Skip to main content
TurnellaBeta
WFM guideMetrics

Contact centre agent productivity

Agent productivity is not a single number. Contacts per hour measures throughput. AHT measures efficiency. Occupancy measures time use. None of them alone tells you whether the agent produced a good outcome. The contact centres that manage productivity through one metric consistently produce worse outcomes than those that don't.

The core productivity metrics — and what each misses

Average Handle Time (AHT)

AHT = (Talk Time + Hold Time + After-Call Work) ÷ Total Contacts

What it measures

The average duration of each contact from answer to completion. The most commonly managed productivity metric in contact centres.

What it misses

AHT does not measure whether the contact was resolved. An agent with a low AHT who never resolves contacts on the first call produces worse customer outcomes and higher total cost than an agent with a 20% higher AHT who resolves contacts cleanly.

Gaming risk

High. Agents under AHT pressure reduce hold, interrupt customers, transfer rather than resolve, avoid complex contacts, and close contacts before they are complete. Each of these produces a repeat contact — making the 'efficiency' gain self-defeating.

Benchmark reference

Varies by contact type: simple account queries 2–4 minutes; complex complaints 10–20 minutes; financial product sales 15–30 minutes. Setting a single AHT target across all contact types is a structural error.

Contacts per hour (CPH)

CPH = Total Contacts Handled ÷ Hours Logged On

What it measures

Throughput: how many contacts the agent completes per hour of logged-on time. A cleaner productivity metric than AHT because it is outcome-focused rather than time-focused.

What it misses

CPH does not capture contact complexity. An agent who handles simple contacts gets a structurally higher CPH than an agent assigned complex complaints — making comparison across agent groups invalid unless contact mix is controlled for.

Gaming risk

Medium. Agents can increase CPH by rushing, transferring, or taking shorter contact types. Controlled by monitoring transfer rate and repeat contact rate alongside CPH.

Benchmark reference

Depends heavily on contact type. Voice simple: 10–15/hr. Voice complex: 3–6/hr. Live chat (concurrent): 8–16/hr (2–4 concurrent × 4–6/hr per chat). Email: 6–12/hr.

Occupancy

Occupancy = (Talk Time + Hold Time + ACW) ÷ (Talk Time + Hold Time + ACW + Available Idle Time)

What it measures

The proportion of logged-on time spent in active contact. A measure of how hard agents are working when they are available — driven by volume and staffing level as much as by individual agent behaviour.

What it misses

Occupancy above 85–87% is associated with agent stress and quality deterioration. Occupancy is primarily a staffing-level outcome, not an individual agent performance metric — if the team is understaffed, occupancy rises whether agents work harder or not.

Gaming risk

Low as an individual metric. Agents cannot game occupancy in isolation because it is driven by contact arrival rates. However, agents can stay in ACW longer to reduce occupancy — monitored via ACW duration analysis.

Benchmark reference

Target range: 80–85% for complex contacts; 85–87% for standard voice; above 87% is a stress indicator. Below 75% suggests overstaffing or low volume.

Occupancy vs. utilisation — the distinction that matters

Occupancy

Contact Time ÷ Logged-On Time

What proportion of the time an agent is logged on and available does the agent spend in active contact? This is an operational metric — driven by the ratio of volume to staffing, not primarily by individual agent effort. If you have 10 agents and 120 contacts/hr at 5 minutes AHT, the team occupancy is mathematically 100% — no agent behaviour changes that.

Utilisation

Productive Contact Time ÷ Total Paid Time

What proportion of total paid time (including shrinkage — breaks, training, meetings, absence) is converted to productive contact handling? This is a capacity planning metric. An agent paid for 8 hours who is available for 5.5 hours (after 31% shrinkage) and has 80% occupancy during available time has 55% utilisation of total paid time.

MetricUsed forAffected byOwner
OccupancyReal-time staffing decision-making; agent wellbeing monitoringContact volume, staffing level, AHTIntraday / WFM team
UtilisationHeadcount planning; cost per contact; budget modellingShrinkage, occupancy, contract hoursWFM / Finance

A balanced agent productivity scorecard

A three-pillar scorecard captures productivity, quality, and attendance together. Agents should be assessed on all three — a high-CPH agent who fails QA is not a productive agent.

PillarMetricTypical weightTarget directionGaming guard
ProductivityAHT vs. team average15–20%Within ±20% of team normMonitor transfer rate, ACW separately
ProductivityContacts per hour (CPH)10–15%At or above team average for same contact mixSegment by contact type before comparing
ProductivitySchedule adherence10–15%≥90–95%Attendance records; exceptions logged
QualityQA score25–35%≥ defined floor (e.g. 85%)Calibration sessions; multi-assessor scoring
QualityFirst contact resolution (FCR)20–25%At or above team averageRepeat contact tracking by agent; definition must be agreed
QualityCustomer satisfaction (CSAT)10–15%At or above team averageSufficient sample size required; survey gaming monitoring
Scorecard design principle:Weight the scorecard so that quality alone cannot compensate for productivity failure, and productivity alone cannot compensate for quality failure. An agent who scores 100% on QA but handles half the contacts of peers is not a productive agent — they are taking time to achieve quality that the team's staffing plan cannot afford. An agent who handles 130% of peer contacts but fails QA on 30% of calls is generating rework. Neither profile is acceptable.

Target-setting principles

Set targets relative to the team, not against an abstract ideal

AHT targets based on 'best agent in the team' or 'theoretical minimum' ignore contact complexity variation and natural communication style differences. Set targets relative to team median — within ±20% is typical. Flag the tail, not the median.

Segment before you compare

An agent handling complex complaints has a structurally higher AHT than an agent handling password resets. Comparing them directly to the same AHT target is incorrect. Either segment targets by contact type or use a complexity-adjusted metric.

Investigate outliers before taking action

An agent with unusually low AHT may be under-handling contacts (rushing, transferring, closing early) or may genuinely be more efficient. An agent with unusually high AHT may be taking longer than needed or may be handling unusually complex contacts. The metric identifies who to investigate — it does not identify why.

Never manage AHT directly with agents

Tell agents the target exists (they should know). Do not actively coach agents to reduce AHT in a performance review context. Coaching on AHT produces quality deterioration. Coaching on the behaviours that generate unnecessary AHT (excessive hold, unnecessary ACW, scripting gaps) does not.

Review targets quarterly

As contact mix changes (new products, new channels, changes in customer skill level with self-service), productivity baselines shift. A target set twelve months ago against a different contact mix is wrong today. Review targets quarterly against current data.

Agent productivity questions

What is agent productivity in a contact centre?

Agent productivity is measured across three dimensions: throughput (contacts per hour — how many contacts the agent handles per logged-on hour); efficiency (AHT — duration of each contact including after-call work); and time use (occupancy — what proportion of logged-on time is in active contact). None of these alone captures whether the contact produced a good outcome. A complete view combines these with quality metrics (FCR, QA score, CSAT) and attendance (schedule adherence, shrinkage contribution).

What is the difference between occupancy and utilisation in a contact centre?

Occupancy is the proportion of an agent's logged-on (available) time spent in active contact — an operational metric driven by volume-to-staffing ratio. Utilisation is the proportion of total paid time (including shrinkage) converted to productive contact handling — a capacity planning metric. An agent on 80% occupancy during 5.5 hours of available time, out of 8 paid hours, has 55% utilisation. Occupancy answers 'how hard are agents working when available?'; utilisation answers 'what proportion of payroll generates productive contact handling?'

Related guides