How do you have a hard conversation with an underperforming rep in 2027?

What should you prepare before the conversation even starts?

Preparation is where most hard conversations are won or lost, and in 2027 preparation means assembling a defensible, shared picture of performance rather than a gut feeling. Before you book the meeting, pull the rep's actual numbers against the target: quota attainment over the last two to three quarters, activity metrics (meetings booked, pipeline created, conversion by stage), and the trend line rather than a single bad month. A single soft quarter is a data point; a three-quarter slide is a pattern, and you owe the rep the honesty of naming which one you are seeing. Write down two or three concrete, recent examples — a deal that stalled because discovery was skipped, a forecast that was wrong two quarters running — because specifics are what make feedback actionable and vague impressions are what make it feel personal.

Equally important is preparing your own hypothesis about the cause, held loosely. Underperformance has many roots: skill gaps, territory or product-fit problems, personal life disruption, unclear expectations, or simple motivation loss. If you walk in certain it is a motivation problem when it is actually a broken territory, you will prescribe the wrong fix and lose the rep's trust. Good managers pre-mortem the conversation by asking themselves what they might be missing and what role management itself played — did onboarding cover this, were expectations ever written down, did the rep inherit a depleted patch. For a deeper framework on separating capability gaps from environmental ones, see the diagnostic ladder at https://pulserevops.com/knowledge/coaching-diagnostics. Arriving with humility about cause, paired with rigor about the facts, is what lets the conversation be collaborative instead of adversarial.

Preparation also includes the logistics people tend to skip. Choose the timing deliberately: never open a hard conversation on a Friday afternoon when the rep will stew over the weekend with no channel to respond, and never wedge it between two other meetings so it feels rushed and transactional. Give it a real thirty-to-sixty-minute block, on camera if remote, in a private room if in office, with your notifications silenced so the rep has your full attention and reads that attention as respect. Decide in advance what a good outcome looks like — usually a shared understanding of the gap and an agreed first step, not a signed confession — because a manager who walks in wanting to "win" the conversation has already lost the coaching relationship. And rehearse the opening line out loud once. The first two sentences carry most of the emotional weight, and managers who improvise them tend to either over-cushion into vagueness or over-blunt into attack. A prepared, plain opening lets the rest of the conversation breathe.

One last piece of preparation is checking your own state before you enter the room. If you are frustrated, behind on your own number, or carrying resentment about the extra work this rep's shortfall has created for the team, the rep will feel it in your tone no matter how careful your words are. The conversation is a coaching act, not a venting act, and the two cannot happen in the same session. If you are activated, delay a day and cool down. Preparation, in the end, is as much about managing your own posture as it is about assembling the rep's numbers.

How do you actually open and structure the conversation?

Open fast and open clear. The single most common mistake is burying the point under five minutes of small talk and vague warmup, which only heightens the rep's anxiety and signals that you are uncomfortable. State the purpose in the first two minutes: "I want to talk about your numbers this quarter and figure out together what is going on and what we do next." Then lead with the specific gap and the data behind it, delivered as an observation rather than an accusation — "You closed 40 percent of quota the last two quarters, and pipeline creation is down about half from your own baseline" lands very differently than "Your performance has been unacceptable." The first invites problem-solving; the second invites defense.

After you name the gap, stop talking and ask. The ratio that separates coaching from lecturing is roughly listen-two-thirds, talk-one-third. Ask open questions — "What is your read on why the numbers moved?" and "Where do you feel stuck?" — and then genuinely listen, because the rep almost always holds information you do not: a competitor undercutting deals, a product gap, burnout, a territory that dried up. The silence after your question is the most productive part of the meeting, and it is the part managers rush to fill. Let it sit. A rep who fills three seconds of quiet with the real reason for the slump has just handed you the actual problem to solve, and you would have talked right over it. The structure below is a reliable arc for the meeting itself.

