Roleplay vs Call Reviews vs Live Coaching: Which One Actually Fixes a Rep

A frontline manager asked me last month which she should be doing with her worst rep: more call reviews, more live listening, or more roleplay. She had time for one of them. Maybe an hour and a half a week per person, if she protected it, which she mostly didn't.

That is the real question. Not which coaching method is best in the abstract — all three do something — but which one changes behaviour on Tuesday's cold call block when nobody is watching. Because that is the only thing that counts. A rep who can articulate what they did wrong on last week's discovery call and still does it again on the next one has not been coached. They have been informed.

So here is how I think about the three, what each one actually costs, and why roleplay is both the most useful and the one most managers run badly.

The three modes are not interchangeable

Start by being honest about what each one is for. Managers get frustrated with coaching because they use the wrong tool and then blame the rep.

Call reviews build awareness. That's it.

You pull a recording, you listen together, you stop it at the moment the rep talked over the objection instead of into it. The rep sees it. They nod. They often see it before you say anything, which is the tell that awareness was never the bottleneck.

Call reviews are cheap and asynchronous. You can do them at 6am with coffee. You can do them on a plane. You can review a call from a rep in another time zone without booking anything. That flexibility is real and it's why call reviews are the default coaching motion in almost every org I've been inside, including the big ones.

But the gap between hearing yourself fumble a pricing question on a Thursday and handling that question well on the next live call is enormous, and nothing in the call review closes it. You are reviewing a performance that has already happened, in a calm room, with no prospect, no adrenaline, no fear of losing the deal. The rep's failure did not happen because they lacked information. It happened because in the moment, under pressure, their mouth defaulted to the thing it always does.

Call reviews are diagnostics. They tell you what to work on. They are not the work.

The other thing they do well: they show you the pattern across reps. If four people on the team all soften the same way when the buyer says "send me something," you have a team problem, not a rep problem, and you now know what to build a session around. I have found more team-level coaching topics in an hour of listening to recordings than in any pipeline review.

Live listening catches habits and costs the rep dignity

Sitting next to a rep, or dialling into the call muted, or listening in real time from three desks away — this catches things recordings don't. Recordings flatten everything. You don't hear the rep take a breath and decide not to ask the hard question. You don't see them check the clock before the third dial. You don't see them skip the prospect who scares them.

Live listening surfaces the behaviours around the call, not just the call. When I was selling, the moments that most needed fixing were often the ninety seconds before I picked up the phone, not the four minutes after.

But there's a cost nobody says out loud. Being listened to live is humiliating for most people. The rep knows you are there. They perform. They use the language they think you want, they ask the discovery question from the deck that they never ask otherwise, they get stiff, and the call you are observing is not the call they normally run. So you are coaching a fictional version of the rep. Then, because they were stiff, the call goes badly, and now they associate your presence with failure. Do that a few weeks in a row with a rep who is already struggling and you will get avoidance dressed up as calendar conflicts.

Live listening also does not scale. It costs you the exact hour the rep is on the phone, and you can only be in one place. If you manage eight people, you are choosing whose hour matters this week.

I'm not saying don't do it. I'm saying use it for what it's uniquely good at — energy, activity behaviour, the stuff between calls — and use it sparingly enough that your presence doesn't distort the thing you came to see.

Roleplay is the only one that rehearses the moment

Here's the thing that took me too long to accept. Selling is a physical skill. It's not knowledge. A rep can explain the correct response to "we already use a competitor" in perfect detail in your 1:1 and then, on a live call, say something completely different, because the live call has fear in it and the 1:1 doesn't.

The only coaching mode that puts a rep back into that moment, on purpose, repeatedly, before it costs them a deal, is roleplay. It's the flight simulator. You crash on the ground so you don't crash in the air.

Roleplay is also the only one where the rep gets more than one attempt. That is the whole game. Call reviews give you one attempt and a post-mortem. Live listening gives you one attempt and an audience. Roleplay gives you three attempts and no consequences.

