The Job Nobody Trains You For

When I first took on people management, my employer sent me to three days of management training. I didn’t think much of it at the time. It took me a few years, and a lot of conversations with other new managers, to realize how lucky I was.

Most new managers get nothing at all. In a 2023 survey by the Chartered Management Institute, 82 percent of people who moved into management said they’d never had any formal management or leadership training. Earlier research from the Center for Creative Leadership landed in similar territory, with about 60 percent of new managers reporting no training for their first leadership role.

So how do people end up in the job? When Gallup asked U.S. managers why they thought they’d been chosen, the two most common answers were their tenure and their success in a previous, non-managerial role. That sounds a lot like the Peter Principle, the decades-old observation that people keep getting promoted until they land in a job they’re not suited for. It’s easy to dismiss as a cynical joke, but economists have actually tested it: a 2019 study of sales workers at 131 firms found that companies leaned on current job performance when deciding whom to promote, even when other traits did a better job of predicting who would succeed as a manager.¹

I’ll admit my own experience doesn’t line up with this so neatly. Over more than two decades of building and growing teams, through hiring pushes and team merges alike, I’ve seen plenty of thoughtful promotion decisions. But even a well-chosen new manager is walking into a different job than the one they just left. The work runs on judgment, context, and conversation: deciding what not to build, telling someone their work isn’t landing in a way they can actually hear, keeping a team steady through a reorg, and noticing that the quiet person in the corner is the one who spotted the real risk. Being great at your previous role doesn’t teach you any of that, no matter how you got picked.

When nobody teaches this stuff, the costs tend to be quiet, and they’re usually paid by someone other than the manager:

Managers fall back on handing out tasks, because that’s the part they already know how to do.
Retrospectives turn into status meetings, or into a kind of theater where everyone agrees things went fine.
Feedback only shows up when something has gone badly wrong, so it lands as punishment instead of information.
People leave. In that same CMI survey, 31 percent of managers said they’d quit a job because of a bad relationship with their own manager.

AI is raising the stakes. Nearly every leader I talk to is being asked to absorb some kind of upheaval right now, from how code gets written to what a career ladder even means when agents are doing work that used to define a senior engineer’s week. Uncertain moments like this one are when command-and-control starts to feel tempting. Crack down, tighten up, add a process! It feels like leadership because it’s visible. In my experience, though, it usually leaves a team less able to adapt, not more.

Three days of training goes by quickly, and no course covers everything. But if I could add a few items to every new manager’s syllabus, they’d map pretty directly onto those failure modes:

Hand out problems, not tasks. Tell people what outcome you need and why it matters, then let them figure out the how. You’ll get better solutions, and you’ll find out sooner who’s ready to grow.
Reflect together and without blame. If people get punished for telling you what broke, they’ll stop telling you, and you’ll hear about the next problem much later and at a much higher cost.
Make feedback routine, not a special event. I’ve always made career development and feedback a standing part of my 1:1s, so nobody hears about a gap for the first time in a performance review. When feedback is ordinary, it tends to land as information rather than a verdict.
Get comfortable saying “I don’t know yet.” When you say it out loud, your team can say it too, which is exactly what you want when the ground is shifting under everyone.
Find somebody who’ll tell you the truth. A mentor, a peer manager, a coach… someone outside your reporting line who can hear “I think I blew that conversation” without it ending up in your review.

None of this is something you should already be good at on Day One. The leaders I admire most stay genuinely curious about the people in front of them, hear hard things without getting defensive, and treat their own mistakes as something to learn from. All of that can be learned. I was lucky enough to get three days’ head start; most people deserve at least that much.

Two caveats worth knowing. The Gallup numbers are managers’ own explanations of why they were picked, and the Peter Principle study only looked at sales, where performance is unusually easy to measure. I’d love to see similar research on engineering and other knowledge work. ↩

Footnotes