Human scoring
Your staff's calls, graded against your standards
Your own staff's calls, graded against your own standards.
What it does
When a call is transferred to a person, that half is recorded too. Each one is scored against criteria you set, with a mark, a short summary and a note on what could have gone better.
You see it by person and by department, so you can tell whether a problem is one member of staff or the whole team.
What it does not do
- It does not grade the AI. This is about your people.
- It does not listen to calls that never reached a person. No transfer, no score.
- It does not decide what good looks like. You write the criteria. It applies them.
Setting it up
Settings, then the scoring section. Write each criterion as something a poor call would visibly fail, not as a virtue.
Good: Confirmed the customer's name and repeated the booking time back.
Bad: Was professional.
The one thing that matters
Watch the spread, not the average.
If every call scores 8 out of 10 on a criterion, that criterion is telling you nothing. It cannot separate your best call from your worst, so it is taking up a slot.
The useful criteria are the ones where scores vary. Those are the ones where behaviour actually differs and where coaching changes something.
Your Data Report lists them worst first, with the spread beside each.
Common mistakes
Vague criteria. "Was helpful" cannot be failed, so everyone passes.
Too many. Five that discriminate beat twelve that do not.
Never changing them. The grader leaves notes about things no criterion asks about. When the same note keeps appearing, that is your next criterion.