Sales Call Scoring: How to Build a Rubric Your Team Actually Uses
A practical framework for scoring sales calls — what to measure, how to weight it, and how to use scores to drive real coaching. Start free.
Most sales call scoring fails before the first call is scored. A manager opens a spreadsheet, lists five criteria they care about, assigns each one a rating from one to ten, and sends it to the team with instructions to score every call. Three weeks later the spreadsheet is abandoned. Some reps scored ten calls. Others scored none. Two managers who graded the same call gave different scores on every single dimension. Nobody coached from the data because nobody trusted it.
The problem is not that scoring is a bad idea. It is that most rubrics are built for the manager's reassurance, not for the rep's development. This guide walks through what separates a rubric that actually drives behavior change from one that collects dust.
The four dimensions worth measuring
Almost every sales call can be evaluated along four dimensions. Not every dimension matters equally on every call type — a cold call is short on discovery, a renewal call is short on opening — but these four capture the work that determines whether a deal moves forward.
Opening and agenda-setting. Did the rep earn the first two minutes? A strong opening establishes why this call is worth the prospect's time, sets a rough agenda, and gets a verbal agreement to proceed. Weak openings dive straight into the pitch, skip the agenda entirely, or spend the first three minutes on small talk with no transition. The specific behaviors to score: upfront contract stated, timing confirmed, prospect's goal for the call surfaced.
Discovery quality. Did the rep understand the prospect's situation before talking about solutions? Discovery is the dimension most managers score wrong because they score it on volume (how many questions were asked) rather than quality (did the questions uncover the real buying drivers). The behaviors that matter: questions that reveal pain, questions that connect pain to business impact, follow-up questions that probe assumptions rather than accepting the first answer, and the ratio of time spent listening versus talking in the discovery portion.
Value articulation. Did the rep connect what Koachly does to the specific problems the prospect described? The failure mode here is generic value propositions — features the rep always mentions regardless of what the prospect said they care about. A good score on this dimension means the rep referenced at least two things the prospect said in discovery and connected them explicitly to the solution. "Earlier you mentioned your reps aren't getting consistent feedback — here's how that maps to what we do" is a strong value articulation. "We use AI to make your reps better" is not.
Next steps. Did the call end with a clear, specific, mutually agreed-upon next step, with a date? "I'll follow up next week" is not a next step. "I'll send you a draft agenda for a discovery call with your VP of Sales — does Thursday at 2pm work?" is. Vague next steps are one of the most reliable predictors of deal stall, and they are almost always the result of the rep accepting the first available escape route rather than pushing for a committed action.
How to weight each dimension
Weighting depends on what you are trying to improve and what type of call you are scoring. There is no universal answer, but there are sensible defaults.
For discovery calls, discovery quality should carry the heaviest weight — typically 40 percent or more. Discovery is the highest-leverage behavior on a first call. Getting it right creates the conditions for everything else. Opening and agenda-setting (20 percent) matters because a weak opening reduces the prospect's willingness to open up in discovery. Value articulation (20 percent) matters less on a discovery call where the goal is mostly to understand, not to sell. Next steps (20 percent) remain important because a discovery call with no committed follow-on is often a polite dead end.
For demo calls, shift the weight toward value articulation (35 percent) because that is the job. Discovery still matters (25 percent) because reps who demo before they understand tend to demo generic features. Next steps carry more weight here (25 percent) because the demo call is often the last time a rep controls the forward momentum of the deal. Opening matters least (15 percent) because both parties already know each other by this point.
Write your weights down and share them with the team. When reps understand why discovery is weighted at 40 percent, they stop treating it as a checklist item and start treating it as the main event.
The scoring scale that works
Use a 1–3 scale, not a 1–10 scale.
A 1–10 scale looks precise but produces noise. When you ask two managers to score the same discovery segment and one gives it a 6 and the other gives it a 7, that is not a meaningful difference — it is ambiguity in the scale. The scores feel objective because they are numbers, but they are not. What does a 6 mean? "Above average but with clear gaps." What does a 7 mean? "Good, one thing to improve." Those definitions are in the scorers' heads, not on paper.
A 1–3 scale forces the question: is this dimension a strength, a development area, or a failure? 3 means the rep executed this dimension well — the specific behaviors were present and effective. 2 means the behaviors were partially present but with a clear, specific gap. 1 means the dimension was largely missing or actively harmful. Scoring the same call with this scale, two managers typically agree 85–90 percent of the time. Scoring it on a 1–10 scale, agreement drops to around 60 percent.
With three dimensions each weighted differently, your composite score range is 3–9. A call that scores 3 across all dimensions (all 1s) needs structural intervention. A call that scores 7–9 is a call you reinforce, not fix.
How to calibrate scores across managers
Inter-rater reliability is the most underestimated problem in call scoring. When different managers score the same call differently, the scores stop meaning anything because they cannot be compared across reps or across time. Two steps fix most of this.
Calibration calls. Once a month, have two managers score the same call independently without sharing their scores. Then compare. Where you agree, nothing needs to happen. Where you disagree by more than one point on any dimension, you have a calibration problem. Work through the specific moments that drove the different scores. The goal is not to agree on the number — it is to agree on what behaviors justify each number so that both managers are applying the same standard.
Anchor examples. For each dimension and each score level, identify a real call clip that exemplifies it. "Here is what a 3 on discovery sounds like. Here is what a 1 on discovery sounds like." Written descriptions of score levels are helpful but vague. Real examples are definitive. When a manager is unsure whether a discovery segment is a 2 or a 3, they can compare it to the anchor clip rather than guessing.
Koachly's scoring feature lets you attach anchor clips to each dimension in your rubric so managers can reference them during scoring — see how the scoring rubric builder works.
