WFM guidePeople management

Contact centre performance management

Q: What metrics should be used for contact centre agent performance management?

The appropriate metrics for individual agent performance management in a contact centre are: (1) Quality score (QA) — assessed by a trained quality analyst against a defined framework. This is the closest measure of agent behaviour and competence. (2) CSAT score (where sufficient sample size exists — typically 20+ surveys per agent per month). (3) Schedule adherence % — a valid compliance metric, though WFM data must be reviewed in context (system errors, unlogged activities). (4) FCR — where measurable at individual level. (5) AHT trend — not as a target, but as a flag for investigation if an agent's AHT diverges significantly from peer average. Metrics that should NOT be used as individual performance targets include: total contacts handled, AHT as a target (creates call-cutting behaviour), occupancy (a function of scheduling, not individual effort), and service level (a team metric, not individual).

Q: How long should a contact centre performance improvement plan (PIP) last?

A contact centre performance improvement plan (PIP) typically runs for 4–8 weeks for operational performance issues (AHT, quality, adherence) and 8–12 weeks where the underlying issue involves behaviour, attitude, or complex skill development. UK employment law does not specify PIP duration, but best practice requires the timeframe to be: (1) long enough for genuine improvement to occur — for most skill-based performance issues, 4 weeks is the minimum meaningful assessment window; (2) proportionate to the severity of the issue; and (3) clearly specified at the outset, with defined success criteria. The PIP should include: weekly check-in meetings with documented outcomes, specific measurable targets, the support being provided (coaching, training, modified duties), and the consequence if targets are not met. Failure to document a PIP rigorously is the most common cause of subsequent unfair dismissal claims.

Performance management in a contact centre is both easier and harder than in most roles. Easier because there is abundant operational data. Harder because that data is easy to misuse — single-metric targets create gaming, AHT targets cut calls short, and contacts-handled counts measure schedule and mix, not effort. Good performance management uses the right data, applied in context, to develop agents — not to justify dismissal for metrics they cannot fully control.

Note on employment law

This guide describes employment law and HR practice as it applies in Great Britain. Employment law varies by jurisdiction and individual circumstances. Always verify the requirements applicable to your situation with your HR team, employment counsel, or ACAS before changing people management practices. This guide is for operational context, not legal advice.

Metric appropriateness matrix

Metric	As a performance target?	Right use	Gaming risk if targeted
QA score (quality)	Yes — the best individual metric	The primary performance indicator. Assess at least 4 contacts/month per agent. Use a defined framework. QA score captures behaviour, not system/mix effects.	Agents can be on best behaviour during known monitoring periods. Use a mix of announced and unannounced monitoring.
CSAT score	Yes — with sample size caveat	Valid at individual level if ≥20 surveys/month per agent. Below this, sample noise makes it unreliable. Use as a secondary indicator alongside QA.	Agents can cherry-pick which contacts to offer surveys to (if they control the prompt). Use ACD-triggered surveys, not agent-initiated.
Schedule adherence %	Yes — a compliance metric	Track and manage adherence. Below 88% warrants coaching. Below 80% warrants formal discussion. Review ACD logs before using adherence data — system errors and unlogged coaching time appear as non-adherence.	Agents mark themselves available but take personal calls or browse. RTA monitoring catches this if real-time status is tracked.
FCR (first contact resolution)	Caution — difficult to measure at individual level	Use FCR as a team metric; individual FCR is noisy. Where individual FCR is measurable and based on repeat-contact match, it is a useful development indicator. Not appropriate as a formal appraisal target without robust measurement.	Agents can 'resolve' contacts prematurely to avoid repeat-contact attribution. Use alongside QA to catch false resolution.
AHT (average handle time)	Never — do not target individuals	Use as a diagnostic indicator only. Flag individuals 25%+ above peer average (same contact mix) for investigation. The investigation may reveal: knowledge gaps, system issues, contact type mix differences, or genuine efficiency problems.	Agents cut calls short, transfer unnecessarily, rush ACW, provide incomplete resolution. FCR collapses. Repeat contacts rise. The cost of gaming AHT exceeds any individual saving.
Contacts handled	Never	A function of AHT, schedule hours, and queue depth. Two agents with identical effort and quality may handle very different contact volumes depending on their shift and the queue. Contacts handled measures none of the right things individually.	Agents rush contacts to inflate count. Quality and FCR collapse. The metric incentivises the wrong behaviour in every scenario.
Escalation rate	Caution — as a development indicator	Above-average escalation rate (vs. peer cohort, same contact mix) is a useful coaching indicator. Not appropriate as a formal appraisal target — some contacts legitimately require escalation, and a low escalation rate may indicate an agent who should be escalating but isn't.	Agents resolve without escalating when they should, leading to incorrect resolutions and higher complaint rates.

