In tech, we obsess over eliminating performance bottlenecks — but often forget the most human one of all: the manager.

We tune our APIs for speed. We optimize our databases to scale. We throw load balancers in front of everything so no one service is overwhelmed.

But what happens when the bottleneck isn’t infrastructure — it’s organizational?

More specifically, it’s a manager who, instead of enabling communication and execution, becomes the hub through which everything must pass. They become a human load balancer — or worse, a proxy — between the people doing the actual work.

And that, ironically, makes everything slower.

Engineering Management as a Bottleneck: A Common Pattern

This isn’t just theory — it’s a real pattern seen across engineering organizations of all sizes.

Imagine a Site Reliability Engineer (SRE) trying to resolve a high-severity incident. They need help from a backend developer who owns a key system. But instead of reaching out directly, protocol (or habit) says: go through the manager.

Or a product engineer needs input from a tech lead about future architecture decisions. But the path forward is once again… the manager.

Instead of streamlined, peer-to-peer conversations, the manager becomes a single-threaded interface for the team. Just like a system that routes all traffic through a single node, it slows everything down.

Why This Happens: Root Causes of Human Bottlenecks

Understanding the why is key to fixing it. Managers don’t become bottlenecks out of laziness or malice — it’s often systemic and unintentional.

1. Legacy Management Culture

Traditional org charts position managers as the primary channel for information and decision-making. In modern, agile environments, this just doesn’t scale.

2. Desire for Control or Oversight

Some managers fear being “left out of the loop.” They centralize communication to stay informed — but end up slowing everything down.

3. Unclear Communication Norms

When it’s unclear who owns what, people default to “ask the manager.” It feels safe — even if inefficient.

4. Lack of Trust

Micromanagement and bottlenecking are often signs of a deeper issue: a lack of trust in the team’s ability to self-direct or problem-solve without oversight.

The Consequences: What Happens When Managers Become Load Balancers

These patterns might seem harmless at first, but the ripple effects across engineering culture and productivity are significant.

  • 🔁 Increased Latency
    Every request is slowed as it waits for triage. This is especially harmful in high-stakes situations like incident response or time-sensitive product work.

  • 📉 Loss of Context and Nuance
    Managers aren’t always the subject matter expert. Requests get rephrased, details get lost, and problems are misunderstood.

  • 🧠 Decision Fatigue and Burnout
    Managers overwhelmed with constant context-switching end up exhausted — and less effective at the work that actually matters: strategy, mentorship, and team growth.

  • 🛑 Team Dependency and Stagnation
    Teams that rely on a manager for every decision become less autonomous, less creative, and more passive.

Engineering Leadership That Scales

So how do we shift from gatekeeping to enabling?

🔓 1. Build a Culture of Direct Communication

Normalize cross-functional, cross-level conversations. If an engineer has a question for another engineer — let them talk. No intermediaries needed.

Use tools like Slack, shared docs, or async video to open up access to knowledge and decision-makers without over-relying on management channels.

🧭 2. Clarify Roles, Responsibilities, and Boundaries

When people know who owns what, they don’t default to “ask the manager.” Clear ownership leads to faster decisions and fewer bottlenecks.

Tools like RACI matrices, team charters, or internal documentation help reinforce this clarity.

🧑‍🏫 3. Redefine the Manager’s Role

Great managers aren’t traffic controllers — they’re multipliers. They empower others to solve problems, encourage initiative, and remove roadblocks.

Their job isn’t to be in every meeting or approve every decision — it’s to build a team that doesn’t need them in order to move forward.

🚦 4. Invest in Psychological Safety

Encourage teams to make decisions, take risks, and communicate directly without fear of stepping on toes. Trust is the foundation for autonomy.

When Managers Step Back, Teams Speed Up

If you’re a manager reading this, ask yourself:

  • Am I involved in decisions I don’t need to be in?
  • Do people default to me for answers that others could provide?
  • Am I accidentally slowing down the very team I’m trying to help?

Stepping back isn’t about doing less. It’s about doing the right things — coaching, aligning, enabling — so your team can operate at full speed.

Because when communication flows freely and decisions are made where the context lives, your team doesn’t just move faster — it grows stronger.

✏️ Personal Notes

  • At the end of the day, this is just one person’s perspective — and someone who’s never actually been a manager. Some points may be incomplete or even a bit shortsighted.

  • That said, these thoughts come from years of observation, experience as an engineer, and many conversations with peers. My opinions may evolve — and honestly, I hope they do.

  • This piece is also a small attempt to help a manager out there — maybe even you — through self-reflection. If it sparks a moment of clarity or curiosity, it’s done its job.

  • Because at the core of it, humans are still the hardest part of tech.