A practical manager's playbook for coaching reps at scale — how to run call reviews, build scorecards, use benchmarks, and layer in real-time coaching that improves calls as they happen.
Works on Zoom, Teams & Google Meet · Mac & Windows · 7-day free trial
Coaching is the highest-leverage thing a sales manager does, and the thing most managers do least well — not from lack of care, but lack of time and data. With six or eight reps, a full pipeline of your own, and no realistic way to listen to every call, coaching collapses into a weekly 1:1 built on the rep's selective memory of how their calls went.
The result is coaching that is vague ("work on your discovery"), infrequent (one conversation a week), and disconnected from what actually happened on calls. Reps don't improve from advice they can't act on, and managers can't fix what they can't see. The whole loop is starved of the one thing it needs: a clear, specific, frequent signal about what to change.
This guide lays out a coaching system that scales: how to run structured call reviews, how to build scorecards that make performance objective, how to use benchmarks to focus effort, and how real-time coaching closes the gap entirely. Then it shows how ConversationPilot gives managers visibility into every call and gives reps live coaching in the moment — so the team improves continuously, not just in the weekly 1:1.
Most call reviews fail because they are unstructured and after-the-fact opinion. A useful review is specific and forward-looking: pick one or two moments from a real call, examine what happened, and agree on one concrete change for next time. "You talked 70% of the time on a discovery call" plus "next call, ask three open questions before you mention the product" beats an hour of general feedback.
Focus on one skill at a time. A rep cannot fix discovery, objection handling, qualification and next steps all at once; pick the highest-impact weakness and work it until it improves, then move on. Reviews that try to cover everything change nothing.
Use real call evidence, not memory. The reason most reviews are weak is that they rely on the rep's recollection, which is selective and self-flattering. Reviewing actual call moments — the exact exchange where the objection was fumbled — makes the feedback undeniable and specific. A call review library that lets you pull the moment that mattered turns coaching from opinion into evidence.
Scorecards turn coaching from subjective to objective. Define what good looks like across the dimensions that matter — discovery completeness, qualification coverage, talk-listen ratio, objection handling, clear next steps — and score calls against it consistently. Now "get better at discovery" becomes "your discovery score is 3/10 because you're not uncovering impact."
The power of a scorecard is comparability. You can see how a rep trends over time, how reps compare to each other, and where the team as a whole is weak. That focuses your limited coaching time on the changes that will move results most, instead of whatever came up in the last 1:1.
Scorecards also align the team on standards. When everyone knows the qualification scorecard requires a confirmed economic buyer, reps self-correct toward it. The scorecard becomes a shared definition of a good call, which is the foundation of any coaching culture. The hard part is scoring consistently at scale — which is exactly where automated scoring on every call changes the math.
Benchmarks tell you where to spend your coaching energy. Team and individual benchmarks — average talk-listen ratio, qualification completeness, objection-handling rate, close rate by stage — reveal the patterns that gut feel misses. Maybe the whole team under-qualifies on authority; maybe one rep's talk time is wrecking their discovery.
With benchmarks you can prioritise. The biggest team-wide gap is usually worth a group session; the individual outliers warrant one-on-one work. Without benchmarks, managers coach whatever feels urgent, which is rarely what would move the number most. Data turns coaching from reactive to strategic.
Leaderboards add a motivational layer when used carefully. Surfacing the reps who run the best discovery, or handle objections most cleanly, makes excellence visible and copyable — and lets you pair a struggling rep with a strong one on a specific skill. The point of benchmarks is not surveillance; it is direction. They answer the manager's hardest question: of everything I could coach, what should I coach first?
Even excellent call reviews, scorecards and benchmarks share one limitation: they all happen after the call. The rep fumbled the objection on Tuesday; you coach it on Friday; the deal is already cooling or gone. The feedback improves the next call but cannot save the one where the mistake happened.
