Performance Reviews in Call Centers — What to Evaluate, How to Structure Them, and How to Make Them Drive Improvement
A performance review in a call center should be the formal version of what the agent already knows. If the agent is surprised by their review — either positively or negatively — the problem is not the review. The problem is that coaching and feedback have not been happening consistently between reviews.
The review itself does three things that ongoing coaching does not: it creates a documented record of performance over a defined period, it connects performance to compensation and advancement decisions, and it sets formal goals for the next period. These three functions are what distinguish a performance review from a coaching session — and they are the reason reviews matter even in operations where supervisors coach regularly.
What to evaluate
A call center performance review should evaluate two categories: metrics (quantitative — the numbers) and behaviors (qualitative — how the agent works). Both matter. An agent who hits every metric target but is hostile to teammates, refuses to follow process changes, or is dishonest on timesheets should not receive a positive review. An agent who is cooperative and professional but consistently misses targets should not either.
Metric evaluation
Use the agent's actual performance data for the review period. Do not rely on the supervisor's impression of how the agent performed — pull the data.
| Metric | Source | What to evaluate | How to score |
|---|---|---|---|
| AHT | ACD reports | Average handle time by call type compared to target. Trend over the review period (improving, stable, declining) | Within 10% of target = meets expectations. Within 20% = developing. Beyond 20% = below expectations |
| FCR | ACD or CRM data | First call resolution rate. Compare to team average and target | Above target = exceeds. Within 5 points = meets. Below target by 5+ points = developing or below |
| QA scores | QA evaluation records | Average QA score, trend, and specific rubric categories that are strong or weak | 90%+ = exceeds. 85–89% = meets. 80–84% = developing. Below 80% = below expectations |
| Adherence | Time tracking / WFM system | Schedule adherence percentage. Pattern of non-adherence (late arrivals, extended breaks, early departures) | 95%+ = exceeds. 90–94% = meets. 85–89% = developing. Below 85% = below expectations |
| Attendance | Time tracking / HR records | Unplanned absence rate. Occurrence count if using a point system | Within policy = meets. 1 occurrence from threshold = developing. At or above threshold = below expectations |
| Calls per hour | ACD reports | Throughput compared to target for the agent's call mix | At or above target = meets/exceeds. Within 10% = developing. Below by 10%+ = below expectations |
Important: Present each metric in context. An agent with AHT 15% above target who also has the highest FCR on the team is thorough but slow — a different situation than an agent with high AHT and low FCR. The review should connect the metrics together, not evaluate each in isolation.
Behavior evaluation
Behaviors are assessed through QA evaluations, supervisor observations, and documented incidents during the review period.
| Behavior category | What to evaluate | Evidence sources |
|---|---|---|
| Call handling quality | Follows the call flow (opening, verification, diagnosis, resolution, documentation, closing). Uses troubleshooting flowcharts when available. Communicates resolution clearly to the customer | QA evaluation scores by rubric category |
| Process compliance | Follows required processes — verification steps, disclosures, disposition coding, escalation procedures | QA evaluations, supervisor observations, compliance audit results |
| Responsiveness to coaching | Implements feedback from coaching sessions. Shows improvement on previously identified gaps | Coaching notes, metric trends after coaching interventions |
| Reliability | Shows up on time, follows the schedule, communicates absences properly, does not extend breaks or leave early | Adherence data, attendance records, timesheet data |
| Teamwork and professionalism | Cooperates with peers and supervisors, participates in team activities, handles shift swaps professionally, communicates respectfully | Supervisor observations, documented incidents, peer feedback where applicable |
The rating framework
Use a defined rating scale so that reviews are consistent across supervisors and defensible for compensation or employment decisions.