The closing third of the conversation shifts from diagnosis to path forward. This is where you and the rep co-author — not dictate — a plan with two or three specific, measurable actions and dates attached. Ownership matters here: a plan the rep helps build is a plan the rep will work, while a plan handed down is a plan the rep will resent and quietly ignore. End by confirming the rep's understanding in their own words and by making explicit what support you will provide, whether that is joining calls, unblocking a resource, or adjusting a territory. For the templated plan structure many RevOps teams use, see https://pulserevops.com/knowledge/performance-plan-template.

A word on tone throughout the arc: match your intensity to the stage of the problem, not to your own stress level. An early, first-pattern conversation should feel like a curious partner leaning in, not a supervisor reading charges. Reserve the graver register for genuinely late-stage situations where coaching has already failed and a formal plan is on the table. Escalating too hard too early is the fastest way to convert a fixable slump into a resignation, because the rep concludes you have already decided the outcome and starts looking for the exit rather than the fix. Calibrating that intensity is a skill of its own, and it is why the same words can heal one conversation and detonate another.

How do you use data and AI signals without weaponizing them?

By 2027 most sales orgs run AI copilots that score deal health, flag at-risk pipeline, and summarize call sentiment. Used well, these tools make hard conversations more fair and less personal, because you are both looking at the same dashboard rather than trading anecdotes. Used badly, they become a cudgel — a manager who reads out fourteen red flags from an algorithm the rep does not understand or trust will trigger pure defensiveness. The discipline is to treat AI output as a starting question, not a closing verdict: "The system is flagging that discovery calls are running short — does that match what you are seeing?" invites the rep into the data instead of using it against them.

Transparency about how the signals are generated is what preserves trust. If a rep believes the activity score is a black box that can end their career, they will game the metric rather than improve the underlying behavior, and you will have optimized for dashboard theater instead of real selling. Show your work: explain which inputs drive the score, acknowledge where the model is imperfect, and invite the rep to challenge readings that feel wrong. The map below shows how the same signal can flow toward either trust or defensiveness depending on how a manager frames it.

There is also a fairness dimension that RevOps leaders increasingly own. Before you hold a rep accountable to an AI-derived number, confirm the model is not penalizing a legitimately different but effective sales motion, and that the territory and quota inputs were fair to begin with. Holding someone to a biased or broken metric is not accountability, it is negligence dressed up as rigor. When the underlying data is sound and transparently shared, it does the hardest emotional work of the conversation for you, because the facts are no longer in dispute and both people can spend their energy on solutions.

It helps to distinguish leading signals from lagging ones when you bring data to the table. Quota attainment is a lagging outcome — by the time it is red, the causes are months old. Activity and behavior signals, like discovery-call depth, multithreading on accounts, or follow-up latency, are leading indicators you can still influence. A skilled 2027 manager leads the conversation with the leading signals precisely because they are coachable, and treats the lagging number as the context rather than the club. "Your close rate is down" invites despair; "your deals are single-threaded and that is where the close rate is leaking" invites a plan. Same data, two different conversations, and only one of them the rep can act on tomorrow morning.

Finally, remember that AI is a preparation tool and a fairness tool, never a delivery tool. Let the copilot draft the deal-risk summary, surface the trend you would have missed, and even suggest the coaching angle — then close the laptop and have the conversation as a human. A rep can tell instantly when they are being managed by an algorithm's talking points versus by a manager who has actually thought about their situation, and the former reliably produces compliance without commitment. The signals get you to the right question faster; only you can ask it in a way that makes the rep want to change.

How do you separate the person from the performance?

The emotional core of a hard conversation is making the rep feel that you are critiquing a set of results and behaviors, not attacking their worth as a person. Language does most of this work. Talk about the gap between current results and the target — an external, fixable thing — rather than labeling the rep as lazy, weak, or not cut out for the job, which frames the problem as an unchangeable trait. "The pipeline is not where it needs to be and here is what I am seeing" keeps the problem on the table between you; "You are just not a hunter" puts the problem inside the rep where nothing can be done about it. The former opens a path, the latter closes one.