So why doesn't everyone do it? Two reasons. Reps hate it, because most roleplay they've experienced was a humiliation ritual in front of the team. And managers do it badly, in a specific and predictable way.

Managers play a nice prospect

This is the failure. You sit down, you say "okay, I'm the CFO, call me," and then you proceed to be the most agreeable CFO in the history of enterprise software. You answer the discovery questions in full sentences. You volunteer your budget cycle. When the rep asks for the meeting, you say "sure, Thursday works."

Every manager does this and I did it too. There are reasons. You want the rep to feel good. You feel awkward being aggressive with someone who reports to you. You are not a good actor and staying in a hostile character is genuinely hard. And underneath it all, you want the session to end on a high note, so you soften.

But a roleplay that the rep wins easily is worse than no roleplay, because it certifies a skill they don't have. You have now both agreed they can handle the CFO. Then they meet a real one, who interrupts on the second sentence, and the rehearsal was worthless.

The standard has to be: the roleplay is harder than the real call. Not slightly harder. Meaningfully harder. If the rep walks out of the session and thinks "that was rough," and then the real prospect next Tuesday feels manageable by comparison, you have done your job. Sparring partners hit harder than the opponent. That's the point of a sparring partner.

Four rules for roleplay that isn't theatre

Pick one moment, not one call

The most common mistake after playing nice is running a whole call. "Take it from the top, do the full discovery." Twenty minutes later you have a sprawling mess with eleven things wrong with it and no idea which one to fix.

Run the moment. Ninety seconds. The moment is a specific place where the rep breaks. Not "discovery" — the moment where the prospect says "just send me some information" and the rep says "sure, what's your email." Not "objection handling" — the moment where the buyer says the price is too high and the rep starts talking about ROI before they've asked what "too high" means.

You find that moment in the call reviews. That's the handoff between the two methods, and it's why the argument about which coaching mode is best is a bit silly — reviews find the moment, roleplay fixes it. But the fixing happens in the rehearsal, not in the review.

Start the roleplay one line before the moment. Not at hello. Give the rep the setup in a sentence — "you're four minutes in, you've just given your value prop, I'm the ops director, go" — and then hit them with it.

Make it hard

Be the prospect who is busy, mildly annoyed, and not interested in helping. Interrupt. Give short answers. Say "why?" to a question that deserves a real answer. Say "we're happy with what we have" without elaborating, and then wait, because waiting is what real buyers do and it is exactly the silence that makes reps panic and fill.

A few things I use. Repeat the objection a second time after the rep has handled it once, because in real life buyers don't accept the first answer and reps have almost never rehearsed the second pass. Go quiet for four or five seconds after the rep asks a good question, so they learn to sit in it instead of rescuing you. Be inconsistent — say budget isn't the issue and then bring up budget again — because real buyers contradict themselves and reps get thrown by it.

What you should not do is be cruel or weird. Hard means realistically difficult, not surreal. Don't hang up. Don't swear at them. The goal is a demanding buyer, not a trauma exercise. If the rep is going red, you have crossed from difficulty into something else and you should stop.

Run it three times

One rep of a roleplay is a test. Three is practice.

First run: let it fail. Don't coach mid-run, don't rescue, let them get to the end of the ninety seconds even if it's ugly. Then give one piece of feedback. One. Not the four things you noticed.

Second run: same setup, same opening line from you, same objection, and they apply the one thing. It will be clunky. That's fine, clunky is what a new motion feels like.

Third run: same again, and now change one variable — you push back harder, or you use a slightly different wording of the same objection — so they're applying the skill rather than reciting the line they just built.

The three runs are non-negotiable because the whole value is in the repetition. A single roleplay is just call review with more embarrassment. If you only have time for one run, do a call review instead and save the roleplay for a week when you have fifteen minutes.

Score one thing

Before you start, say out loud what you are scoring. "I'm listening for whether you ask one question before you answer." Then at the end, that is the only thing you comment on.

