How to Have Difficult Conversations with Veterinary Staff (Without Carrying Them Home)

Most veterinary leaders assume difficult conversations are hard because of who is involved: the team member's defensiveness, the relationship, the discomfort of the moment itself. I carried that assumption for years, and it kept me avoiding conversations I knew I needed to have. What I had labeled as patience had actually been avoidance.

The assumption is understandable. It is also wrong. Difficult conversations drain leaders when they arrive too late, when the expectation was never defined before the moment of correction, and when nothing is structured to carry the accountability forward after the conversation ends. The weight does not come from the conversation. It comes from everything the leader was already carrying before it started, and from the open question of whether it will have to happen again.

That is a structural problem, and structural problems have structural fixes. What follows is the working structure I use and teach: what to do before the conversation, a three-part script for the conversation itself, and the follow-through that decides whether you ever have to have it twice.

Before the conversation: three steps that do most of the work

1. Have it early, while it is still light.

In most veterinary hospitals the threshold for naming a pattern is set far too high. A leader notices something early, files it as something to watch, and weeks pass while the threshold quietly adjusts upward. By the time the conversation happens it is no longer a course correction, it is a confrontation carrying months of accumulated frustration. A five-minute conversation about one observed behavior needs almost no emotional preparation. A thirty-minute intervention about an entrenched pattern costs both people days of recovery. Every week a pattern goes unnamed, the eventual conversation gets more expensive. The kindest version of the conversation is the earliest one.

2. Confirm the standard existed before the correction.

The most common reason a difficult conversation fails is not what gets said in the room. It is that the expectation was never clearly defined before the room. If you cannot point to a standard that was documented, communicated, and applied to everyone, your feedback will land as opinion, and the team member's defensiveness will be rational: they are being corrected against a benchmark they never knew existed. When that is the case, the first conversation is not a correction at all. It is a reset: name the standard, own that it was never made explicit, and start the clock from today.

3. Separate the observation from your interpretation.

Write down the specific behavior you observed, then strip out every attribution of motive. A technician arriving late three times in two weeks is an observation. A technician who does not care about the team is an interpretation. One is the basis for a productive conversation, the other is the basis for a conflict. This step matters most with your strongest people, because with them the interpretation is usually wrong: the behavior is real, but the story you have attached to it rarely is.

During the conversation: behavior, impact, ownership

The script I teach in TRIAGE Module 4 has three parts, delivered in order, without padding and without apology. Behavior. State the specific, observable thing. "The last three overnight handoffs went out without treatment notes attached."

Impact. Connect it to the system, not to your feelings. "The day team is rebuilding the plan from scratch each morning, and yesterday that delayed two rechecks."

Ownership. Hand the responsibility over explicitly, with a clear next step. "I need the notes attached before you leave the building. Can you own that from tonight?" That is the whole shape. Notice what is missing: no compliment sandwich, no "how do you think it's going," no fifteen minutes of context-setting to soften the landing. Padding does not reduce discomfort, it smuggles in ambiguity, and ambiguity is what makes these conversations feel personal. Defensiveness rises when identity feels attacked, and it falls when responsibility is clear. Clarity is not cruelty. In practice it is the most respectful thing you can offer, because it treats the other person as someone capable of meeting a standard rather than someone who needs to be managed around.

Two things protect the conversation while it happens. First, regulate before you enter the room: if you are having the conversation while irritated, the team member will respond to your irritation and not your standard, and what you call accountability will just be a reaction finding its way out. Second, let them talk. The goal is not to win the exchange. It is to leave the room with a shared, specific agreement about what changes and by when.

After the conversation: close the loop

Most leadership credibility is not lost in the first correction. It is lost afterward, when nothing visibly follows. Three follow-through steps decide whether the conversation worked.

Summarize the agreement the same day.

Two sentences, written, to the person: what was agreed, and when you will check in. This is not documentation-as-threat. It is the difference between a conversation and a standard.

Check in when you said you would.

If the behavior changed, say so plainly. Correction without acknowledgment teaches people that the only feedback they will ever get is negative, and that is how you turn a fixable pattern into a resignation.

If the pattern returns, change the conversation.

A repeated pattern after a clear agreement is no longer a feedback problem, and repeating the identical conversation teaches the team that your standards are suggestions. The second conversation names the agreement, names the recurrence, and introduces a consequence with a timeline. Standards are rarely tested by first mistakes. They are tested by second chances.

Why this is a retention issue, not a comfort issue

There is now direct evidence for how much of this lands on the leader's shoulders. A 2026 study of 367 veterinary professionals in Veterinary Record found that psychologically safe teams report lower intention to leave, and that the link runs through the supervisor's feedback skill rather than through co-worker support. Retention follows the leader's trainable skills. Gallup's performance-management research points the same direction: only 26% of employees say performance management motivates them to do outstanding work, which describes environments where accountability is inconsistent and expectations are vague, not environments with too many hard conversations.

Avoidance does not remove discomfort. It redistributes it, usually onto your most reliable people, who absorb the extra load while the pattern continues. The structure above is how you stop asking them to.

The conversation you are avoiding right now is the lightest it will ever be.

Written by Dr. Kaelyn Petras, DVM
Founder of PIVOT Vet Strategies and Emergency Medical Director with experience across emergency, specialty, hospital leadership, intern training, and veterinary leadership systems design.