There is also a scale problem. A manager can deeply review maybe two or three calls per rep per week — a tiny sample of the conversations actually happening. Most calls go uncoached entirely, which means most coachable moments pass without any feedback at all. The loop is simply too slow and too narrow to change behaviour fast.
The solution is to move some coaching into the call itself. If a rep gets a nudge to ask the discovery question they were about to skip, the behaviour improves on that call and every future one — and it happens on every call, not just the sampled few. Real-time coaching doesn't replace review; it makes review far more powerful by raising the baseline of every conversation.
ConversationPilot gives managers visibility into every call and gives reps real-time coaching on every call. During the conversation, reps get the next best question, objection responses and qualification prompts on screen in under two seconds — so the coaching happens in the moment it can change the outcome, on every call, not just the ones you review.
For managers, the dashboard, call leaderboards and team benchmarks make performance objective: talk-listen ratios, qualification completeness, objection handling and playbook compliance across the whole team. Every call produces a structured report and a call score, so you can coach against real evidence instead of the rep's memory, and the call review library lets you pull the exact moment that mattered.
It runs as a discreet overlay on Zoom, Teams and Meet, hidden from screen sharing, with no bot in the meeting. AI Playbooks encode your standards so reps run the right call automatically. The combination — live coaching plus full visibility — closes the loop that manual coaching can't: continuous improvement on every conversation, with the data to direct it.
The goal is not a few well-coached calls but a culture where every rep improves continuously. ConversationPilot makes that achievable for a manager who doesn't have time to listen to every call. Live prompts raise the floor on every conversation; scorecards and benchmarks tell you where to focus; the call library makes coaching evidence-based.
New reps ramp faster because the copilot models the right behaviour live while they learn — the next question, the objection response, the qualification prompt are right there on every call. Instead of months of trial and error, they perform closer to your best reps from week one, and your 1:1s focus on refinement rather than basics.
Over a quarter, the compounding is significant. A team that improves a little on every call — better discovery, cleaner qualification, sharper objection handling — outperforms one that improves only in occasional reviews. Coaching stops being a bottleneck constrained by your hours and becomes a system that works on every call, with you steering it where the data says it matters most.
| Capability | ConversationPilot AI | Doing it manually |
|---|---|---|
| Calls coached | Every call, live | 2–3 sampled per rep per week |
| Timing of feedback | In the moment it matters | Days later in a 1:1 |
| Evidence | Scores, transcripts, call library | Rep's selective memory |
| Scorecards | Automatic on every call | Manual, inconsistent |
| Benchmarks & leaderboards | Team-wide, objective | Gut feel |
| New-rep ramp | Live modelling from week one | Slow trial and error |
Run structured call reviews focused on one skill at a time, use scorecards to make performance objective, and use benchmarks to decide what to coach first. Then add real-time coaching so reps improve on every call, not just in the weekly 1:1. ConversationPilot provides all four.
They happen after the fact, rely on the rep's selective memory, and only cover the two or three calls a manager has time to review. Most coachable moments pass without feedback. Real-time coaching plus full-call visibility — what ConversationPilot offers — closes that gap.
They make performance objective and comparable. Instead of "get better at discovery," you can say "your discovery score is 3/10 because you're not uncovering impact." ConversationPilot scores every call automatically across discovery, qualification, talk-listen ratio and objection handling.
Yes. ConversationPilot surfaces the next best question, objection responses and qualification prompts on the rep's screen in under two seconds during the call — so the coaching happens in the moment it can change the outcome, on every call, hidden from screen sharing.
The copilot models the right behaviour live while new reps learn — the next question, the objection response, the qualification prompt appear on every call. They perform closer to your best reps from week one, so your 1:1s focus on refinement instead of basics.
Use benchmarks. ConversationPilot's manager dashboard shows team and individual patterns — talk-listen ratios, qualification completeness, objection handling, playbook compliance — so you spend limited coaching time on the gaps that will move results most, not whatever felt urgent.
Real-time prompts, objection handling and qualification — while the call is happening.