The Sales Manager's Complete Guide to Call Scoring

Call scoring is one of the most underused tools in sales management. Most managers know that some calls are better than others. Very few have a consistent system for measuring which ones, why, and what to do about it. The result is coaching that depends on memory and opinion, feedback that reps push back on, and improvement that is slower than it should be.

A well-designed call scoring system solves this. It makes what “good” looks like concrete, gives every rep a comparable baseline, and turns 1:1s from opinion sessions into evidence sessions. This guide covers how to build one, calibrate it, and use it without demoralising your team.

What call scoring actually tells you (and what it doesn’t)

Call scoring tells you how a rep behaved on a specific call, measured against a defined set of criteria. It does not tell you whether the outcome was right or wrong, why the deal advanced or stalled, or whether the prospect was ever going to buy. A rep can score 90 on a call they lost, because the loss was about fit rather than execution. A rep can score 50 on a call they won, because the prospect bought despite the call rather than because of it.

This is a feature, not a bug. Call scoring measures inputs, not outputs. Inputs are what managers can actually improve. Quota is an output. What the rep said in the first 30 seconds of a cold call is an input. Score the inputs consistently and the outputs improve over time.

How to choose what to score: the 6 key dimensions

The dimensions you score should be the behaviours that most predict a good outcome for your specific type of call. A cold call scoring rubric looks different from a discovery call rubric. Here are the six dimensions that apply broadly across most B2B sales calls:

• Opening and rapport: Did the rep establish relevance quickly? Did they earn the right to continue the conversation?
• Discovery depth: Did they ask questions that went beyond surface-level? Did they uncover the real problem and its consequences?
• Talk-to-listen ratio: Did the rep dominate the call, or give the prospect enough space to talk? A healthy ratio is roughly 40–60 for discovery calls. See our full guide on the talk-to-listen ratio.
• Objection handling: How did the rep respond to pushback? Did they acknowledge the objection and re-engage, or collapse or skip over it?
• Value communication: Did the rep connect the product to the prospect’s specific problem, or pitch generic features?
• Next steps: Did the call end with a clear, agreed-upon next step? Vague closes (“I’ll send some info”) predict deal stalls.

For teams using custom playbooks, additional dimensions might include methodology adherence (MEDDIC qualification, upfront contract usage) or compliance requirements. Keep the total number of dimensions to six or fewer. More than that and the rubric becomes unwieldy. Custom rubrics in Koachly let you define these dimensions for your team specifically.

Building a call scoring rubric your team will use

A rubric that managers do not use consistently is worse than no rubric, because it creates an illusion of rigour without delivering it. The best rubrics are simple enough to apply quickly and specific enough to be actionable.

For each dimension, define what a 1, 3, and 5 look like in plain language. Avoid abstract descriptors (“excellent,” “average”). Use behavioural descriptions:

• Discovery depth — 1: Rep accepted the prospect’s first answer and moved on. No follow-up questions.
• Discovery depth — 3: Rep asked at least one follow-up question but did not explore consequences or quantify the problem.
• Discovery depth — 5: Rep asked follow-up questions, surfaced a specific consequence of the problem, and got the prospect to put a number or timeframe on it.

Write the descriptions with input from your top reps. They know what good looks like in practice. Their language makes the rubric more credible when you share it with the rest of the team.

How to calibrate: getting all managers to score the same way

Calibration is the step most teams skip and then wonder why their scores feel inconsistent. Without it, two managers can score the same call differently enough that the scores become meaningless — one sees a 70 where another sees a 45, because they are measuring different things.

The calibration process is straightforward:

1. Select a call that all scoring managers will review. Pick one that has interesting moments — not a perfect call, not a terrible one.
2. Each manager scores the call independently, dimension by dimension, before the group discussion.
3. Share scores and discuss the gaps. Where did you disagree? What does each score mean to each of you?
4. Update the rubric descriptions to close the interpretation gaps you found.
5. Repeat monthly until variance between managers is consistently small.

The goal is not for every manager to agree on every score. It is for the same call to reliably produce a similar score across the team. That is what makes cross-team comparisons and trend data meaningful. Team analytics in Koachly surfaces score variance across your team so you can see where calibration gaps exist.

Using call scores in 1:1s without demoralising reps

Scores can feel threatening if they are introduced as a verdict rather than a starting point. The framing matters as much as the number.

The most effective approach in 1:1s:

• Lead with a specific moment, not the overall score. “I want to look at a moment from your call on Tuesday” is less defensive than “your score this week was 52.”
• Ask before you tell. Play the clip and ask the rep what they would change. Most reps spot the issue themselves. An insight they find is one they keep.
• Coach one dimension at a time. A rep who needs to improve their discovery depth and their objection handling and their talk ratio will improve none of them if you try all three at once. Pick one. See our sales coaching questions guide for how to structure these conversations.
• Track the trend, not the number. A rep moving from a consistent 58 to a consistent 65 over four weeks is coaching well, regardless of what the absolute number is. Show reps their trend line, not just the latest score.

Common mistakes: what bad call scoring looks like

The most common failure mode is a rubric that measures activity rather than quality. “Did the rep ask a discovery question?” is not the same as “Did the rep ask a discovery question that went deep enough to surface the real problem?” Activity-based scoring rewards checkbox behaviour and produces reps who ask the right questions in the wrong way.

A second common mistake is using scores without a coaching conversation. A score that lands in a rep’s inbox with no context produces anxiety, not improvement. The score is only useful as a prompt for a coaching conversation. Without the conversation, it is just a number.

A third mistake is scoring too infrequently. A monthly score tells a rep what was happening four weeks ago. A weekly score is close enough to the call to be actionable. Daily AI scoring is ideal: every call scored, every week.

Automating call scoring with AI

Manual call scoring does not scale. A manager reviewing one call per rep per week across a team of ten is already spending ten hours — before the 1:1s. AI call scoring changes this by analysing every call automatically and surfacing the moments worth a manager’s time.

The most effective use of AI scoring is as a triage layer. The AI scores all calls and flags the ones with low scores or specific problem moments. The manager reviews those flagged moments — not full recordings — and uses them as the anchor for the weekly 1:1. This compresses the call review work from hours to minutes and makes the coaching conversation more specific.

Koachly’s AI call scoring works across Zoom, Google Meet and Teams. Each call receives a score from 0 to 100 across five skill dimensions, with three specific fixes tied to moments in the recording. Managers can view every rep’s scores in one dashboard without listening to recordings. Reps can practice the moments that scored lowest with the AI roleplay feature. See pricing for what is included on each plan.

Sales call scoring FAQ

Conclusion

A call scoring system does one thing well: it makes “good” concrete. When good is concrete, reps know what to aim for, managers know what to coach, and 1:1s become conversations about specific moments rather than abstract impressions. Start with six dimensions, write behavioural descriptions for each level, calibrate monthly across managers, and keep the 1:1 conversation focused on one skill at a time. AI scoring handles the coverage. The manager handles the conversation.