Seven People, Twenty-One Relationships

If you coach teams, or lead one, you’ve probably had this experience. You can describe each person on the team clearly. Sarah is sharp and direct. Marcus is thoughtful and measured. Priya is the one who keeps the room calm. Each of them, easy. The three of them in a meeting together, harder. The whole team at a quarterly offsite, hardest of all.

That isn’t a failure of attention. It’s a structural problem, and it’s the central problem of team work.

The math nobody warns you about

By the time a team is large enough to need outside help, no single observer (not the coach, not the leader, not even the members themselves) is holding the full map.

This is why coaching teams feels different from coaching individuals, even when you’re good at both. With one person, the unit of analysis matches the person. With seven people, the unit of analysis is the relationship, and there are too many of them. You can be excellent at reading rooms and still not have language for the structure underneath what you’re reading.

Most of the tools that exist for this work were built to describe individuals. They give you a profile per person and leave the relationships to your intuition. That’s a fine approach when teams are small and time is plentiful. It breaks down quickly when neither of those is true, which is most of the time.

Here’s what actually makes the work hard.

Individual reports answer the wrong question

When a coach is hired into an executive team and given access to the personality assessments everyone has on file, they end up with seven documents about seven people. None of those documents are about how those seven people work together. The coach is supposed to construct the relationship picture in their head from the individual pictures. That’s a real cognitive load, and it scales badly.

It also assumes that pairwise dynamics are predictable from individual traits, which they aren’t. Two direct people don’t necessarily get along better than one direct person and one harmonizer. They sometimes collide harder, because there’s no one in the dynamic willing to let something go. Two careful introverts don’t necessarily have great meetings; they may both wait for the other to speak. The traits don’t add up. They interact, and the interactions are what the team actually experiences.

Every observer has a partial, filtered view

The leader’s view of the team is one of N possible views, not the view. Their experience of “what Sarah is like” is shaped by their own personality interacting with Sarah’s. Marcus has a different Sarah. Priya has a third Sarah. None of them are wrong. They’re each having a slightly different relationship.

This matters more than it sounds. When a leader says “Sarah’s a strong communicator but she gets defensive about feedback,” they’re describing the Sarah they get, which is partly a function of how they deliver feedback. A peer giving feedback the same way might get a different Sarah. A peer giving feedback differently might get a third Sarah. The leader treats their experience as the data; it’s actually a sample of one.

Coaches run into the same problem from the outside. You meet with the CEO Tuesday. You meet with the COO Thursday. You hear two versions of last week’s argument and have to construct a third version that explains both. Without a structural model of what’s happening between those specific people, that triangulation is mostly intuition. You’re often right; you’re sometimes badly wrong; you usually can’t tell which.

The signal isn’t where the noise is

The pair that’s loudly disagreeing in meetings often has a healthy dynamic. They’re disagreeing because they trust the room enough to disagree out loud. The pair that’s quietly aligned on paper may be the one carrying the actual strain, because the strain went underground a long time ago and nobody’s surfaced it.

This is one of the most counterintuitive parts of team work, and it’s where surface-level reading fails most predictably. Coaches develop their nose for it over years. The nose is real, but it’s not transferable, and it doesn’t scale. You can only have so many noses on so many teams.

You can’t fix what you can’t compare side by side

Even when a coach correctly identifies that something is off between two specific people, the next move is hard. What’s actually different about how those two people communicate? What does each of them experience the other as doing? What would it mean for them to adapt to each other versus build norms around the difference?

Most assessment outputs aren’t structured to answer those questions. They’re structured to describe individuals. To compare two people, the coach has to flip back and forth between two reports, mentally extracting the points of contrast, then reconstruct what those contrasts probably mean for the relationship. It’s possible. It’s slow. It rarely happens between sessions, which is when the coach actually needs the answer.

What would actually help

Tools that match the shape of the problem instead of the shape of the easier problem next to it. That means a few things.

It means describing each pair, not just each person. The relationship is the unit, so the artifact should be the relationship.

It means asymmetry. Sarah’s experience of Marcus is not the same as Marcus’s experience of Sarah, and the artifact should reflect that. The same difference between two people lands differently from each side. A guide that pretends otherwise is rounding off the part the coach actually needs.

It means findings rather than predictions, and questions rather than prescriptions. The coach isn’t being replaced by the tool. The tool’s job is to put a clean signal on the desk so the coach’s judgment can do its work faster.

And it means holding the structural map between sessions, so the coach isn’t reconstructing it every time. The map should be something you can return to, share with the leader, work from in the next conversation. Right now, for most coaches, that map lives only in their heads. That’s an enormous amount of cognitive load to carry through a week of back-to-back sessions.

Why this matters now

For a long time, the structural problem we just described was unsolvable at the level of tooling. You couldn’t generate a useful per-pair guide for every relationship on a team without a small army of analysts, so tools compressed people into types and let the coach do the rest. That was a reasonable shortcut for the era it was built in. It’s a lot less reasonable now.

The shape of the problem hasn’t changed. Twenty-one relationships is still twenty-one relationships. What’s changed is what’s possible. We can now produce specific, asymmetric, research-grounded guidance about how each pair on a team is likely to communicate, where they’re likely to align, and where they’re likely to need adaptation. That doesn’t replace the coach’s judgment. It gives the coach’s judgment a much better starting point.

If you’ve spent years feeling that the personality tools you have access to don’t quite match the problem you’re actually trying to solve, you’ve been right. The tools were built for individuals. The problem was always the team.