TL;DR: Shuttle diplomacy is ineffective, because it quickly degrades into “he said, she said”. Choose supervised confrontation instead. Or hold people accountable to having a direct conversation.


Patrick is an engineering manager and has twelve direct reports. These direct reports are members of several teams, so Patrick cannot sit with all of them every day, and does not notice everything that is going on first-hand. One day, during a one-on-one with Robert, an engineer, there is a complaint: Robert says that he is not entirely happy with his teammate Christopher. Christopher seems to look for little arguments, takes every word literally and holds it against people, and wants to “win” discussions instead of finding a good outcome collaboratively. Robert concludes: “Sometimes, it feels like we are opponents rather than teammates. I found myself avoiding discussions with him, because they are usually so exhausting.”

This worries Patrick a bit. He has been sensing that the atmosphere in Robert’s team was not entirely healthy, and that there were some tensions. Always wanting to address issues early before they turn into something big, he decides to find out about Christopher’s view of things. He tries that in his next one-on-one with Christopher a few days later:

Patrick: “How are things in your team?”

Christopher: “It’s good. Everything all right.”

Patrick: “How does it work in your team, do you collaborate a lot and have discussions about problems, or does everyone usually work alone?”

Christopher: “Hm, it’s both, I would say. At the beginning of a bigger feature or project, we often have discussions and try to find the best ideas together, but then everyone usually works alone, until he is blocked somehow, or needs somebody to review his code.”

Patrick: “When you have discussions, do opinions sometimes clash? Are there arguments?”

Christopher: “Hm, not really. I mean, it’s not like we share 100% the same opinion all the time, but usually, we agree on something pretty quickly. Everybody has been here for some time, and knows the system and the processes. Usually, we quickly find a solution that everybody supports.”

What to do?

Hmm. This did not bring much insight. Patrick is now faced with Robert telling him one story, and Christopher telling a completely different story. According to Robert, communication on the team is suffering. According to Christopher, everything is fine. Either Christopher really does not see anything wrong with the communication on the team, or he does not want to talk about it. Be that as it may, Patrick could not get much out of him.

This is the great weakness of shuttle diplomacy, which I have written about before. It can quickly degrade into a slightly more grown-up version of “he said, she said”. Patrick now has to make a decision if he wants to leave it at that or follow up on it. He could tell himself: “If Christopher does not see an issue, then probably Robert is being a bit oversensitive.” However, since Patrick has been an engineering manager for some time, he knows that these little things usually do not go away on their own, but rather become bigger. If Christopher is really displaying conflict-seeking behaviour, then ignoring it sends the message that this is okay. Patrick does not want to send this message.

Forced intervention

I used to do something drastic — and a little mean — from time to time: I would call both persons into a room and have them talk about the issue at hand, without any previous announcement. You might call it “supervised confrontation”, or “forced intervention”, because I was forcing them to have the conversation. I used do that because it sends a very clear message: “I will not let little conflicts spoil an entire team and turn into something big at some point. I want you to get your shit together and deal with this.”

However, doing it this way is also little mean. Firstly, I was using and, in a way, spreading, knowledge that I had gathered in one-on-ones with employees. This is dangerous, because people must be able to trust that the contents of one-on-ones are confidential.

Secondly, this technique is mean because none of the participants except me agreed to have the conversation. It is a forced intervention, and everybody will be unprepared. Without agreement to have the conversation, there might also be reduced buy-in from the conversation partners with respect to the outcome of the discussion.

Strengthen responsibility

Show them the way

While it can be necessary to force people to talk, it is not always the best idea for the reasons just mentioned. Nathan Barry described a different approach that is less directive. When somebody tells you about a people problem, you answer (I’m quoting from Nathan’s post here):

“I’m going to assume that you’re either telling me this because…

A. You need to tell someone in order to get your thoughts clear before you have the direct conversation with your co-worker or…

B. You’re worried about how that conversation will go and so you want someone else to be there as an observer or to help keep the conversation on track.

So, which one is it and when are you going to have the conversation?”

I think this is brilliant, because it spreads responsibility more naturally. You, as the manager, refuse to take full responsibility to resolve your employee’s problem. Instead, you throw the ball back and …

  1. …make clear that it is their responsibility to work towards a solution.
  2. …hold them accountable to solving it soon (“When are you going to have the conversation?”).
  3. …offer your help.

The only responsibility the manager assumes is the responsibility to coach the employee through the situation. After all, every problem that happens in the manager’s scope is also his problem. The employee, in turn, assumes the responsibility to actually do something about it. Thus, responsibility is shared. In contrast, forced intervention takes responsibility off employees’ backs entirely. The manager decides if, when, and how the conflict will be tackled. If you do that too often, people will get used to the idea that “the boss” takes care of all difficult situations. They will be even less prepared to handle things themselves, and the manager will have to handle more and more situations, thus becoming a people skills bottleneck.

As a rule of thumb, it is always a good investment of your time and effort to empower people to deal with conflict, because a certain level of conflict is unavoidable as soon as your organization consists of more than one person.

There is another way in which Nathan’s method might protect manager time: If the employee refuses to assume her share of the responsibility, then the problem was probably not big enough, and the manager does not have to do anything about it. If this happens once or twice, people will be more thoughtful about which issues they discuss with their managers.

This is the only potential downside I see: Will very shy employees stop mentioning problems, because they know that their manager will make them have a critical conversation? When you perceive this danger, you have to coach and encourage your employee extra carefully, and probably act as a moderator or facilitator during the conversation. So there might still be supervised confrontation, but this time, it is voluntary. This secures more buy-in from the employee.

Conclusion

Many issues between teammates will not just go away by themselves. As a manager, you should address them. However, you cannot handle every single problem on your own, and you should avoid shuttle diplomacy like the plague. Instead, empower people to handle conflict themselves. Supervised confrontation is a good first step to prepare them for that. Ultimately, only a direct conversation between conflicting parties can bring lasting results. Let your people choose if they want your help in it or not, and hold them accountable to actually having the conversation.

Time investment

This blog post took me about 2.5 hours to write.