The Black Peter Principle

What has this to do with Data and AI??

In Germany there is a card game for children called Black Peter. It is a simple game where you collect pairs, and whoever is left holding the Black Peter at the end has lost. Everyone here knows it. Nobody wants to have the Black Peter.

It is the same in teams.

Each of us has already engaged with teamwork, either in workshops, training or through books. Often it is a mandatory training for employees, even for juniors. Managers receive training on it, coaches, project managers too. And yet one point is consistently overlooked, with devastating consequences. I would even say: very few of us have ever truly worked in a real team. At least not in one like this:

Although we have all learned that teams go through different phases, we tend to forget it when it matters most. During my work I was part of many different teams, and one thing above all others distinguishes a high performing team from a ragtag bunch. I call it the Black Peter principle.

Sooner or later, teams reach a point where it gets hard. And I am not talking about obvious conflict. There does not have to be a fight, a difficult person, or a crisis. It is more of a feeling. Nothing particularly unusual has happened, but something is off. It is like standing in front of a really difficult task that you haven’t mentally worked through yet. A heavy, diffuse feeling. In German you call that the “inner pig-dog”, that voice of inner resistance. In teams, this exists too, only much, much stronger. I call this unknown something the Black Peter, that everyone wants to get rid off.

And now comes the decisive moment.

Teams want to get rid of it. That is natural. But what happens next is what separates the good teams from the great ones. The organizations that allow the Black Peter to leave the team, through a manager stepping in, or simply by allowing people to only stay loosely connected to the team, these organizations will never have high performing teams. The relief is real. But so is the missed opportunity.

I compare it to a marriage. As soon as difficulties arise, and it doesn’t even have to be a fight, it can simply be something unspoken hanging between you, and then you bring in a third party. That never ends well. It takes on a new dynamic that sometimes even worsens the situation. The issue doesn’t get resolved. It gets a life on its own. The Black Peter.

I have seen teams that managed this well. Those teams couldn’t bring in anyone from the outside at all, because they were fully empowered and authorized, with senior leaders inside the team. That was in several large organizations. So it is absolutely possible. Those teams learned to sit with the discomfort, work through it together, and they delivered outstanding results.

And I have seen the other kind. All managers outside the team, authority never delegated, individuals pulled in different directions by people above them. The Black Peter just wanders through the entire organization until it is eventually dropped. And with it, the chance to grow together is gone.

The structure of the organization around the team matters just as much as the team itself. A team is not a team just because you put people in a room and give them a name.

When it comes to scaling AI across a company with large amounts of data, this is not a nice-to-have insight. It becomes the difference between transformation and expensive stagnation.

The challenge in Data and AI is even more complex, more daunting, more unclear and bigger than in a pure software development team. There is no simple roadmap you follow step by step and everything will work out.

That is exactly why building teams that truly function becomes even more important.

AI is new. We are all still learning. Maybe you are sitting on a pile of legacy data systems, or some roll-out that is only half finished. Maybe you have no real overview at all, maybe your systems are not even talking to each other yet. Maybe nobody involved fully understands the end-to-end process. Maybe it is completely unclear how the data should be used or how the whole thing should be structured. Even figuring out where to start is a challenge in itself.

The task is enormous. And it needs teams that are not afraid of Black Peter.