Belief is the other half of separation, and it has to be sincere. High-performing managers pair candor about the gap with genuine confidence that the rep can close it, because a rep who believes their manager has already written them off has no reason to fight. That does not mean false reassurance — you should never promise an outcome you cannot control or soften the numbers into meaninglessness. It means being clear that the reason you are investing a hard, honest conversation is that you think the rep is worth it. When you must deliver a consequence, such as a formal performance plan, you can still do it while affirming the person: the plan exists because you want them to succeed here, and you are willing to spend your time to get them there. That combination of high standards and high support is what the research on managerial effectiveness consistently identifies as the pattern that actually turns performance around.

There is a practical grammar to separation that managers can drill until it becomes reflex. Attack the process, not the character: "the discovery step is getting skipped" instead of "you're sloppy." Attach behavior to outcome so the fix is obvious: "when the demo comes before qualification, the deal stalls" instead of "your deals stall." And describe, do not diagnose the personality: report what the data and calls show, then ask the rep to supply the why, rather than assigning them a motive they will instantly reject. Most defensiveness in these conversations is not the rep resisting the truth of the number; it is the rep resisting a character verdict smuggled in alongside the number. Strip the verdict out and the same feedback becomes something a person can actually hold and work with.

Separation also protects you from the halo-and-horns trap, where one strong or weak quarter colors your read of everything else the rep does. A rep in a slump is still, in most cases, the same competent professional you hired, temporarily off their game for a reason you have not fully identified. Treating them as a fundamentally broken performer will produce a self-fulfilling prophecy; treating them as a capable person facing a specific, solvable obstacle tends to summon the capable person back to the table. The stance is not soft — you are being ruthlessly clear about the gap — but it is grounded in the assumption of competence, and reps can feel the difference in the first thirty seconds.

What do you do after the conversation ends?

The conversation is not finished when the meeting ends — arguably the most consequential part is the follow-through, and it is where distracted managers most often fail. Within 24 hours, send a written recap: the gap you discussed, the root causes you surfaced together, the specific actions each of you committed to, and the dates you will check in. This document protects everyone. It gives the rep an unambiguous reference so there is no drift about what was agreed, it creates the paper trail HR and legal will expect if the situation escalates to termination, and it forces you to be precise about your own commitments rather than letting them evaporate.

Then you actually hold the checkpoints. A performance conversation with no follow-up teaches the rep that the talk was theater and the standard is negotiable, which is worse than never having the conversation at all. Show up to the checkpoints you scheduled, acknowledge progress specifically when it happens, and escalate honestly when it does not. If the rep improves, close the loop explicitly so they know they have climbed out — nothing kills motivation faster than fixing the problem and still feeling under a cloud. If they do not improve despite a fair plan and real support, the written trail you built means the harder decision that follows is defensible, documented, and free of surprise. Either way, consistency between what you said in the room and what you do in the weeks after is the entire game.

Structure the follow-up cadence so momentum is visible rather than vague. Tighter is better early: a light-touch check within the first week to remove any blocker before it festers, then a substantive review at the two-week and thirty-day marks against the specific actions you both named. Keep the checkpoints on the calendar as standing appointments, not as things you will schedule "when there's time," because the rep reads a cancelled checkpoint as evidence that the plan was never real. At each one, look at the same leading signals you framed in the original conversation, so progress is measured on behaviors the rep can control rather than on a lagging number that may not have moved yet even though the work has genuinely improved. Recognizing that early behavior change before the results catch up is one of the most motivating things a manager can do, because it tells the struggling rep that their effort is being seen.

Finally, run a short debrief on yourself after the dust settles. Did the rep engage or shut down, and what in your framing produced that? Was the root cause you eventually found something your management could have surfaced earlier through better one-on-ones or clearer expectations? The best RevOps leaders treat every hard conversation as data about their own system — if the same failure pattern keeps appearing across different reps, the problem is rarely the reps and usually the enablement, the territory design, the comp plan, or the hiring profile. Fixing the system upstream is what turns the hard conversation from a recurring event into a rare one, which is the real goal. A manager who has to have this talk constantly has a process problem wearing a people costume.

Related questions

What is the difference between coaching and a performance improvement plan?

