I almost called this post “Leadership Lessons I Wish I’d Learned Sooner” — then realized that framing misses the entire point. The lessons aren’t late. They arrived exactly when I was ready to absorb them, which is the whole problem with leadership development: you can’t speed it up the way you speed up a code deployment.

I’ve managed engineering teams for the better part of a decade now. Small teams (3 people at Backer), distributed teams (9 engineers across 4 time zones at TaskRabbit), and DevRel teams at Google spanning multiple offices. The contexts differ wildly, but one pattern has been consistent: the leadership skills that matter most are the ones that resist being learned quickly.

The workshop illusion #

The tech industry loves compressed learning. Bootcamps. Crash courses. “Become a senior engineer in 12 weeks.” There’s a version of this for management too: the two-day leadership offsite, the weekend workshop on emotional intelligence, the lunch-and-learn on giving feedback.

I’m not going to tell you these are useless. I’ve attended several, and I picked up frameworks and vocabulary that proved helpful. But there’s a gap between knowing a framework and having the judgment to apply it. That gap is measured in years, not hours.

Take feedback. Every leadership workshop covers feedback. The SBI model (Situation, Behavior, Impact) takes about ten minutes to learn. Using it effectively with a defensive senior engineer who’s been at the company longer than you have? That takes dozens of uncomfortable conversations, a few badly handled ones, and the slow accumulation of intuition about when to push and when to back off.

The framework is the easy part. The judgment is the hard part. And judgment develops on its own timeline — largely indifferent to how many workshops you attend.

What actually builds leadership capacity #

Looking back at my own development (and watching the managers I’ve coached), three things compound over time more than anything else.

Coaching relationships that persist. Not a one-off coaching session; a sustained relationship with someone who sees your blind spots and has permission to name them. I resisted coaching for longer than I should have. It felt like admitting I didn’t know what I was doing. Which, in retrospect, was exactly the point. The best coaching relationship I’ve had lasted over a year, and the growth came from pattern recognition — my coach noticing the same avoidance behavior in different situations until I couldn’t deny it anymore.

Feedback loops with actual teeth. 360 reviews are fine. But the feedback that changed me most came from engineers on my team who trusted me enough to be honest. At TaskRabbit, one of my direct reports told me that I tended to “solve too fast” — jumping to solutions before the team had finished thinking through the problem. That stung. It was also completely right. I couldn’t have heard that feedback in my first year of management. By my fourth year, I could hear it and actually do something about it.

Accumulated reps in high-stakes situations. There’s no substitute for having navigated a performance improvement plan, delivered a layoff message, mediated a conflict between two senior engineers who genuinely disliked each other, or made a hiring decision you weren’t sure about. These situations don’t come up daily; they might come up twice a year. Each one teaches you something that no simulation can replicate.

The compounding problem #

Leadership development compounds. But unlike financial compounding (where the math is clean and predictable), leadership compounding is messy and non-linear.

Year one of management: you’re learning to give feedback without breaking relationships. Year three: you’re learning to calibrate feedback intensity to the person and situation. Year five: you’re coaching other managers on how to give feedback. Year seven: you’re building a team culture where feedback happens organically without you being involved.

Each stage requires the previous one. You can’t skip to year five without the accumulated experience of years one through four. And the rate of compounding varies — some lessons click fast; others take years of repetition before they land.

I promoted two engineers during my time at TaskRabbit. The conversations around those promotions — the calibration, the documentation, the advocacy — taught me more about developing people than any course on talent management. But I wouldn’t have had the context to learn from those conversations if I hadn’t already spent years observing how promotions (including my own) were handled at previous companies.

This is what I mean by marathon. The distance isn’t optional.

Self-awareness as the foundation #

Every leadership framework I’ve encountered eventually bottoms out at self-awareness. It doesn’t matter how good your feedback model is if you can’t recognize your own patterns. It doesn’t matter how many books you’ve read about delegation if you don’t understand why you’re holding on.

Building self-awareness is its own marathon. I’m Brazilian, and the cultural norms around directness that I grew up with don’t always map cleanly to engineering management in the Bay Area. I had to learn (slowly, through mistakes) when my natural directness was an asset and when it was landing as abrasiveness. That calibration is ongoing; I don’t think it ever finishes.

The tools help — journaling, coaching, 360s, peer feedback. But the real mechanism is time plus reflection. You need enough experiences to identify patterns, and enough reflective capacity to examine those patterns honestly. Remove either ingredient and growth stalls.

What I’d tell a new engineering manager #

I almost didn’t write this section because “advice for new managers” is the most overdone genre in tech leadership writing, and I’m skeptical of my own authority on the subject. But here’s what I actually believe, stripped of the usual platitudes.

The first year will feel like you’re failing. You probably are, in small ways. That’s how it works. The engineers on your team had years to develop their technical skills before they were expected to perform at a senior level. You get approximately three months before people start expecting you to be good at management.

Find one person who will tell you the truth. Not your manager (their incentives are complicated). Not your peers (they’re navigating the same uncertainty). Find a mentor or coach outside your reporting chain who has no reason to be diplomatic.

Track your own patterns, not your team’s metrics. Dashboards are seductive. Velocity charts, deployment frequency, incident counts — these give you the illusion of understanding what’s happening. They’re useful, but they’re lagging indicators. Leading indicators are your own behaviors: how you run 1:1s, how you respond to conflict, how you make time for strategic thinking versus reactive firefighting.

Accept that this will take years. Not months. Years. The marathon metaphor isn’t motivational; it’s descriptive. There are no shortcuts that don’t cost you something later.

The long game #

I’m writing this at a moment of transition — reflecting on several years of engineering management across very different companies and contexts. The leader I am today would be unrecognizable to the one who started managing a team of three at Backer in 2018. Not because I had some dramatic transformation. Because hundreds of small lessons accumulated into a different way of seeing teams, systems, and my own role within them.

That accumulation is the marathon. It’s not glamorous. It doesn’t make for a compelling LinkedIn post. But it’s the only thing that actually works.

If you’re early in your leadership journey and feeling impatient, I get it. I was impatient too. The only honest thing I can tell you is that the impatience itself fades as you accumulate enough experience to appreciate how much you still don’t know. And paradoxically, that’s when the real growth accelerates.