Call evaluation scorecard: the 10 criteria that matter
Anyone can listen to a call. Evaluating it fairly, consistently from one agent to the next and in a way that helps people improve, is another matter. Without a scorecard, every manager scores on gut feeling, agents don't know what they're being judged on, and feedback stays vague: "be more upbeat", "listen to the customer more".
A good call evaluation scorecard solves all three problems. Here's how to build one, whether your teams make sales calls, handle customer service or manage after-sales.
What a call evaluation scorecard should do
Before listing the criteria, let's set the goal. A call evaluation scorecard is there to:
- score every call the same way, whoever the evaluator is;
- show the agent what went well and what to work on first;
- track progress over time, agent by agent and across the team.
If your scorecard can't do these three things, it will be filled in… then forgotten.
The 10 criteria of a good scorecard
The criteria below suit most calls. Keep the ones that make sense for you, adapt them to your script and weight them by call type.
1. The opening
Does the customer know who they're speaking to from the very first sentence? Company name, agent's first name, a courteous greeting. A good example: "NextCall, Jean speaking, good morning!"
2. Needs discovery
Does the agent ask open questions to understand the situation before suggesting anything? In sales as in after-sales, an answer given too early is often the wrong one.
3. Active listening
Do they let the customer finish their sentences? Do they pick up on what has just been said rather than reading their script word for word?
4. Restating
This is the most often forgotten criterion, and one of the most effective. "So, to sum up, you need…": the customer feels understood, and the agent checks they aren't heading down the wrong track.
5. Subject knowledge
Is the information given accurate and precise: product, pricing, procedure, timescales? A hesitant agent erodes trust, even with the right attitude.
6. Handling objections or complaints
In sales, does the agent respond to "it's too expensive" or "I'll think about it"? In customer service and after-sales, do they deal with the customer's frustration without making excuses or shifting the blame? The right response is measured by what follows: does the customer stay in the conversation?
7. Clarity and tone
Simple sentences, no internal jargon, a pace that suits the customer. The tone stays courteous even when the call gets tense.
8. The solution or proposal
Does the agent provide a concrete answer: a suitable offer, a solution to the problem, a clear procedure? This is the heart of the call.
9. Closing and next step
A successful call ends with a clear next step: a meeting booked, a quote sent, a ticket resolved or a call-back scheduled. This is often where a "pleasant" call and a useful call part ways. It's also an excellent metric to track: the share of calls that end with a concrete action.
10. Process compliance
Identity checks, mandatory disclosures, logging in the CRM, following the script: everything that is non-negotiable in your business.
Adapting the scorecard to the call type
The ten criteria don't carry the same weight everywhere:
| Call type | Most heavily weighted criteria | Metric to track |
|---|---|---|
| Sales | Needs discovery, objections, closing | Share of calls with a next step |
| Customer service | Listening, clarity, solution | First call resolution rate |
| After-sales | Complaint handling, subject knowledge, process | Resolution rate, repeat calls avoided |
| Account follow-up | Restating, solution, closing | Satisfaction, repeat purchases |
A single scorecard for every call produces unfair scores. Plan one scorecard per main call type.
A scoring scale that leaves no room for ambiguity
"Good", "average", "needs work"… these words don't mean the same thing to everyone. For two evaluators to give the same score, describe what you observe at each level. For example, for the opening:
- 0: no company name and no first name;
- 4: partial or hesitant introduction;
- 6: complete but clumsy introduction;
- 8: in line with the script, with a minor slip;
- 10: natural, complete, usable as a model for the team.
Two simple rules make scores fairer:
- No opportunity, no penalty. If the customer raised no objection, the "objections" criterion doesn't count towards the score.
- One score, one piece of evidence. Every score should rest on a sentence that was actually spoken. That's what makes feedback indisputable and useful to the agent.
The 4 most common mistakes
- Too many criteria. Beyond a dozen, nobody fills in the scorecard properly any more.
- Criteria you can't observe. You can't hear "being motivated" in a call; you can hear "suggesting a meeting".
- Evaluating only a few calls. One call per agent per month tells you nothing reliable: chance weighs more than skill.
- Scoring without feedback. An evaluation that doesn't lead to a conversation with the agent, and to one specific point to work on, is pointless.
And across a whole team?
Applying a scorecard by hand takes around 15 to 20 minutes per call: listening, scoring, finding the relevant passages, writing feedback. That's why most teams only evaluate a handful of calls.
That's the work NextCall does: your scorecard is applied automatically to your teams' calls, whether sales, customer service or after-sales. Every score is backed by the exact sentence from the call, and every agent gets the point to work on in their next call. Managers know which calls to listen to and what to coach on, instead of searching at random.
See NextCall on your own calls →
In summary
- A good scorecard contains 8 to 12 observable criteria, adapted to the call type.
- Each score level is described in concrete terms, so everyone scores the same way.
- Every score rests on a sentence from the call.
- Evaluation only has value if it leads to one specific improvement point for the agent.