Call Center Metrics: The 12 KPIs Every Manager Should Track

Why Metrics Matter in a Call Center

Call centers generate a large volume of measurable activity. Every call is a data point: how long it lasted, whether the issue was resolved, how the customer felt at the end. Without a structured approach to measuring that activity, managers are left making decisions based on intuition and anecdote.

The right metrics create a feedback loop. They make it possible to identify where the operation is performing well and where it is not, to test the effect of changes, and to hold teams accountable to outcomes rather than just activity. Metrics also enable resource planning — understanding how call volume, handle time, and staffing interact is fundamental to running a contact center that is neither understaffed nor wastefully overstaffed.

The 12 KPIs below are grouped into four categories: inbound volume metrics, quality metrics, efficiency metrics, and agent performance metrics. Each has a clear definition, a formula where applicable, and a note on how it is typically used.

Inbound Volume Metrics

Volume metrics tell you how many contacts are arriving, how many are being handled, and how many are lost before they reach an agent.

1. Calls Handled

The total number of inbound calls answered by an agent in a given period. This is the foundation metric — it tells you the scale of the operation. Tracking calls handled over time reveals seasonal patterns, campaign-driven spikes, and long-term growth trends. It is used as the denominator in many other KPI calculations.

2. Call Abandonment Rate

The percentage of inbound callers who disconnect before reaching an agent.

Formula: Abandon rate = (abandoned calls ÷ total inbound calls) × 100

A caller who disconnects after one second (often a misdial or accidental call) is typically excluded from the abandoned count using a short-abandon threshold — commonly calls that disconnect within the first 5–10 seconds. Rising abandonment rate is a leading indicator of increasing queue wait times or staffing shortfalls. Track abandonment by time interval (hourly or by 30-minute period) to identify when in the day the problem occurs.

3. Service Level

The percentage of calls answered within a target wait time. A common formulation is "X% of calls answered within Y seconds" — for example, 80% of calls answered within 20 seconds. Service level is typically measured in real time and over daily or weekly intervals. It is the primary operational metric that staffing models are built around.

4. Average Speed of Answer (ASA)

The average time a caller waits in queue before being connected to an agent. ASA is a useful complement to service level — a team can meet its service level target while still having a high average wait time if the distribution of wait times is uneven. Track both.

Quality Metrics

Quality metrics measure whether interactions achieved their purpose and how customers perceived them.

5. First Call Resolution (FCR)

The percentage of customer contacts resolved without requiring a callback, escalation, or follow-up interaction.

Formula: FCR = (calls resolved on first contact ÷ total calls handled) × 100

FCR is widely regarded as a high-signal quality metric because it reflects both agent capability and the adequacy of the processes and tools available to agents. Low FCR drives higher call volumes (customers calling back) and higher costs. It also correlates with customer satisfaction. For more context on how FCR relates to satisfaction scores, see our guide on how to calculate NPS, CSAT, and CES.

6. Customer Satisfaction Score (CSAT)

CSAT measures customer satisfaction with a specific interaction, typically collected via a post-call survey asking the customer to rate the experience on a scale (commonly 1–5 or 1–10). The CSAT score is expressed as the percentage of respondents who gave a positive rating (typically the top 1–2 options on the scale).

Formula: CSAT = (positive responses ÷ total survey responses) × 100

CSAT is interaction-level feedback. It is actionable — a low CSAT score on a specific call type or with a specific agent group points to a targeted improvement opportunity. Survey response rates are typically low, so CSAT should be interpreted alongside other quality indicators rather than in isolation.

7. Net Promoter Score (NPS)

NPS measures customer loyalty at the relationship level rather than the interaction level. It asks: "How likely are you to recommend us to a friend or colleague?" on a 0–10 scale. Respondents are classified as Promoters (9–10), Passives (7–8), or Detractors (0–6). NPS = % Promoters − % Detractors.

NPS collected in a contact center context reflects the cumulative customer experience, not just the most recent call. It is a useful strategic metric but less immediately actionable for day-to-day operations than CSAT or FCR.

Efficiency Metrics

Efficiency metrics measure how much time and resource each interaction consumes.

8. Average Handle Time (AHT)

AHT is the average total time an agent spends on a contact, including talk time (time speaking with the customer), hold time (time the customer is placed on hold during the call), and After Call Work (ACW — time spent on tasks related to the call after it ends, such as updating records or sending a follow-up).

