The Hardest Part of Scaling to 50+ Engineers Wasn't Technical—It Was Letting Go of Direct Management and Building a Leadership Team

I want to talk about something that doesn’t get discussed enough in CTO circles: the hardest part of scaling past 50 engineers wasn’t technical—it was letting go.

Letting go of direct management. Letting go of reviewing every architecture decision. Letting go of knowing every line of code. Letting go of being the most technical person in every room.

The transition from hands-on technical leader to executive building a leadership team was harder than any technical challenge I’d faced in my career. And I suspect I’m not alone.

What Worked at 20 Engineers (But Doesn’t Scale)

At 20 engineers, my leadership model was deeply hands-on:

  • I reviewed every significant architecture doc personally
  • I knew everyone’s code and could jump into any PR
  • I made most technical decisions (with input, but ultimately my call)
  • I had 1-on-1s with every engineer
  • I was involved in every hiring decision
  • I could keep the entire system in my head

This felt like good leadership. I was adding value. I was unblocking people. I was making sure we maintained technical excellence.

At Microsoft and Twilio, I’d seen leaders “ascend to management” and lose their technical edge. I was determined not to be one of those CTOs who became disconnected from the work.

What Broke at 50+ Engineers

By 50 engineers, my hands-on model became the bottleneck:

  • Architecture docs piled up waiting for my review (I became the constraint)
  • Engineers stopped making decisions and waited for “Michelle’s guidance”
  • I was in 40+ hours of meetings per week, leaving no time for deep work
  • My calendar was 95% booked 3 weeks in advance
  • I knew some engineers’ names but not their work
  • I couldn’t keep the system in my head anymore

I was the single point of failure for the entire engineering organization.

And here’s the uncomfortable truth: I was holding on because letting go felt like losing my identity.

I’d built my career on being the best engineer in the room, the person who could solve the hardest technical problems, the architect who saw the whole system. If I let go of that, who was I?

The Transition: Building a Leadership Team

The shift from managing engineers to leading leaders required three fundamental changes:

1. Building the Leadership Bench

I promoted 3 Directors of Engineering and 5 Senior Engineering Managers. Not because we needed the titles, but because we needed distributed decision-making and leadership capacity.

The Directors own:

  • Architecture decisions within their domain
  • Hiring for their teams
  • Team process and delivery
  • Technical strategy for their area

I don’t review their architecture docs. I don’t approve their hiring decisions. I don’t attend their sprint planning.

That was terrifying at first. What if they make decisions I disagree with? What if they hire the wrong people? What if the architecture diverges in ways I wouldn’t have chosen?

2. Trusting Leaders to Make Different Decisions

This was the hardest part: learning to trust my Directors to make decisions I would have made differently.

Not wrong decisions—just different ones.

Early example: One of my Directors chose to build an internal tool instead of using an off-the-shelf SaaS product. I would have bought the SaaS product. But it was his call to make, not mine.

My job changed from making the decision to ensuring there was a good decision-making process. Did they consider alternatives? Understand trade-offs? Align with business needs?

If yes, then trust them—even if I would have chosen differently.

3. Shifting Focus to Systems and Culture

My time allocation shifted dramatically:

Before (at 20 engineers):

  • 60% technical decisions and architecture
  • 30% management (1-on-1s, hiring, performance)
  • 10% strategy and culture

After (at 50+ engineers):

  • 15% technical guidance (high-level, not detailed)
  • 35% leadership team development
  • 50% organizational design, culture, strategy

I moved from first-order impact (writing code, reviewing architecture) to second-order impact (enabling leaders who enable teams who build products).

That required redefining what “adding value” means at the executive level.

What I Wish I’d Known

1. Delegation Is Not Abdication

I initially swung too far—from doing everything myself to completely hands-off. Neither works.

Good delegation is:

  • Clear ownership and decision rights
  • Context and constraints (business goals, technical principles)
  • Support when needed (coaching, unblocking)
  • Accountability for outcomes (not micromanagement of process)

2. Your Leadership Team Will Make Mistakes (And That’s How They Grow)

Some of my promoted Directors made decisions I had to unwind. Architectural choices that created tech debt. Organizational structures that didn’t work. Hiring mistakes.

My instinct was to step in and fix it. But that would have undermined their authority and prevented their growth.

Instead: coaching conversations, learning retrospectives, adjusting for next time. They learned more from those mistakes than from any success.

3. Letting Go of Technical Identity Doesn’t Mean Losing Technical Credibility

I worried that if I wasn’t in the code daily, I’d lose credibility with engineers. But credibility at this level comes from:

  • Asking the right questions (even if you don’t know the answers)
  • Understanding trade-offs and constraints
  • Making engineers better through coaching and resources
  • Protecting the team from organizational chaos

I’m not the best engineer in the room anymore. But I can make the engineers in the room better at their jobs. That’s a different kind of technical leadership.

My Questions for Other Leaders

For CTOs and VPs who’ve made this transition:

  1. How did you identify who was ready for Director/Senior Manager roles? Our best engineers aren’t always our best managers.

