The Call Centre QA Scorecard: Template, Examples and Best Practices

Your QA scorecard defines what "good" looks like. Get it right and coaching, calibration and analytics all fall into place. Get it wrong and you measure the wrong things consistently. Here is how to build one that works.

Start with sections, not a flat list

Group criteria into logical sections that mirror the call flow:

1. Opening — greeting, branding, identity verification
2. Compliance — mandatory disclosures, data handling, consent
3. Resolution — accuracy, product knowledge, problem-solving
4. Soft skills — empathy, active listening, tone
5. Closing — recap, next steps, professional close

Sections make the scorecard readable, let you weight whole areas, and produce far more useful analytics than a flat checklist.

Weight criteria to reflect what matters

Not every criterion is equal. Assign weights that add up to 100% and reflect business priority. If empathy drives your CSAT, weight it accordingly. If you run collections QA, compliance carries more weight than upsell.

Separate fatal criteria

Some failures are not "partial". A missed mandatory disclosure or a data-protection breach is catastrophic regardless of how good the rest of the call was. Mark these as fatal criteria: triggering one scores the whole evaluation zero and raises an escalation. This ensures critical breaches are never averaged away by an otherwise strong call.

Track an independent compliance score

This is the step most teams miss. An agent can deliver a wonderful experience while still breaching a regulation. A single blended score hides that risk. Run a separate compliance score built from your regulatory criteria, with stricter escalation, so quality and compliance are visible independently. In regulated industries — insurance, banking, healthcare — this is non-negotiable.

Write criteria as observable behaviours

"Was empathetic" is unscoreable. "Acknowledged the customer’s situation before moving to resolution" is. The more concrete and observable each criterion, the higher your inter-rater reliability and the more actionable your coaching.

Calibrate relentlessly

Even a perfect scorecard fails if two evaluators score the same call differently. Run regular calibration sessions and watch evaluator drift. Consistency is what makes scores trustworthy to agents and useful to the business.

Example: a retentions scorecard

Section	Example criterion	Weight	Type
Opening	Professional greeting & branding	10%	Standard
Compliance	Mandatory disclosure given	20%	Fatal + compliance
Compliance	Identity verified before account info	15%	Compliance
Resolution	Genuine save offer made	20%	Standard
Soft skills	Empathy & active listening	15%	Standard
Closing	Correct close & next steps	20%	Standard

Don’t build it alone — let AI draft it

Modern platforms can draft a scorecard from your policy document, flagging fatal and compliance criteria automatically, then let a human refine it. That turns a multi-week committee exercise into an afternoon.

Ready to build yours? Start a free 30-day trial and use the scorecard builder, or try the live scoring demo first to see a scorecard in action.