Formula: AHT = (total talk time + total hold time + total ACW) ÷ calls handled

AHT has a direct impact on staffing requirements — higher AHT means fewer calls per agent per hour. Reducing AHT without reducing quality typically requires improving agent access to information (knowledge base, screen-pop from CRM), reducing unnecessary hold time, and streamlining ACW through better tools. Chasing AHT reduction without measuring quality outcomes risks pushing agents to rush calls, increasing repeat contact rates and harming CSAT.

9. After Call Work (ACW) Time

ACW is the time an agent spends on post-call tasks — logging call notes, updating CRM records, completing follow-up actions — before becoming available for the next contact. ACW is a component of AHT but is worth tracking separately because it often reveals process inefficiencies. High ACW relative to talk time suggests agents are spending more time on administrative work than on customer-facing activity. Reducing ACW typically involves better integration between telephony and CRM systems, or AI transcription and summarisation tools that automate call logging.

10. Occupancy Rate

Occupancy measures the percentage of time agents are actively handling contacts (including ACW) versus waiting for the next contact. Very high occupancy leaves no buffer for agents between calls, leading to burnout and quality degradation. Very low occupancy indicates overstaffing relative to demand. Occupancy is closely related to service level — the right occupancy target depends on the staffing model and the service level you are targeting.

Agent Performance Metrics

Agent-level metrics are used for individual coaching, performance reviews, and identifying training needs.

11. Quality Assurance (QA) Score

QA scores are assigned by reviewers (team leaders, quality analysts, or automated scoring tools) who evaluate call recordings against a defined rubric. The rubric typically covers elements such as greeting and tone, adherence to script or compliance requirements, accuracy of information provided, and handling of objections or complaints. QA scores are most useful when the evaluation criteria are consistent and when scored calls are representative of each agent's overall call mix. Reviewing call recordings alongside QA criteria allows managers to give specific, evidence-based coaching feedback.

12. Schedule Adherence

Schedule adherence measures how closely agents follow their assigned schedules — being on the phones during scheduled call time, taking breaks at the designated times, and completing ACW within the allotted period. High schedule adherence is necessary for workforce management models to function accurately. If agents are frequently off-queue during their scheduled times, actual staffing levels diverge from planned levels, and service level targets become difficult to meet predictably.

Real-Time vs Historical Metrics

Call center metrics serve two distinct purposes depending on the time horizon over which they are viewed.

Real-time metrics are displayed on supervisor dashboards and wallboards. They show what is happening now: current queue length, agents available, calls in queue, and service level in the current 30-minute or hourly interval. Supervisors use real-time data to make immediate operational adjustments — moving agents from one queue to another, triggering outbound callback offers when inbound queues spike, or escalating staffing. Call analytics platforms typically provide real-time dashboards alongside historical reporting.

Historical metrics aggregate data over days, weeks, and months. They are the primary tool for performance reviews, trend analysis, capacity planning, and evaluating the impact of operational changes. A one-week dip in CSAT tells you little; a consistent decline over three months points to a structural problem worth investigating.

Both views are necessary. Real-time data without historical context produces reactive management. Historical data without real-time visibility means problems are only discovered after the fact.

How to Build a Metrics Dashboard

A useful metrics dashboard is not simply a collection of every available number — it is a curated set of indicators aligned to your operation's current priorities. A few principles:

• Start with your strategic goals. If your primary objective is customer experience, FCR and CSAT should be prominent. If your objective is cost reduction, AHT and occupancy take priority. The dashboard should reflect what you are trying to achieve.
• Separate real-time and historical views. Operational dashboards (displayed on supervisor screens) should focus on current-state metrics. Reporting dashboards for managers and leadership should focus on trends over defined periods.
• Provide agent-level and team-level views. Aggregate metrics tell you whether the operation is performing. Agent-level metrics tell you where coaching is needed. Both views should be available.
• Avoid metric overload. Twelve metrics is a reasonable upper bound for a primary dashboard. If a metric is not regularly acted upon, it is consuming attention without generating value.
• Build in comparisons. A single data point has limited meaning. Each metric is most useful when compared to the same period last week, last month, and against target. Trends and variances from target are more informative than absolute numbers.

For teams building or reviewing their contact center technology, see our overview of the MCC Contact Center platform for details on analytics, reporting, and agent performance tools.

Frequently Asked Questions