  2. How do you balance letting leaders grow through mistakes vs. protecting the business from costly errors? Where’s the line?

  3. How did you personally handle the identity shift from “best technologist” to “leader of leaders”? What helped you let go?

  4. What does “staying technical” look like at the executive level? How do you maintain technical credibility without being in the code?

  5. How do you develop leadership capacity in your team? Are there specific programs, coaching, or experiences that accelerate leadership development?

This transition is lonely. You can’t really talk about it with your team (“hey, I’m struggling to let go of control”) without undermining confidence. But I suspect many of us are struggling with it.

Let’s talk about it openly.


Michelle Washington | CTO | Learning that leadership at scale is about enabling others, not doing it yourself

Michelle, thank you for the vulnerability in this post. The identity shift from “best technologist” to “leader of leaders” is something I’ve lived through, and it’s as hard as you describe.

The Journey from Google EM to VP Engineering

My path: IC engineer at Google → Engineering Manager → Senior Engineering Manager at Google → Director at Slack → VP Engineering at current EdTech startup.

Each transition required letting go of what made me successful at the previous level. And each time, I resisted because my identity was wrapped up in what I was good at.

What “Staying Technical” Means at VP Level

Your question about maintaining technical credibility without being in the code hits home. Here’s what I’ve learned:

VP-level “technical” isn’t about coding—it’s about:

  • Technical judgment: Understanding trade-offs, asking probing questions, spotting architectural risks
  • Technical strategy: Aligning technical decisions with business outcomes
  • Technical culture: Building environments where engineers do their best work
  • Technical talent development: Growing engineers’ capabilities and careers

I don’t write production code anymore. But I can:

  • Review an architecture proposal and spot the scalability concerns
  • Ask “have we considered the data migration path?” in design reviews
  • Push back on product timelines when technical complexity is underestimated
  • Coach a senior engineer through their first architecture design

That’s technical leadership at scale. It’s different from IC technical excellence, but it’s equally valuable.

Identifying Leadership Potential vs. Engineering Excellence

You asked how to identify who’s ready for Director/Manager roles. This was my biggest mistake early on: I promoted my best engineers, not my best potential leaders.

Result: some excellent engineers became mediocre managers. They missed coding. Their teams struggled. Everyone was unhappy.

What I look for now:

Leadership Readiness Signals:

  • Do other engineers come to them for guidance (informal leadership)?
  • Do they think about team outcomes, not just individual contributions?
  • Do they give credit to others and take blame themselves?
  • Can they influence without authority?
  • Do they develop others and make their teammates better?

Engineering Excellence Signals:

  • Deep technical skills
  • High individual output
  • Elegant code and architecture

These are different skill sets. Some people have both. Many have one or the other.

Now I explicitly create two career tracks: Individual Contributor (IC) path and Management path. Both go to executive levels (Principal Engineer / Staff Engineer is equivalent to Director).

This lets best engineers stay engineers and best leaders become leaders.

Developing Leadership Capacity

You asked about programs for leadership development. Here’s what’s working:

1. Leadership Rotations: Before promoting someone to Director, they do a 3-month “acting Director” rotation with coaching. Tests readiness without permanent commitment.

2. External Executive Coaching: We provide executive coaching for all Directors and above. Having a confidential space to work through leadership challenges is invaluable.

3. Peer Leadership Circles: Our Directors meet monthly (without me) to share challenges and learnings. Peer support matters.

4. Explicit Leadership Skill Development: We teach:

  • Giving difficult feedback
  • Strategic thinking and priority-setting
  • Managing up and across
  • Developing and coaching others

These aren’t intuitive skills. They must be taught and practiced.

The Mistake Balance Question

Your question about letting leaders make mistakes vs. protecting the business is the art of leadership. Here’s my framework:

Let them make the mistake if:

  • Impact is reversible
  • Cost is manageable
  • Learning value is high
  • They have context to make an informed decision

Step in if:

  • Impact is irreversible (e.g., major customer commitment we can’t deliver)
  • Cost exceeds learning value (e.g., multi-million dollar technical debt)
  • Safety or compliance issue
  • They lack critical context that changes the decision

Example: A Director wanted to rewrite a core service from scratch. I disagreed (high risk, unclear value). But I let them proceed with a prototype to prove/disprove the value. They learned it wasn’t worth it. Valuable learning, contained risk.

The Lonely Leadership Truth

You said this transition is lonely. You’re right. There are things you can’t share with your team, with your CEO, sometimes even with your peers.

What helped me:

  • CTO/VP peer group outside my company: Monthly dinners with other engineering leaders. Confidential space to share struggles.
  • Executive coach: Someone to process the messy emotional parts of leadership.
  • Mentors who’ve been there: Reaching out to leaders 5-10 years ahead of me for perspective.

Leadership at scale is isolating. Building support structures is essential.

Thank you for starting this conversation. It’s rare to see CTOs talk openly about the psychological challenges of leadership transition. We need more of this.