to confront the previous manager or to relitigate what happened. You’re building it so that when you make your first visible decisions, you can make them on a real diagnosis instead of a generic “let’s get to know each other” play that the survivors have already seen run before.

Four moves that make the survivors stay

Once you have a working theory, the survivors need evidence in the first thirty days that something specific has changed. Vague reassurance (“I’m a different kind of manager”) makes it worse, not better. They’ve heard versions of that before. What works is demonstrated movement on a small number of things they actually care about.

Name what you inherited, in a team meeting, within your first two weeks. Not in a way that throws the previous person under a bus. In a way that proves you see what the survivors see. Something close to: “I know this team has lost a lot of people in the last year. I’m not going to pretend that didn’t happen. I’m going to spend the next month figuring out what I think the right calls are, and I’m going to check those calls with you before I make them.” That sentence does three things at once. It validates the experience as real. It signals you won’t fake-fix it in week one. And it puts you on the hook publicly for consulting the team, which they will hold you to.

Then ship a visible fix on one specific friction point everybody mentions. Not a strategic change. A thing. A standing meeting that everybody hates. A reporting cadence that takes three hours every Friday and nobody reads. A tool the team needs and somebody upstream has been blocking the purchase. Whatever the unprompted complaint is in your first round of 1-on-1s, fix it in week three or week four and tell the team why you fixed it. The survivors aren’t asking you to be a savior. They’re asking you to demonstrate that the small frustrations they were screaming about will land somewhere when they tell you about them.

The harder discipline is holding every structural decision for at least sixty days. No reorg. No role changes. No hires for the open seats. Not because you’re paralyzed, but because every structural change you make in your first sixty days reads to the survivors as the same playbook the previous manager ran, just with different boxes on the org chart. The cost of waiting two months on a hire is small. The cost of hiring the wrong person into a role you don’t yet understand is enormous, because the next departure after that hire is a resignation the team will read as your responsibility, not the previous manager’s.

By day forty-five, write down what kind of person and what conditions would have kept the last three leavers. Not the people themselves. The conditions. If the leaver was a senior engineer who left for a smaller team, your conditions list might be “more autonomy on architectural choices, less context-switching across products.” If the leaver was a project manager who left for a remote-first company, the list might be “fewer in-office days, asynchronous-first decisions.” This document is the basis for every backfill hire and every retention conversation you’ll have for the next year. Without it you’re guessing. With it, you have a written diagnosis you can test, share, and refine.

What you don’t do

Promising the survivors that everything will be different is the obvious move and it backfires. They won’t believe you, and the moment any old pattern recurs even slightly, they’ll read it as proof you lied.

Bringing in your old team’s process from your last role and imposing it on this one is the second obvious move, and it does the same thing in reverse. The survivors are pattern-matching for any sign of a top-down rewrite, because the previous manager almost certainly did one and it’s part of what made people leave.

Calling the people who left to ask why feels diligent. It also gets back to the survivors as “the new manager is talking to the people who quit,” which reads as either nostalgic for a team you don’t have or hunting for ammunition against the people who stayed. Whatever you wanted to learn from those calls, you can learn from the documentation gaps and the unprompted mentions instead.

And do not start the conversation about retention bonuses, equity refreshes, or other financial retention plays in your first month, even if the survivors are clearly at flight risk. Money this early reads as a panic move. It also commits the org to a precedent before you have the diagnosis to defend it. Make the case for retention spend in week eight or nine, when you can defend it with a written theory of what broke.

The window closes

Around day sixty, the Survivor Read ends. Not because the team stops watching you (they’ll keep watching for the rest of your tenure) but because the acute evaluation period closes. By then, the people who decided you’re a different signal will start to relax, contribute more, share more, and the team will start to feel like a team again. The people who decided you’re the same signal will be in the late stages of an external job search, and the next two or three departures will land in months three and four.

You don’t get a way to extend the window. You get one shot at it, and the four moves above are the highest-yield ways I’ve seen people use the time. Do them, and the team you have at day ninety will be smaller than you wish it were, but stable. Skip them, and the team you have at day ninety will be smaller than the one you inherited, and the next round of departures will land on your record rather than the previous person’s. The clock on whose-fault-this-is starts the day you sign the offer letter, even though the math feels unfair. By month two the survivors are no longer separating you from the conditions you walked into. They’re treating you as the conditions.