De-escalating peer friction without making it worse

The default workplace advice for peer friction is to have a difficult conversation. Be direct. Bring it up. Don't let it fester. The advice is correct. It is also dangerously underspecified.

If you have been in the workforce for any length of time, you have watched someone follow this advice and make the situation worse. They went in raw. They said the thing they had been thinking for three weeks. The other person responded badly. The relationship got more brittle, not less. The next time the friction came up, both people remembered the conversation and avoided each other harder.

The advice is right that the conversation needs to happen. What the advice usually leaves out is what you have to do before the conversation to make it go well. That preparation is what distinguishes a de-escalation from an escalation.

What goes wrong in a raw conversation

When you walk into a difficult conversation with a colleague without having processed what you actually want to say, three predictable things happen.

You overstate the case. Three weeks of accumulated frustration produce a version of the story in your head that includes every instance, every slight, every undermining comment. You compress this into the opening minute of the conversation. The other person hears not "the specific thing I want to address" but "everything I have ever done wrong." The defensive response is appropriate to what they heard.

You attribute intent. Underneath frustration with a colleague is almost always an interpretation of why they did what they did. Sometimes the interpretation is right. Often it is the worst-case version of an ambiguous behavior. When you express it in the conversation, you are not stating an observation; you are stating an accusation about their motivation. They respond to the accusation.

You lose the ask. Difficult conversations work when you have a specific behavioral change you want to request. Raw conversations rarely make it to the ask. They get stuck in the description of the problem, the back-and-forth about whose interpretation is correct, the litigation of past instances. The conversation ends without anyone agreeing to do anything different.

These three failure modes are not character flaws on your part. They are what happens when you try to do the structured thinking in real time, in a conversation that is already emotionally loaded. Most of us cannot do that work in real time. The structured thinking has to happen before.

The preparation that makes the conversation work

There are three things you need to have written down — in your own private space, where nobody else can read them — before you have the conversation.

First, the specific behavior. Not "they are dismissive." That is an interpretation. "In the standup on Tuesday, they cut me off mid-sentence; in the design review on Thursday, they reframed my proposal as the team's idea." Specific instances. Two or three is enough. If you cannot generate two or three specific instances, the friction may not yet be at the threshold where a conversation is the right move.

Second, the impact on you. Not "they are undermining me." That is an interpretation about their intent, plus a description of the impact, fused together in a way that will land as an accusation. Separate them. "When that happens, I lose the chance to be associated with ideas I brought, and I am less likely to bring forward new ones in those settings." This is about you. It is about what changes for you. It is not a claim about them.

Third, the ask. What specific behavioral change would address the impact? "Could we agree that when one of us has surfaced an idea, the other will reference it back to its source in the conversation?" The ask makes the conversation actionable. Without it, the conversation has nowhere to go.

Writing these three things down — privately, with the freedom to phrase them badly the first time and then revise — is the preparation that distinguishes a de-escalation from an escalation. You do this thinking in writing because in conversation, the version that came out of your head in the moment is the version the other person responded to. You only get one shot at the framing they hear first.

Why private writing matters specifically

If you process the friction in a shared document, a group chat, a message to a sympathetic peer, you are doing one of two things. Either you are venting, which feels good and does not change the conversation. Or you are recruiting allies, which raises the stakes and makes the actual conversation harder.

Writing the three items above in a private space — where the audience is only you, and you will read what you wrote a second time before saying any of it out loud — is structurally different. You can phrase the impact in its rawest form and then revise to the version that is true but lands. You can write the worst-case attribution of intent and then notice you are attributing intent and revise to behavior. You can draft the ask, look at it, decide it is not actually what you want, and try again.

This is the work the friction conversation requires. It happens in writing, in private, before the words come out of your mouth. The advice to "have the difficult conversation" assumes you have already done this work. If you have not, the advice is the thing that escalates.

What this looks like in product terms

Open a 1:1 agenda. Name it for the conversation you are preparing for — your 1:1 with your manager where you are going to bring up the peer dynamic, or, if you are going to address the peer directly, the conversation with them. Write the three items above. Sit with the draft for a day or two if you have it. Edit it down. Mark the version you are willing to bring into the conversation.

If you are going to your manager first, the agenda becomes the prep for that 1:1. If you are going to the peer directly, the agenda is your private rehearsal — you read it before the conversation, and you bring the structure but not the words exactly as written. The conversation is human. The preparation is what makes the human conversation go somewhere.

Nothing in the agenda leaves your account. The privacy guarantee matters here in a specific way. If your manager or HR could read the draft, you would not be able to write the early version that includes the raw attribution of intent — and that early version is the one you have to write down to be able to revise it. The privacy is what makes the structured thinking possible.

When the conversation should not happen

One honest note. Not every peer friction is at the threshold where a conversation is the right move. If you cannot identify two or three specific behaviors, you may be working with an accumulated mood rather than a pattern you can address. The right move there is usually different: a conversation with your own manager about the dynamic, or a structural change to how you work with the peer (reducing the surface area of interaction), or sometimes just patience.

The preparation work above also surfaces this. When you sit down to write the three items and you cannot complete the first one with specific behaviors, the writing itself is telling you something. Honor that. The private space is the layer where you find out whether the friction is at the level where direct conversation is the right tool.

How Nela Helps

Use Nela to log your wins, track your challenges, and build a private 1:1 agenda from your own evidence for your next conversation. Your data is owner-only at the database — enforced by Postgres Row-Level Security, not just hidden in the UI — and only you can read it back through the app. Request pilot access.

Further reading

• Stone, D., Patton, B., & Heen, S. (2010). Difficult Conversations. Penguin.
• Edmondson, A. (1999). "Psychological Safety and Learning Behavior in Work Teams." Administrative Science Quarterly.
• Grenny, J., Patterson, K., McMillan, R., Switzler, A., & Gregory, E. (2021). Crucial Conversations. McGraw Hill.