The Engineering Leadership Multiplier
A leader's impact is measured less by personal heroics and more by the decisions, ownership, and leadership capability they build into the organization.
Most engineering leaders I know were promoted for being exceptionally good at solving problems.
They were the engineer who could find the root cause when nobody else could. The one who understood how the systems really fit together. The one people called when something went badly wrong.
Then they were given a team. Then a larger one. Then several teams.
And many of them, quite naturally, kept doing what had made them successful: solving the hardest problems personally.
I recognize that instinct because I've had it myself. When something difficult surfaces, stepping in feels like the responsible thing to do. It's faster. It's satisfying. The team is grateful.
But somewhere along the way, I realized that the skill that gets you into leadership is often the one that limits you once you're there.
A leader's real impact isn't how many problems they can solve. It's how many people they've developed who can solve problems without them.
The hero pattern
Consider a situation I've seen play out more than once.
An engineering manager leads a group of around forty engineers across several teams. They are technically outstanding. They still review the most important pull requests, join every serious incident bridge, and are the final word on most design decisions.
For a while, this works. Quality is high. Incidents are resolved quickly. Senior stakeholders trust the manager because, whenever something goes wrong, they're already on it.
Then the cracks begin to show.
Decisions queue up whenever the manager is in back-to-back meetings. Tech leads stop proposing designs and start asking what the manager would prefer. The strongest engineers, the ones with leadership potential, begin to leave, because there's no room for them to grow. And the manager is working sixty-hour weeks and still feels behind.
Nobody intended this. The team didn't become dependent because they lacked ability. They became dependent because, every time a hard problem appeared, they learned that the real decision-maker would step in.
Every rescue teaches the team something. Usually, it teaches them to wait.
I wrote separately about why leadership is about enabling people rather than micromanaging them. Here I want to focus on what a leader should do instead: how to deliberately build capability that multiplies.
Look at where your time actually goes
One of the most useful exercises I know is uncomfortable but simple: look honestly at your calendar for the last four weeks.
When I've done this, and when I've asked other leaders to do it, the result is usually surprising. A large share of time goes to status meetings, approvals, and problem-solving that someone else on the team could have handled. A small share goes to the things only a leader can do: setting direction, developing people, managing important relationships, and making the genuinely difficult trade-offs.
A leader I once worked with found that nearly two-thirds of their week was spent in meetings whose primary purpose was to keep them informed. The team was effectively spending hours each week preparing updates for one person.
They made a few changes. A weekly ninety-minute status meeting became a short written update that took five minutes to read. A focused thirty-minute risk review replaced several ad hoc check-ins. Routine approvals were delegated to tech leads with clear criteria.
They didn't lose visibility. They gained something like a day a week, and they spent it on coaching and on the cross-team problems that had been waiting for months.
Your calendar is the most honest statement of your leadership priorities.
Delegate decisions, not just tasks
There is a difference between giving someone work and giving them ownership.
Handing out tasks while keeping every meaningful decision isn't really delegation. It's dispatch. The person does the work, but the judgment, and therefore the growth, stays with the leader.
Real delegation means giving someone a decision area. For example: "You own the reliability of the payments platform. Here are the constraints: budget, compliance requirements, and the availability target we've committed to. Here's how we'll know it's working. Come to me when you need help or when something crosses these boundaries."
Then comes the hard part: living with their decisions.
Early on, those decisions may not be the ones you would have made. Some will be slower. Some will be worse. I've learned to accept that, within reason, because the alternative is worse. A good-enough decision made without me is often more valuable than a perfect one that depends on me. Only one of them scales.
What I watch closely is not whether every decision matches mine, but whether the person's judgment is improving, and whether they know when to escalate.
Build leaders, not just teams
The leaders I admire most measure themselves partly by the leaders they leave behind.
That doesn't happen by accident. It requires deliberately creating opportunities for others to lead, before they feel completely ready.
Some of the most effective ways I've seen:
- Let a tech lead run quarterly planning for their area, with the leader as a reviewer rather than the author.
- Rotate incident command so that more than one person learns to lead under pressure.
- Send someone else to the steering committee to present their team's work, and support them rather than speak for them.
- Name potential successors explicitly and give them stretch responsibilities with real stakes.
Each of these carries some risk. The planning might be rougher. The presentation might not land as well. But each one builds judgment and confidence in a way that watching never will.
If no one on your team could step into your role within a year, that is worth reflecting on. It's a risk for the organization, and it usually means people aren't getting the opportunities they need to grow.
Build a management system, not a dependency
Strong teams don't rely on a leader's memory and attention to stay on track. They rely on a reliable operating rhythm.
The systems I've found most effective are surprisingly simple:
- A small set of meaningful measures, such as delivery flow, reliability, and quality, reviewed regularly and trusted by everyone.
- A regular risk review, where problems are raised early, before they become escalations.
- Explicit commitments, written down, so that progress is visible without anyone needing to ask.
- A decision log, so that context survives beyond the meeting where a choice was made.
The aim is to make work visible before it becomes a crisis. And one caution: status reporting can quietly become a substitute for progress. If teams spend more time describing their work than doing it, the system needs simplifying.
My perspective: measure yourself by what happens without you
I have come to believe that the most honest test of a leader is what happens when they're not there.
Take a real holiday. Two weeks, genuinely offline. Then come back and look carefully. What broke? What waited for you? Which decisions didn't get made? Who stepped up?
That list is your leadership backlog. It is a far better measure of effectiveness than how many fires you personally put out last quarter.
There are times when a leader needs to be hands-on. During a major incident, a turnaround, or when a team has lost confidence, strong direct involvement can be exactly what's needed. I don't think stepping in is wrong. The risk is that crisis mode is comfortable for people like us, and "for a while" quietly becomes the way we operate.
My view: the multiplier isn't how much a leader can do personally. It's how much keeps happening, and keeps happening well, after they step out of the room.
Great engineers solve hard problems. Great engineering leaders build teams that solve the next ones without them.
Reactions & comments
Sign in with GitHub to react or join the conversation.
Discussion is not configured yet.
Add your giscusrepoId and categoryId to site.config.ts.