Balanced scorecard design

A balanced scorecard prevents single-metric gaming by making performance a weighted composite. No single metric can be optimised to the detriment of others without the overall scorecard score falling. A typical contact centre agent scorecard:

Dimension	Weight	Metrics included	Why this weight?
Quality	40%	QA score (primary). Compliance flags from QA. Complaint rate.	Quality is the primary measure of what the agent actually does on a call. It covers accuracy, process adherence, customer experience, and regulatory compliance.
Customer outcome	25%	CSAT score (where sample is adequate). FCR (where measurable). Escalation rate (as development indicator, lower weight).	Customer outcomes confirm that quality behaviour translates to customer experience. CSAT and FCR are the validation of the QA score — if QA is high and CSAT is low, the QA framework may need reviewing.
Availability and reliability	20%	Schedule adherence %. Bradford Factor / absence score. Attendance record.	An agent who is unavailable when scheduled, regardless of quality, reduces the team's capacity to serve customers. Adherence is a valid performance dimension — being present and available is a fundamental job requirement.
Development and conduct	15%	Completion of training/development activities. Feedback response (coaching actions taken). Team contribution. Behaviour (where relevant).	Long-term performance is a function of an agent's development trajectory. An agent engaging in training and coaching is investing in future performance. Conduct is included here to avoid it being excluded from the formal performance framework entirely.

Performance improvement plan (PIP) structure

Required components of a fair and legally sound contact centre PIP

Clear performance concern

State the specific performance shortfall with data: 'Quality score of 58% over the past 8 weeks against a target of 75%. Specific areas: product knowledge (avg 48% of framework) and call structure (avg 52%).' Not 'your quality is not good enough.'

Measurable success criteria

Define exactly what constitutes successful completion: 'Quality score of ≥72% (sustained over 4 consecutive monitoring weeks), with no individual framework area below 65%, and zero compliance failures.' Vague targets ('significant improvement') are unenforceable.

Support to be provided

Document the support the employer will provide: additional coaching sessions (frequency, duration), targeted training, access to knowledge resources, regular check-ins. Failure to provide support while running a PIP is both unfair and legally problematic.

Review points and duration

State the PIP duration and review cadence: '6-week PIP with weekly check-in meeting on [day/time]. Midpoint review at week 3. Final assessment at week 6.' All meetings must be documented with signatures.

Outcome statement

State clearly what happens at the end of the PIP: 'If targets are met: PIP concludes, performance placed under 3-month observation. If targets are not met: escalation to [next step in disciplinary procedure per company policy].'

Employee response opportunity

Allow the agent to respond to the PIP content before it is finalised: 'Do you have any context about the performance shortfall we should be aware of — health, workplace factors, systems issues?' Document the response. This is both best practice and legally important.

Performance management questions

What metrics should be used for contact centre agent performance management?

Appropriate for individual performance: QA score (primary), CSAT (with ≥20 surveys/month), schedule adherence %, FCR trend (as development indicator). Never as individual targets: AHT (creates call-cutting), contacts handled count (measures schedule and mix, not effort), occupancy (a staffing outcome), service level (a team metric).

How long should a contact centre performance improvement plan (PIP) last?

4–8 weeks for operational performance (quality, adherence, AHT trend). 8–12 weeks for behaviour or complex skill development. UK best practice requires: long enough for genuine improvement, defined success criteria stated at outset, weekly documented reviews, explicit support provision, clear outcome statement. Failure to document rigorously is the most common cause of unfair dismissal claims following PIPs.