Agent Scorecard

What goes on a good agent scorecard

Five to seven measures is the practical limit. Past that, agents cannot tell what matters and team leaders spend coaching sessions explaining math. Most contact centers settle on a mix like this:

• Quality score from quality assurance evaluations. Usually the heaviest weight, 30% to 40%.
• Resolution, measured as first call resolution or repeat contacts within 7 days.
• Customer feedback, such as CSAT from post-call surveys.
• Efficiency, typically average handle time against a target for the agent's queue.
• Attendance and adherence, meaning schedule adherence and unplanned absence.

Sales and collections teams add conversion or dollars collected. Back-office teams swap AHT for items processed per productive hour. Pick from the wider list of call center KPIs only those an agent can actually influence. Service level, for example, belongs on the team or site scorecard, not the individual one.

How to set weights and targets

Weights say what the operation values. If quality is 35% and AHT is 10%, agents get the message that a slower, correct call beats a fast, wrong one. Flip those weights and you will see transfers and short calls within a month.

Set targets by queue and tenure, not one number for the floor. A billing agent and a technical support agent should not share an AHT target, and an agent in week three after nesting should be on a ramped target. A common pattern is 80% of the full target in month one, 90% in month two and 100% from month three.

Cap attainment on each metric. Without a cap, an agent who crushes AHT can mask poor quality and still land at the top of the ranking.

Using the scorecard for coaching, not just ranking

The scorecard tells a team leader where to look. It does not tell them why. An agent with low FCR might be missing product knowledge, skipping discovery questions or working a queue that gets the hardest calls. That is what call coaching sessions are for: pull two or three calls, listen together and agree on one behavior to change before the next review.

Review weekly for new hires and monthly for tenured agents. Show the trend, not just the latest number, and let agents see their own data every day so the monthly review holds no surprises. If you need a ready layout, start from our call center agent scorecard template and adjust the weights to your queues.

Common scorecard mistakes

• Too many metrics. Twelve measures at 8% each means nothing moves the score.
• Too few QA evaluations. Four evaluations a month is a small sample. One bad call can swing the quality score by 10 points, so treat month-to-month quality changes with caution unless you score more calls or use automated scoring.
• Metrics the agent cannot control. Abandonment rate and service level depend on staffing. Putting them on an agent scorecard feels unfair because it is.
• No calibration. If two QA analysts score the same call 15 points apart, the quality line on the scorecard is noise. Run regular calibration sessions.
• Tying pay to a broken scorecard. Fix the measures first, then link incentives. Agents will optimize exactly what you pay for.

Scorecards in high-volume and BPO settings

In outsourced operations the agent scorecard usually rolls up into a client-facing scorecard tied to the contract's service levels. Keep the two separate. The client cares about site results; the agent scorecard should reflect what one person controls.

When a center hires a new class of 30 agents every month, the scorecard also becomes a hiring feedback loop. Compare month-three scores by recruiting source, assessment result and trainer. If agents from one source consistently score lower on quality, that is a screening problem, not a coaching one.