This is hard. You will hear five other problems. Write them down for another week and shut up about them today. A rep who receives five corrections applies none. A rep who receives one correction, three times, applies one — and one applied is infinitely more than five discussed.

Say the score out loud too. "Run one, you answered immediately. Run two, you asked but then answered your own question. Run three, you asked and waited. That's the rep I want on Tuesday." Reps need to hear that they got better inside the session, because the improvement is small and they will not notice it themselves.

A weekly cadence a frontline manager can sustain

Here is what I'd actually run, and the constraint is that it has to survive a bad week — a quarter-end, a sick kid, an escalation. If your coaching system only works in calm weeks, you don't have a system.

Monday, before the team is on the phone. Fifteen minutes of listening to recordings, on your own, no rep present. You are not coaching. You are hunting for the moment. One moment per rep, written in a note. If you have eight reps, you won't get to all of them; rotate.

One live listening block, once a week, one rep. Half an hour, sat with them or dialled in, and you tell them in advance so it isn't an ambush. Rotate through the team so each person gets it every month or two. Use it for the between-call behaviour, not the call craft.

Fifteen minutes of roleplay inside every 1:1. This is the load-bearing wall. Not a separate meeting that gets cancelled — inside the 1:1, at the front, before the pipeline chat, because if you put it at the end it never happens. One moment, hard, three runs, one score. Fifteen minutes. Everybody groans for the first month and then it becomes the normal shape of a 1:1.

One team session a month. Take the moment that showed up across several reps and run it as a group. Two people spar, everyone else watches and scores the same single thing. Group roleplay is bad for developing an individual and good for setting a standard, so use it for the standard.

That's it. It fits. What kills coaching cadences is scope, not time — managers design a beautiful programme and abandon it in week three.

When the rep is further gone than a bad habit

Everything above assumes a functioning rep with a specific broken moment. Sometimes that isn't the situation. Sometimes the rep has missed two quarters, the confidence is gone, and the conversation you need to have isn't about the objection handling at all — it's the one where you sit down and both admit the number isn't happening on the current trajectory.

That conversation is itself a skill, and it's one managers get almost no practice at, because you only get a handful of them a year and each one matters enormously. It's worth rehearsing before you walk in. I'd run it the same way you'd have a rep run a cold open — pick the moment, make the other side difficult, three passes. The two-quarter miss 1:1 playbook walks through that conversation as a roleplay you can practise on yourself, and there are versions shaped for the specific pressures of SaaS teams, where the miss usually shows up as pipeline coverage, and for staffing and recruiting teams, where it's more often activity and candidate flow.

Do not go into that meeting cold. You will either soften it into meaninglessness or overcorrect into a threat, and both outcomes cost you the rep.

The honest comparison

Call reviews are the cheapest and the slowest to change anything. Use them to find the moment and to spot team-wide patterns. Live listening is expensive per hour, catches things recordings can't, and distorts what it observes — so use it rarely and tell the rep it's coming. Roleplay is the only one that rehearses the moment under pressure with more than one attempt, which is why it changes behaviour, and it's the one most managers run as a friendly chat instead of a hard rep.

If you only fix one thing after reading this, be a worse prospect. Interrupt more. Answer in three words. Repeat the objection after they've handled it. Let the silence sit. Your rep will find it unpleasant and then find Tuesday easier.

The practical problem with all of this is that fifteen minutes per rep per week is a lot of manager hours, and the quality of the roleplay depends entirely on your acting on a day when you're tired. That's the gap I built DrillCall to sit in — reps run the hard version of the moment as many times as they need without needing a manager in the room, and you spend your fifteen minutes on the scoring conversation rather than on being the CFO. If I were running a team again, that's the order I'd do it in: find the moment in the recordings, let the rep grind the reps, then use my own time for the part only a human can do.

But the sequence matters more than the tooling. Find the moment. Make it hard. Three runs. Score one thing. Everything else is commentary.