Roleplay vs Call Review: Which One Actually Changes What a Rep Does on Monday

The comfortable default

If you ask a sales manager how they coach, most of them will describe call review. They pull a recording, they scrub to a timestamp, they say "listen to what you did here," the rep nods, they book the next one for two weeks out. Everyone leaves feeling like work happened.

I have done this. I have done a lot of it. And for a long time I thought I was coaching, right up until I noticed the same rep making the same mistake on a call I reviewed a month later. Not a related mistake. The same one, in the same spot, with almost the same words.

Call review is comfortable for a manager for three reasons. It is evidence-based, so nobody can argue with you. It requires no performance from you — you sit next to the rep and point at a screen. And it fits neatly into a calendar block. Roleplay is the opposite on all three counts. There is no recording to hide behind, you have to actually play the buyer, and it feels stupid for the first ninety seconds every single time.

Reps hate roleplay for a simpler reason. It is exposing. In a call review, the bad moment already happened and is safely in the past — you are both looking at a corpse. In roleplay, the rep has to fail live, in front of their boss, in real time, and then do it again. That is a genuinely unpleasant thing to ask someone to do, and pretending it isn't is why so many managers get quiet compliance instead of effort.

So the honest comparison is not "which one is better." It is: call review is better at diagnosis, roleplay is the only thing that reliably changes behaviour, and doing either one alone is why your coaching doesn't survive the weekend.

What call review is genuinely good at

Diagnosis. That is the whole list, and it is a big one.

A rep cannot hear themselves on a live call. They are managing their own nerves, tracking what the buyer said, looking at a CRM tab, and trying to remember the next question. There is no spare processing power to notice that they said "just wanted to reach out" or that they answered a pricing question the buyer had not asked yet. Recording removes that constraint. On playback, the rep hears what the buyer heard, which is often nothing like what the rep thought they were doing.

The second thing call review does well is settle arguments. When I sold inside AWS and Dell, the single fastest way to end a debate about why a deal stalled was to go to the tape. Reps have a story about their calls. The story is usually flattering and usually wrong — not because reps are dishonest, but because memory edits. You remember the good question you asked. You do not remember the eleven seconds of silence after the buyer said "we're pretty happy with what we have."

Third, call review finds patterns across calls that no single call reveals. One rep I worked with was convinced her problem was closing. It wasn't. Across five recordings, she opened every discovery call by summarising what she already knew about the account, which took ninety seconds and trained the buyer to sit back and be talked at. The close was fine. The first ninety seconds had already set the shape of the conversation.

What call review is bad at is the part everyone assumes it does: changing what the rep does next.

Here is the mechanism, and once you see it you cannot unsee it. Call review produces understanding. Understanding lives in a different part of the brain than performance. A rep can explain to you, in precise and articulate detail, exactly what they should have said when the buyer pushed back on price — and then, on Tuesday afternoon, on a live call, with adrenaline up and a buyer talking over them, they will say the old thing again. Not because they forgot. Because the old thing is what comes out under pressure, and nothing about sitting in a conference room analysing a recording changed what comes out under pressure.

Every rep I have watched has this gap. It is not a knowledge problem. Knowing what to say and being able to say it while a stranger is being difficult at you are two different skills, and only one of them gets built by listening.

What roleplay is genuinely good at

Roleplay builds the reflex. That is also the whole list, and it is also a big one.

What happens on a real call is not deliberation. It is retrieval. The buyer says "send me some information" and the rep does not weigh options — a response fires. Whatever response has fired most often in the past is the one that fires now. If a rep has said "sure, what's your email" two hundred times, that is what will come out on call two hundred and one, regardless of how many times they have nodded along to a coaching session about it.

Roleplay is the only tool I know that changes which response fires. Not because it is magic, but because it is the only one that makes the rep produce the words out loud, under a small amount of pressure, more than once. That is the entire mechanism. Say the new thing enough times in a mildly uncomfortable setting and it becomes the thing that comes out when the setting is genuinely uncomfortable.

The part managers get wrong is treating roleplay as a test. A test is one attempt, graded. That builds nothing. The rep performs their best version once, you say "good, much better," and you both go home. Repetition is not the packaging around the coaching. Repetition is the coaching.

Roleplay also does something call review structurally cannot: it lets you run the situations that matter but rarely happen. The economic buyer who joins twenty minutes late and asks a hostile question. The champion who goes silent after a great demo. The prospect who says the incumbent just cut their price. You cannot review a call that has not happened yet, and the highest-stakes moments in a sales cycle are exactly the ones a rep has had the fewest live attempts at.

Why one without the other fails

Call review alone produces a rep who can describe their own weaknesses fluently and has not fixed any of them. I have met a lot of these. They are often the smartest reps on the team and it is genuinely painful to watch, because they know. They can tell you the exact moment the call went sideways. They just cannot not do it.

Roleplay alone produces a rep who gets very good at handling the objection their manager finds interesting, which is frequently not the objection that is actually costing them deals. Managers have hobbyhorses. Mine was discovery depth for years, and I ran roleplay after roleplay on discovery with a rep whose real problem was that he never got a next step booked. He got sharper at discovery. His pipeline stayed exactly as leaky as it had been, because I had been drilling the wrong moment.

That is the whole argument for pairing them. Review is the instrument that tells you where to point. Roleplay is the thing that fires. Pointing without firing is analysis. Firing without pointing is exercise.

The ratio that works

Here is the shape I would run, and the ratio matters more than the total time.

Spend a short block on review and a long block on roleplay. Not half and half. Review is diagnosis, and diagnosis should be quick, because you are not trying to catalogue every flaw in the call — you are trying to find the one moment to drill.

