Being a group leader involves a lot of people skills, and even those who undergo leadership training face countless challenges to their skillset. Most enter research leadership because they are drawn to tackling interesting questions or problems—not because they want to resolve interpersonal conflicts or give hard feedback. Yet, many PIs receive little training in navigating tough conversations with the people who are so central to the work. Avoiding these conversations typically doesn’t make the issues go away; it usually makes them more entrenched or tougher to address.
Today, we look at a few of the most common difficult conversations research leaders face, and a way of thinking about each one that makes them easier to start.
Challenge 1: Lagging Progress or Low Motivation
Every research group has moments when a project stalls or a student stops making expected progress. The natural instinct is to give it more time, to assume the person knows they need to pick up the pace, or to hope the next set of results will turn things around.
But delay usually makes things worse. A student may not realize how worried you’ve become—or they may know, and feel too ashamed to bring it up. By the time the conversation happens, frustration has built on both sides, and what could have been a straightforward check-in feels like a confrontation.
A different approach is to catch these moments early and frame them as problem-solving rather than evaluation. A conversation that starts with “Let’s look together at what has been working and what hasn’t” differs from “Your progress is not where it needs to be.” The first invites collaboration; the second triggers defensiveness.
The goal isn’t to lower standards but to intervene before a small gap becomes a large one. PIs who handle this well tend to build regular checkpoints into their mentoring rhythm, not to micromanage people, but to create natural opportunities to adjust course before anyone feels like they’re failing.
Challenge 2: Mismatched Fit
Some mismatches can’t be solved with clearer expectations or more frequent check-ins. A student who is very skilled could be working on a project that doesn’t suit their strengths or interests. A postdoc may have realized halfway through their fellowship that their career goals have shifted in a direction the current group can’t support.
These conversations tend to be challenging because they touch on identity and belonging. No one wants to tell a junior colleague that the group may not be the right place for them, and no colleague wants to hear it.
But letting a poor fit drag on helps no one. The trainee stays in a situation that’s draining rather than developing. The group loses momentum. And you’re carrying an unspoken tension that affects every interaction.
The frame that works here is honesty paired with support: “I don’t think this is a failing on your part or mine. I think the match between what you need and what this group can offer isn’t what we hoped it would be. Let’s figure out together what a good next step looks like.”
This conversation is never easy. But it’s far kinder to have it early than to let someone spend months or years in a setting that isn’t serving their development.
Challenge 3: Working Styles
Every PI eventually encounters a situation where a trainee or staff member is great on paper but difficult to work with. Maybe they dominate every group meeting, making it hard for quieter members to speak. Maybe they resist feedback or become defensive when their work is questioned. Maybe they have a pattern of missing deadlines that forces others to scramble.
These are the conversations that most often get avoided. They feel personal, like they’re about someone’s character rather than their work. And it can be hard to pinpoint the specific issue when it feels more like a pattern of behavior that’s slowly eroding group functioning.
Instead of “you talk too much in meetings,” try deploying the situation–behavior–impact model: “When meetings run long because one person speaks more than others, it makes it harder for the rest of the team to contribute“. This type of statement centers the impact on group functioning and enables a productive discussion.
It also helps to remember that you don’t need to resolve everything in one conversation. The first conversation is simply to name the pattern and ask for the other person’s perspective. Often, people aren’t aware of how their behaviors affect others. Once they know, many will adjust without needing to be managed further.
Challenge 4: Someone’s Leaving the Group
Personnel turnover is the norm in research. Postdocs finish their appointments. Students graduate. Staff moves on. But even when departures are expected and positive, the conversations around them can feel awkward or unresolved.
When speaking with the person who is leaving, a useful practice is to separate the logistical conversation from the reflective one. The logistics—timeline, knowledge transfer, and remaining responsibilities—are necessary and straightforward. The reflective conversation asks something different: What worked well during this person’s time with the group? What could have been better? What would they tell a friend who was about to join? Not every departing person will want to have this second conversation. But those who do may share insights that help you improve.
When speaking to remaining team members, consider what they want to know versus what you can and are willing to share about the situation. For example, they may want to know why someone is leaving, and there will be cases where you might not be able to share that information. When possible, be transparent about how the departure affects them directly or indirectly, how knowledge and resources will transfer to others, and whether the role will be filled or not.
Challenge 5: Dynamics with a Peer
Research groups don’t operate in isolation. They share equipment, space, and sometimes people with other groups. From time to time, something goes wrong: a collaboration becomes unbalanced, a shared resource gets mismanaged, a team member reports problematic behavior from a colleague.
These conversations are difficult because the power dynamics differ. The person on the other side of the table is a peer, not someone you are supervising, and the stakes can feel higher if there could be longer-term consequences at risk.
The principle that helps here: separate the problem from the person, and assume good intent to start. “I want to talk about how the shared equipment scheduling has been working. Lately, my team has had trouble getting time during regular hours, and I’m wondering if there’s a mismatch in how we’re both understanding the arrangement” opens a different conversation than “Your group is monopolizing the equipment.”
Most people want to get along with their colleagues; starting from that assumption creates space where problems might actually be discussed. While we can never control how someone else might act or react, we can be attentive with our approach and our own responses.
The Remaining Challenge: Self-reflection
The most important difficult conversation any research leader has is the one they have with themselves to bring awareness to what they might prefer to avoid.
A few questions worth asking oneself a couple of times a year:
- Is there a tension or issue in the group I’ve been hoping will resolve on its own?
- Have I let a pattern develop that I would have addressed immediately if it had started last year instead of gradually over time?
- Am I avoiding a conversation I already know I need to have?
- What have I been putting off, and what is that avoidance costing the group?
These questions aren’t meant to provoke anxiety or constant reinvention. They’re meant to keep you intentional. The PIs who sustain healthy cultures over the long haul tend to be the ones who’ve made honest self-check-ins a habit to ensure that problems aren’t swept under the rug.
Building the Skill
Handling difficult conversations is a skill. And like any skill, it improves with practice and reflection.
- Reframe these conversations as acts of care. Addressing a problem early is kinder than letting it grow. Giving someone honest feedback about fit is more respectful than letting them struggle in silence. Naming a pattern of behavior is more supportive than quietly resenting it. As Brené Brown says, “Clear is kind.”
- Prepare without over-preparing. A few notes about what needs to be said and how to open are useful. A full script usually backfires, because real conversations never go exactly as planned. The goal is to go in clear about the outcome you need, but open to how it unfolds. Consider practicing different possible reactions from the other person so that you are more likely to respond and less likely to react.
- Reflect afterward. What went well? What could have gone better? What will you do differently next time? This reflection creates a learning opportunity that will make the next difficult conversation just that much easier.
The Cost of Avoidance
Thriving research groups are not the ones where small problems were allowed to grow, where feedback was delayed until it felt punitive, or where tensions simmered until someone left in frustration. Avoiding difficult conversations feels safe in the moment because it goes with the flow. But the unresolved issues are an undercurrent that gets stronger in the long run.
On the other hand, a group where the PI models direct, respectful conversation about hard topics becomes a place where people trust each other more, not less. They know where they stand. They know that problems will be addressed rather than ignored. They know that feedback is a sign of investment, not a threat. This fosters a culture of resolution rather than avoidance.
Start Small
The next time you’re facing a conversation that feels too hard to have, consider starting smaller than the whole thing. It’s not necessary to try to solve everything at once. A small step can put you on the path to resolution.
“I’ve been meaning to talk with you about how things are going. Is now a good time, or should we find time next week?”
By taking this small step, you’ll get past the initial discomfort, making the next step a little lighter.