Over the past couple of years, as we grew from under ten people to around thirty, I watched the same pattern several times. Someone strong takes on one or two direct reports and does well. We add a third. The job gets harder in a way that effort alone does not fix.

For a while I read that as a ceiling. I was wrong.

The third report is not where good people stop growing. It is where the job itself changes. And it is the point where coaching starts to pay for itself. We learned to treat the jump as a transition to invest in, not a verdict to hand down. The same people who strained at three reports now run healthy teams of four and five.

Why the third report is different

The work of leading a team does not grow with headcount. It grows with the connections between people, and those multiply fast. Six people already have fifteen connections between them. Double the team to twelve and you do not get thirty, you get sixty-six. Fred Brooks made the same point about software teams in The Mythical Man-Month.

The role inverts at the same time. The moves that win at one or two reports are obvious ones: jump in and fix it, be the technical anchor, hold the whole picture in your head, give fast feedback in a tight one-to-one loop. At four or five reports, every one of those moves turns into the bottleneck. Marshall Goldsmith put it in one line: what got you here will not get you there. The Leadership Pipeline calls each of these steps a passage, and the price of every passage is giving up the job that earned it.

What did not work: waiting for it to sort itself out

Our first instinct was to hope people would grow into it on their own. They did not, and that is not a character flaw. Almost nobody crosses this on instinct. The skills that make someone a strong individual contributor are the wrong skills for the new job, so left alone they reach for what has always worked: more heroics. The lead becomes the single point everything waits on, works longer to compensate, and the team learns to bring every problem to one desk.

That is the honest failure mode. It is also completely fixable.

What worked: coaching and structure at the third report

We started treating the third report as a planned transition with real support attached. Two investments did the work.

First, coaching. We put external coaching in place for the lead and paired it with internal coaching from people who had already made the jump. The outside coach gave them a place to work on the new job that was not a performance review. The internal side made it concrete to how we actually operate.

The part that mattered most was coaching them on how to coach. The hardest habit to break at the third report is solving the problem yourself. So we taught one move first: ask a good question before you give an answer. Then the bigger ones: hand over a whole problem rather than a task, and sit with a first attempt that is slower and rougher than yours would have been. Developing people is a skill, and it is learnable. It does not arrive with the title.

Second, structure. Heroics are what a lead falls back on when there is no process to fall back on. So we built the lightweight structures that let them stop being the funnel: clear ownership, real handoffs, and somewhere to look things up that is not the lead’s memory. Structure is what makes delegation safe enough to actually do.

With coaching and structure together, the same player-coaches who were underwater at three reports now run teams of four and five. The teams are steadier than they were when one person carried them.

What we do now, before the third report lands

We no longer treat a third report as a quiet headcount change. We treat it as a transition we plan for. Before we add the report, we ask three questions. Is there a coach in place, internal or external. Are we actively teaching this person how to coach, not just how to do the work. And is there enough process that the team does not route everything through one desk.

Two things worth trying this week

If you have a lead approaching their third report, start here.

Put coaching in place before the headcount arrives, not after they are underwater. Even one of your own one-to-ones spent modelling a single move, asking a good question before giving an answer, changes how they lead.

Then find the one thing that works today only because it lives in the lead’s head, and turn it into a small written process this week. That single step lifts load off the lead and gives the team somewhere other than the lead to look.

I do not think we have this fully solved. We are still learning how early to start and how much structure is enough. But the lesson has held. The third report is a coaching problem, and coaching is something you can invest in. Treated that way, the wall becomes a step.