You have 11 direct reports, your 1:1s keep slipping, and decisions are piling up in your inbox. The question isn't whether you're bad at this. It's whether your span of control has outgrown you, and that's a math problem with a structural fix.

You have 11 direct reports. Growth is fast. Your 1:1s keep getting bumped, decisions are piling up in your inbox, and you have not had a real strategy thought in three weeks. So you ask the obvious question. Do I add a manager, or am I just bad at this?

That question has a name. It is called span of control, and getting it right is one of the quiet decisions that separates companies that scale cleanly from ones that grind.

Span of control is the number of people who report directly to one leader. Too few, and you are paying for management nobody needs. Too many, and quality collapses, people stop getting coached, and you become the bottleneck. This is the same trap behind the founder bottleneck, just one layer down.

‍

What is a good span of control?

The number first, then the nuance.

For knowledge work and startups, the sweet spot is 5 to 10 direct reports. Engagement data points to a peak around 8 to 9. Below 5, you usually have an over-managed, expensive layer. Above 10, management quality starts to slide.

The big tech companies land in that band too, though the specific figures come from industry reporting rather than the companies themselves. Reporting on their org design puts Amazon managers at roughly 6 to 8 direct reports and Google closer to 7 to 10, and both have been deliberately widening spans, Amazon through what it calls its "builder ratio" and Google by cutting manager and VP roles about 10%. They are running the flattening experiment at scale, and not by accident.

The work changes the number, though. Repetitive, well-defined roles can stretch much wider, sometimes 15 to 25. Complex, creative, or fast-changing work needs a tighter span, because each person needs more of your attention. A founder juggling strategy on top of people rarely sustains more than 5 to 7 well.

‍

The cost of getting it wrong

Span is not a tidy org-chart detail. It moves real numbers.

Gallup found managers drive 70% of the variance in team engagement. Stretch a manager too thin and that lever snaps. A widely cited study by Doran (2004) found that every extra 10 people in a manager's span lifts staff turnover by around 1.6%, and first-year employees feel it most. Wider spans quietly cost you your newest people.

There is a cost on the other side too. Pile on too many management layers and decisions slow to a crawl. Freshworks estimated that organizational complexity, mostly excess layers, can eat up to 7% of annual revenue, roughly the size of a real R&D budget. And the appetite for fixing it is there: in a McKinsey survey of more than 2,500 leaders, two-thirds called their own organizations overly complex. Flattening, done right, speeds decisions and cuts that drag.

So you are threading a needle. Spans too wide break management. Layers too deep break speed. The goal is the leanest structure that still lets every person get coached.

‍

The warning signs your span is too wide

You will feel this before you see it on a chart.

If three of those are true, your span has outgrown you. That is not a character flaw. It is a math problem with a structural fix. The same logic applies to single points of failure on your team, which we cover in the bus factor test for startups.

‍

When to add a layer (and when not to)

A simple trigger: when you cross roughly 8 direct reports and growth is steady, start planning a layer. Most founders should add one by 10 to 12 reports. A founder also carrying strategy should pull that trigger earlier, closer to 7.

But adding a layer is not free, so do it with intent.

Promote or hire? You can grow a manager from within or hire one in. Internal promotion keeps context and rewards loyalty, but only if the person actually wants to lead and you train them for it. The single biggest mistake here is promoting your best doer and assuming leadership will follow. It will not, not without support. Our manager competency framework lays out the skills to look for and build.

Add the layer without bloating the org. A new layer should never push your company past about six total layers. The most agile orgs run three or four. If a new manager just adds a box that decisions have to pass through, you have made things slower, not better. The fix is to pair the new layer with real delegation. The manager has to own decisions, not relay them. If you skip that part, you have built a bottleneck with a fancier title. Our guide on how to delegate as a founder is the missing half of this move.

‍

Try this audit this week

You can make this decision with data, not vibes. Spend 30 minutes here.‍

1. ‍Count your real span. List everyone who actually comes to you for decisions, not just the names on the org chart. Dotted lines count.‍
2. Score the work. Mark each report as simple, complex, or new. The more complex and new ones eat more of your time, so weight them double.‍
3. Check the warning signs. Run the five signals above. How many are true right now?‍
4. Pick the move. If your weighted span is past 8 and the signs are flashing, name the layer you need and who fills it. If you are unsure whether you need a manager, a chief of staff, or a COO, our breakdown of chief of staff versus COO at a startup sorts it out.

If you are about to build new managers, do not throw them in cold. A cohort program shortens the learning curve and protects the teams underneath them, which is exactly what our leadership training for companies is built to do.

‍

Grow the org without growing the gridlock

Span of control is not about drawing a neat org chart. It is about making sure every person in your company still gets the leadership they need to do their best work.

Get the number right, add layers on purpose, and pair every new manager with real ownership. That is how you grow the org without growing the gridlock.

Frequently Asked Questions:

‍