Every sales manager has been there. A rep fumbles a discovery call. You pull up the recording, add a timestamp comment at the exact moment it goes wrong, and send it over with a note: "Watch 14:32 — this is where you lost them."
The rep watches it. Nods. Says they get it.
Two weeks later, the same thing happens on a different call.
The Recording Trap
Call recording platforms — Gong, Chorus, Modcloth — are genuinely useful tools. They surface patterns across hundreds of calls, flag deal risk, and give managers visibility they didn't have before. No argument there.
But somewhere along the way, "we have recordings" became synonymous with "we have training." And that's the trap.
Reviewing a call recording is a diagnostic tool, not a development tool. It tells you what went wrong. It doesn't build the reflex to do it differently next time.
What the Science Actually Says
In the 1990s, psychologist Anders Ericsson spent years studying elite performers — chess grandmasters, concert pianists, Olympic athletes. His conclusion, which became the basis for the "10,000 hours" idea, was more specific than the popular version suggests.
It wasn't just hours that mattered. It was deliberate practice: structured repetition in a controlled environment, with immediate feedback, focused specifically on the areas where performance breaks down.
Watching a pianist perform doesn't make you a pianist. Watching game tape doesn't make you an athlete. Reading about objection handling doesn't make a rep better at it.
Skill lives in the muscles — or in this case, in the neural pathways that fire when someone says "your price is too high" and you have three seconds to respond.
The Gap Between Knowing and Doing
There's a specific failure mode in sales training that every manager recognises but few can name: the rep who knows the right answer.
Ask them after a call what they should have said differently, and they'll tell you exactly. They understand the framework. They've read the book. They watched the recording.
But in the moment — when the prospect pushed back, when the silence stretched out, when the conversation took an unexpected turn — the trained response didn't fire. The old pattern did.
This isn't a knowledge problem. It's a practice deficit.
Knowledge is acquired by watching and listening. Skill is acquired by doing — repeatedly, in conditions that approximate the real thing, with feedback that arrives quickly enough to reshape the behaviour.
The Analogy That Sticks
Think about how you learned to drive.
At some point, someone sat in the passenger seat and told you about the clutch. Maybe you watched a video. Maybe you read the manual. None of that was the moment you learned to drive.
The moment you learned to drive was the first time you stalled the car three times in a row on a hill, and then — on the fourth attempt — you got it. Your foot found the friction point. Your body remembered it.
The same principle applies to every conversation skill your team needs: handling a pricing objection, de-escalating an angry customer, delivering feedback that lands without damaging the relationship.
These aren't skills you can learn by watching someone else do them. They require repetition in a context where getting it wrong has low enough stakes that you can actually afford to fail.
The Real Cost of Skipping Practice
The default training stack for most teams looks like this: onboarding deck, product certification, call shadowing, recorded call review. Some teams add manager role-play — once a quarter, if everyone's calendar aligns.
The problem isn't that these things are bad. The problem is that they're all passive. The rep receives information and observes performance. They rarely produce it, under pressure, in a realistic context.
The consequences compound quietly:
- New hires take longer to ramp because their first live calls are also their first real practice attempts.
- Experienced reps fall back on old habits because new techniques never get enough repetitions to become automatic.
- Managers spend more time on individual coaching because team-wide skill doesn't improve.
What Practice Actually Looks Like
Deliberate practice has three components Ericsson identified as non-negotiable:
1. Realistic conditions. The practice environment needs to approximate the real thing closely enough that the skills transfer. Reading a list of objections and their responses isn't practice. Saying the words out loud, in response to a real push, is practice.
2. Immediate feedback. The gap between action and feedback determines how quickly skill develops. A manager reviewing a call two days later is useful. An AI that scores your response immediately — before you've forgotten the moment — is faster.
3. Focused repetition. You can't practice everything at once. The rep who needs to improve their discovery questions needs to run discovery conversations — not full call simulations. Constraint accelerates skill development.
Where This Leaves Your Team
None of this means recordings are useless or training decks are a waste. They're diagnostic and foundational, respectively. You need them.
But if practice — real, repetitive, feedback-rich practice in realistic conditions — isn't part of your training stack, you're optimising for knowledge acquisition in a job that requires skill.
The recordings tell you where the gap is. Practice is what closes it.
Decibl is built for the practice layer — AI-powered voice roleplay that gives your team realistic reps before the real thing.