Coaching is ongoing, informal skill development for any rep; a PIP is a formal, documented process with defined timelines and consequences, usually triggered after coaching has not closed a persistent gap. Coaching assumes growth; a PIP assumes the job is at risk without change.

How long should you wait before addressing underperformance?

Address a genuine pattern early, once you see two to three periods of decline rather than one soft month. Waiting too long is unfair to the rep, who loses time to correct, and to the team carrying the shortfall.

Should hard conversations happen over video or in person?

In person is best when feasible because tone and body language soften hard news, but a well-run video call with camera on beats a delayed in-person meeting. Never deliver serious performance feedback by chat or email.

How do you keep the rest of the team motivated during this?

Keep the specific conversation confidential and hold the same standard visibly for everyone. Teams stay motivated when accountability feels consistent and fair, and they lose faith when they sense favoritism or when underperformance is ignored.

What if the rep gets emotional or defensive?

Slow down, acknowledge the emotion without abandoning the point, and return to shared data. Give them room to respond, but keep the conversation anchored to specific results and the forward plan rather than letting it become an argument about character.

FAQ

How do I start a hard conversation without making it awkward? State the purpose in the first two minutes, lead with a specific data point, and frame it as solving a problem together. Awkwardness comes from vagueness and delay, so clarity is actually the kinder choice.

What if the underperformance is caused by something outside the rep's control? Then the plan should address the real cause — a broken territory, product gap, or unclear expectations — rather than pressuring the rep for effort. Fixing the environment is often the fastest path to recovered numbers, and misdiagnosing it destroys trust.

How much should I talk versus listen? Aim for roughly one-third talking and two-thirds listening. Name the gap clearly, then ask open questions and let the rep surface information you do not have about deals, competitors, or personal factors.

Do I need to involve HR? Loop in HR before you move to a formal performance plan or if there is any risk the situation ends in termination. For an early coaching conversation, HR is usually not required, but documentation still matters.

How do I document the conversation properly? Send a written recap within 24 hours covering the gap, agreed actions, owners, and checkpoint dates. Keep it factual and specific, store it consistently, and use the same format for every rep to keep the process fair.

What if the rep does not improve after the plan? If you provided a fair plan, real support, and clear checkpoints and the gap persists, escalate to a formal process or exit with a clean documented trail. Follow-through on consequences is what makes your standards credible for the whole team.

Can AI tools replace the manager in these conversations? No. AI can surface signals, draft recaps, and make the data fairer, but the judgment, empathy, and trust required to change a person's behavior are irreducibly human. Use AI to prepare, never to deliver the message.

How often should I have performance check-ins to avoid surprise conversations? Weekly one-on-ones and regular pipeline reviews mean underperformance surfaces continuously, so the hard conversation is a natural escalation rather than an ambush. Ongoing dialogue is the single best way to make the rare formal talk less painful.

What if I am the reason the rep is underperforming? Own it plainly in the conversation, because a rep who hears a manager take responsibility for unclear expectations or a bad territory will trust everything else you say. Then fix your part of the plan as visibly as you ask them to fix theirs.

How do I stay calm if the rep pushes back hard on the data? Treat the pushback as information, not insubordination, and ask them to walk you through their view of the numbers. Often the challenge reveals a flawed input or a real obstacle, and even when it does not, hearing them out lowers the temperature enough to keep the plan on track.

Sources

• Harvard Business Review — How to Give Feedback People Can Actually Use
• Gallup — How to Manage a Struggling Employee
• Radical Candor — Kim Scott on Caring Personally and Challenging Directly
• SHRM — Managing for Performance and Documentation Best Practices
• Crucial Conversations — VitalSmarts Framework for High-Stakes Talks
• Sales Management Association — Coaching Effectiveness Research
• MIT Sloan Management Review — Managing Performance in Hybrid Teams
• Center for Creative Leadership — The Feedback That Works Model

Related on PULSE

• Coaching diagnostics: capability versus environment
• The performance plan template RevOps teams actually use
• Running one-on-ones that prevent surprises
• How AI deal-scoring changes the manager role