Fifteen minutes of review, one moment

Pick one recording. Not five. Listen with the rep and stop at the first moment where the conversation changed shape for the worse — where the buyer went from engaged to polite, or from open to guarded, or where the rep started talking and did not stop.

Then stop reviewing. This is the discipline that most managers cannot hold. You will hear four other things worth fixing in that same call and you will want to mention all of them. Do not. A rep who leaves a 1:1 with four things to work on works on zero of them. A rep who leaves with one thing has a chance.

Use the review to establish two facts: this is the moment, and here is what the buyer actually did in response to what you said. Do not explain the fix yet. Ask the rep what they would say instead. Whatever they come up with is your raw material.

Forty minutes of roleplay, one moment, eight times

Now run that exact moment. Same objection, same buyer persona, same point in the call. Over and over.

Not a full call. Not "let's roleplay a discovery call from the top." That is theatre, and it dilutes the reps you get on the moment that matters. If the broken moment is the buyer saying "we already have a vendor for this," then you set up two sentences of context and go straight to the line. The rep responds. You respond as the buyer. Thirty seconds, maybe sixty. Stop. Reset. Go again.

Eight times is roughly what it takes before something loosens. The first two attempts are the rep performing what they think you want. Attempts three through five are usually worse than attempt one, which is normal and worth warning them about in advance, because they are actively trying to drop an old habit and have not built the new one yet. Somewhere after that, the response starts to sound like a person talking instead of a rep executing.

The manager coaching playbooks I put together for running the 1:1 after two missed quarters are built on exactly this shape — a short diagnostic conversation followed by repeated live reps on one specific moment — because the hard part of that conversation is never the diagnosis. Both people usually know what is wrong. The hard part is the rep leaving able to do something different.

How to run roleplay without the theatre

Most roleplay is bad, and reps are right to dread it. Here is what makes it bad and how to fix each thing.

Bad roleplay changes the scenario every time. The manager, bored, invents a new objection on rep four. That resets the rep to zero and turns the session into improv practice. Keep the buyer identical. Same person, same company, same objection, same tone. The rep should be the only variable in the room.

Bad roleplay has a manager who plays the buyer as either a pushover or a monster. Pushover buyers teach nothing. Monster buyers teach the rep that the exercise is a hazing ritual. Play the buyer as mildly disengaged and slightly impatient, which is what most real buyers are. If the rep says something good, respond the way a real person would — a bit more open, not delighted.

Bad roleplay stops to give feedback after every attempt. Long feedback between reps means the rep spends more time hearing your voice than using theirs. Keep it to one sentence. "Too long, cut the first half." "You apologised again." "Good — same again." Then go.

Bad roleplay ends when it gets good. The moment the rep nails it is the moment most managers stop, because it feels like a natural ending and everyone gets to feel good. That is the worst possible place to stop. The rep has produced the right response once, under favourable conditions, with no adrenaline. Do it three more times after it gets good. That is where the reflex actually forms.

Bad roleplay is not written down. At the end, the rep should say the version they landed on out loud one final time and write it in their own words. Not your words. Theirs. A phrase a rep would never naturally say will not survive a live call.

The signal you are looking for: it gets shorter

Here is the single best tell that roleplay is working, and it is almost universal in my experience. Attempt one is long. The rep hedges, explains, justifies, adds a caveat, apologises implicitly, and finishes with a question that gives the buyer an easy way out.

By attempt six or seven, if the drilling is working, the response is half the length. The hedging is gone. There is one clear sentence and then silence, and the rep is comfortable letting the silence sit.

Length is a proxy for confidence. Reps pad when they are not sure the thing they are saying will hold. So do not coach the words directly — coach the length, and the words fix themselves. "Say that again in one sentence" is the most useful instruction I know in a roleplay session.

This is also the thing that transfers to Monday. On a live call the rep will not remember your framework. They will remember that this objection gets a short answer, and short is a much easier thing to retrieve under pressure than a five-part structure.

What to ask for if your manager only does one

If you are a rep and your manager only runs call review, you do not need to change their process. You need to change the last ten minutes of the meeting. Say this: "That makes sense. Can we run that moment three or four times before I go, so I actually say it out loud?"

Almost no manager says no to that. It is a small ask, it flatters their diagnosis, and it costs them ten minutes. Then hold them to the rules above — same objection each time, one sentence of feedback, do not stop when it gets good.

If your manager only runs roleplay and never reviews recordings, ask the opposite question: "Before we drill this, can we listen to where it actually went wrong on a real call?" You are protecting yourself from being drilled on your manager's hobbyhorse instead of your actual problem.

And if you get neither — which happens more often than anyone admits, especially in industries where managers carry their own number — you can run the review half yourself. Listen to your own recordings on your commute with one question in mind: where did the buyer's tone change? That is your moment. Then find a colleague and drill it. A peer playing the buyer badly is worth more than a manager who never gets around to booking the session. I wrote the industry versions of the coaching 1:1 — for healthcare teams, real estate and professional services — partly because the objections differ enough that generic drills do not transfer, and partly because reps in those markets are often coaching themselves.

What I would do next

Pick one recording this week. Find the moment the buyer went from engaged to polite. Write down the objection in the buyer's exact words, then say your response out loud eight times until it gets shorter. That is the entire method, and it does not require a manager, a platform or a budget.

If you want the reps without needing a colleague free at the right time, that is why I built DrillCall — an AI buyer that holds the same objection, in the same tone, for as many attempts as you want, and does not get bored on rep five the way a human partner does. Use it for the drilling half. Keep doing the diagnosis half with a real person who has heard your calls, because a machine cannot tell you which moment matters yet.

The uncomfortable truth in the roleplay versus call review debate is that the tool everyone prefers is the one that changes the least. Review is where you find out what is broken. Roleplay is where you fix it. If you only have time for one this week, you already know which one you would pick — and you should probably pick the other one.