4-level rating scale
| Rating | Definition | Metric criteria | Behavior criteria |
|---|---|---|---|
| Exceeds expectations | Consistently performs above target on most metrics. Demonstrates behaviors that other agents should model | 4+ of 6 metrics above target. No metric below expectations | Positive QA trend, strong coaching responsiveness, zero attendance issues, recognized for quality or teamwork |
| Meets expectations | Consistently performs at or near target. Reliable and professional | All metrics at or within acceptable range of target. No more than 1 metric in "developing" | Follows processes, responds to coaching, meets attendance standards |
| Developing | Performance is below target on 2+ metrics but trending in the right direction, or the agent is within the first 6 months of the ramp period | 2–3 metrics below target but showing improvement over the review period | Responds to coaching but improvement is not yet reflected in metrics. Or new agent still building proficiency |
| Below expectations | Performance is below target on 3+ metrics with no improvement trend despite coaching. Or behavior issues that are not being corrected | 3+ metrics below target with flat or declining trend | Does not respond to coaching, attendance or adherence issues, process compliance failures |
Calibrating across supervisors
Different supervisors will rate differently unless the operation calibrates. Supervisor A may rate an agent "exceeds" for the same performance that Supervisor B rates "meets."
| Calibration step | How to do it |
|---|---|
| Define anchor examples | For each rating level, document 2–3 example profiles (specific metric ranges + behavior descriptions) that represent that rating. All supervisors reference these examples |
| Pre-review calibration session | Before the review cycle, supervisors review 3–5 borderline cases together and discuss ratings. Align on where "meets" ends and "exceeds" begins |
| Ops manager review of ratings distribution | After supervisors complete initial ratings, ops manager reviews the distribution. If one supervisor rates 50% of agents "exceeds" and another rates 10%, investigate the disparity |
How to structure the review conversation
The review conversation should follow a consistent format. A 30-minute review that is structured produces better outcomes than a 60-minute review that meanders.
The 30-minute format
| Segment | Time | What happens |
|---|---|---|
| 1. Performance summary | 5 min | Supervisor presents the overall rating and the metric summary. No surprises — the agent should have been seeing these metrics in their scorecard throughout the period |
| 2. Strengths | 5 min | Identify 2–3 specific strengths with evidence. "Your FCR is 78%, which is 3rd on the team. On the call on March 2 at 10:15, you identified the billing discrepancy on the first attempt — that is the kind of resolution that prevents callbacks" |
| 3. Development areas | 10 min | Identify 1–2 areas for improvement with specific examples and a concrete plan. "Your ACW averages 85 seconds. We discussed documentation standards in coaching on February 15. For the next period, the target is 60 seconds. We will review your ACW weekly and I will listen to 3 calls per week to provide specific feedback on your post-call process" |
| 4. Agent input | 5 min | Ask the agent: "What would help you perform better?" and "Is there anything about the role, schedule, or process that is making your job harder than it needs to be?" Listen and take notes |
| 5. Goals for next period | 5 min | Set 2–3 specific, measurable goals for the next review period. The agent should leave knowing exactly what "success" looks like for the next 3–6 months |
What to avoid in the conversation
| Mistake | Why it fails | Instead |
|---|---|---|
| Starting with weaknesses | The agent becomes defensive immediately and stops listening | Start with the overall rating and strengths. The agent is more receptive to development areas after hearing what they do well |
| Vague feedback | "You need to improve your customer service skills" tells the agent nothing actionable | Be specific: "On 3 of 6 evaluated calls, you did not confirm the resolution before ending the call. The QA rubric scores this as a process compliance miss. The specific change is: before ending every call, ask 'Is there anything else I can help you with?' and confirm the resolution" |
| Discussing every metric | Walking through all 6 metrics and 5 behavior categories takes too long and dilutes the message | Focus on the 2–3 strongest areas and the 1–2 most important development areas. Provide the full scorecard in writing for the agent to review |
| Comparing to other agents by name | "You're slower than Sarah" creates resentment, not motivation | Compare to targets and team averages, never to named individuals |
| Making promises the supervisor cannot keep | "If you improve your AHT, I'll get you the Monday–Friday shift" may not be within the supervisor's authority | Only commit to actions within the supervisor's control. If the agent asks about schedule changes, promotions, or raises, explain the process honestly rather than making promises |
Connecting reviews to outcomes
The review must connect to real outcomes — otherwise agents learn that the review is a formality with no consequence.