Using scores to decide what to coach
The point of scoring is not the score. The point is to make coaching decisions faster and more consistent. Here is a simple decision framework based on the composite score.
Score 7–9: reinforce what's working. The rep is executing well. Coaching here should name the specific behaviors that drove the high scores and connect them to outcomes. "On this call your discovery of the budget timeline at 12 minutes was exactly right — you pushed past the first answer and got the real constraint. That is why you got the follow-on meeting." Positive reinforcement on specific behaviors is more powerful than generic praise.
Score 5–6: one specific skill to develop. The call was mostly competent but something held it back. Look at which dimension scored lowest and pick the one highest-value behavior within that dimension to work on. Not five things — one. Bring a specific call moment to the coaching session, describe what happened, describe the impact, and roleplay the alternative. See how to coach a sales rep in a one-on-one for the full framework.
Score below 5: structural issue. When a rep consistently scores below 5, something more fundamental is wrong — either they have not internalized the ideal call flow, they are misapplying their energy on the wrong types of calls, or there is a confidence or mindset issue blocking execution. Single-session coaching rarely fixes this. It needs a structured development plan, more frequent observation, and often a reset on fundamentals.
Common mistakes
Scoring the outcome, not the behavior. A rep who followed a perfect discovery process but got a "no" from a prospect who was never going to buy deserves a high score on discovery. A rep who skipped discovery and happened to close a deal deserves a low score on discovery. Scoring should measure what the rep did, not whether it worked. Mixing outcome and behavior scores teaches reps that process is irrelevant as long as they hit quota, which is exactly the wrong lesson.
Scoring every call. You do not need to score every call. Scoring is a sampling method. Five scored calls per rep per month is enough to identify patterns. Twenty scored calls per rep per month creates administrative overhead that kills compliance. Score enough calls to see whether the behaviors you are coaching on are changing — not every call that exists.
Scoring without follow-through. Scores that do not lead to coaching conversations are a waste of time. The rubric is a trigger for a coaching conversation, not a report card. If you score calls but never reference those scores in one-on-ones, your reps will correctly conclude that scoring is theater.
Too many dimensions. Rubrics with eight or ten dimensions produce more data but less clarity. When everything is measured, nothing is prioritized. Three to five dimensions is the practical maximum. If you find yourself wanting to add a sixth dimension, ask whether it is a distinct behavior or a subcategory of one you already have.
What a good rubric looks like
Here is a concrete example of a rubric for a discovery call. Three dimensions, weighted, with a 1–3 scale.
Discovery quality (40%). 3: Rep asked at least three open questions that revealed pain, probed at least one answer to understand root cause, and the prospect's primary concern was explicitly named before the rep pivoted to solution. 2: Probing questions were present but surface-level, or the rep pivoted to solution before the prospect's main concern was confirmed. 1: Rep dominated the conversation, asked mostly leading questions, or moved to pitch within the first five minutes.
Opening and agenda (30%). 3: Rep stated the purpose of the call, proposed a rough agenda, and got verbal confirmation within the first two minutes. 2: Agenda was partially set or confirmed vaguely ("that works for me") without explicit agreement. 1: Call started mid-pitch with no agenda, or rep spent more than four minutes on small talk with no transition to purpose.
Next steps (30%). 3: Call ended with a specific action, a named person responsible, and a confirmed date or time. 2: Next step was agreed but vague ("I'll send you something") or undated. 1: Call ended with no next step or with "let me know if you're interested."
This rubric takes ninety seconds to apply to a call. That brevity is a feature, not a limitation. If your rubric takes fifteen minutes to fill out, it will not get filled out.
For teams using Koachly, the built-in rubric builder lets you create this structure once and apply it across every scored call automatically. You can also read our guide to reviewing a sales call for the full call review process that scoring fits into.
Frequently asked questions
Frequently asked questions
- How many dimensions should a call scoring rubric have?
- Three to five. Fewer than three and you lose granularity — a rep can score well on two dimensions while a critical gap goes undetected. More than five and the rubric becomes unwieldy. Managers stop applying it consistently, and reps get overwhelmed feedback on too many things at once. Three dimensions with clear weights is the most practical starting point for most teams.
- How often should you score calls?
- Five scored calls per rep per month is enough to identify trends and inform coaching. Scoring more than that rarely improves coaching quality and significantly increases the administrative burden on managers. The goal is a representative sample, not exhaustive documentation. For new reps in their first 90 days, scoring eight to ten calls per month accelerates ramp — you want to catch bad habits before they become ingrained.
- Should reps score their own calls?
- Yes, and it is one of the most underused coaching tools available. Self-scoring before a coaching conversation forces reps to evaluate their own behavior before hearing their manager's perspective. When rep and manager score the same call, the disagreements surface the most interesting coaching conversations. Reps who score their own calls consistently improve faster than reps who only receive scores from their managers — the metacognitive habit of evaluating your own performance is itself a skill worth developing.
- What's the difference between a scorecard and a rubric?
- A scorecard records what happened — metrics like call duration, talk-to-listen ratio, number of filler words, whether a next step was set. A rubric evaluates quality — was the discovery thorough, did the rep articulate value specifically, was the opening effective? Scorecards are easy to automate. Rubrics require judgment. Both are useful; neither replaces the other. Automated scorecards flag calls worth reviewing. A rubric tells you why the call went the way it did.
- How do I get my managers to score calls consistently?
- Two things: calibration sessions and anchor examples. Calibration sessions (two managers scoring the same call independently once a month, then comparing) surface disagreements before they become systemic. Anchor examples — real call clips attached to each score level for each dimension — replace subjective interpretations with concrete references. Without these, managers drift toward their own implicit standards over time, and your scoring data loses its comparability.