The four review outcomes
| Outcome | When it applies | What happens next |
|---|---|---|
| Recognition and advancement | Agent rated "exceeds expectations" for 2+ consecutive review periods | Eligible for advancement (senior agent, team lead), preferred shift selection, remote work eligibility, or merit increase. Communicate what they earned and why |
| Continuation | Agent rated "meets expectations" | Standard coaching cadence continues. Set goals for the next period that push toward "exceeds" if the agent is interested in advancement |
| Improvement plan | Agent rated "developing" for 2 consecutive periods (not improving), or rated "below expectations" | Formal performance improvement plan with specific targets, timeline (typically 30–60 days), support provided, and consequences if targets are not met |
| Separation | Agent on improvement plan does not meet targets within the defined timeline | Employment decision. This should never be a surprise — the agent was told the timeline and consequences during the improvement plan conversation |
Compensation connection
| Rating | Compensation outcome | Notes |
|---|---|---|
| Exceeds expectations | Merit increase (if budget allows) + eligible for advancement | The increase should be meaningful enough that the agent perceives a real reward for sustained high performance |
| Meets expectations | Standard increase (cost-of-living or step increase per policy) | Adequate, but the agent should understand that higher performance leads to higher rewards |
| Developing | No increase. Or standard increase if the trajectory is positive and the agent is within ramp | Explain that the increase is tied to reaching targets. Set a check-in to reassess in 90 days |
| Below expectations | No increase. Improvement plan initiated | Focus the conversation on the improvement plan, not the lack of increase. The path forward is clear: meet the targets, retain the role |
Review cadence
| Review type | Frequency | Duration | Purpose |
|---|---|---|---|
| Formal performance review | Every 6 months (or annually, depending on operation) | 30 minutes | Documented evaluation against all metrics and behaviors. Connects to compensation and advancement |
| Coaching session | Weekly or biweekly | 15–20 minutes | Focus on 1–2 specific behaviors or metrics. Not formally documented as a review but notes are kept |
| QA evaluation debrief | After each evaluation cycle (monthly) | 10–15 minutes | Review specific calls, discuss rubric scores, set immediate improvement focus |
| Improvement plan check-in | Weekly during an active improvement plan | 15 minutes | Review progress against improvement plan targets. Documented |
Why 6-month reviews are better than annual reviews for call centers: In an operation where average tenure is 12–18 months, an annual review means many agents receive zero or one formal review before they leave. A 6-month cycle ensures that every agent gets at least one formal review, and agents who stay 18+ months get three — enough data to support advancement decisions.
Performance reviews for BPO operations
BPO operations have additional considerations because agent performance affects client SLAs and contract health.
| BPO-specific element | What it adds to the review |
|---|---|
| Client-specific metrics | If the agent works on a specific client account, evaluate against that client's SLA targets, not just internal targets. An agent who meets internal targets but misses client SLA requirements is underperforming for that account |
| Multi-account performance | For cross-trained agents, evaluate performance per account. An agent may exceed on Account A but struggle on Account B — the development plan should address the specific account gap |
| Client feedback | If the client has flagged specific agents (positively or negatively), incorporate that feedback into the review with attribution |
| Account profitability awareness | Agents do not control profitability directly, but supervisors should understand how agent performance (AHT, overtime, adherence) affects the account's margin. An agent with consistently high AHT on a per-call billing model directly reduces margin |
Documentation and record keeping
Every formal performance review should be documented and retained.
| Document element | What to include |
|---|---|
| Review date and period | The specific dates covered by the review |
| Metric summary | All metrics for the period with targets, actuals, and trend |
| Behavior assessment | Rating per behavior category with specific examples |
| Overall rating | The rating from the 4-level scale with justification |
| Development areas | 1–2 specific areas with the improvement plan |
| Goals for next period | 2–3 measurable goals |
| Agent acknowledgment | Agent signature confirming they received the review (not that they agree with it) |
| Supervisor signature | Supervisor signature confirming the review was delivered |
Retain review documents per your HR policy — typically 3+ years or the duration of employment plus 1–2 years. These documents are the evidence base if an employment decision is challenged.
