Contact centre metrics dashboard
Not all metrics belong on the same dashboard. A real-time wall board showing CSAT is useless — agents cannot change CSAT in the next 15 minutes. A strategic leadership view showing interval-level AHT is noise — senior leaders cannot act on 30-second variances. The right metrics dashboard design matches data freshness and metric type to the audience and their decision-making window.
The four dashboard tiers
Real-time wall board
Primary metrics
- ▸Current service level % (vs. target)
- ▸Contacts in queue / longest wait
- ▸Agents available vs. scheduled
- ▸Agents not ready / unavailable count
- ▸Abandon rate (current interval)
- ▸Volume vs. forecast (current interval)
Not for this view
AHT, FCR, CSAT, NPS, attrition — anything that cannot be acted on in the next 15 minutes. If agents can't change it right now, it doesn't belong on the real-time board.
Tooling note
ACD / telephony platform real-time reporting. WFM intraday module. Some operations use large screens; most modern operations use browser-based dashboards accessible to team leaders on desktop.
Intraday WFM dashboard
Primary metrics
- ▸Volume vs. forecast by interval (% variance)
- ▸AHT vs. forecast (actual vs. planned)
- ▸Adherence % by team
- ▸Occupancy % (current interval and rolling hour)
- ▸Headcount available vs. required (Erlang gap)
- ▸Escalation / break queue flag
Not for this view
Strategic metrics (attrition, annual leave utilisation, FCR trends). These belong in the daily or weekly view. The intraday WFM dashboard should surface re-planning decisions, not trend analysis.
Tooling note
WFM platform intraday module (Calabrio, Verint, NICE WFM, or equivalent). Requires ACD integration for real-time volume and state data. Often combined with the wall board data in the WFM platform.
Daily operations dashboard
Primary metrics
- ▸Service level % (daily actual vs. target)
- ▸Abandon rate (daily)
- ▸Average AHT (daily vs. forecast vs. trend)
- ▸FCR % (daily if measurable; often 24-hour lag)
- ▸Adherence % (daily by team)
- ▸Contacts handled vs. forecast
- ▸Agent absence / Bradford score flags
Not for this view
Month-to-date CSAT or NPS (too few data points for single-day view to be meaningful). Long-term headcount trends (strategic view). Individual agent performance (managed in 1:1 cadence, not daily dashboard).
Tooling note
Most ACD and WFM platforms generate end-of-day reports automatically. Operations managers often review in email or a BI tool (Power BI, Tableau, Looker) rather than a live dashboard. Auto-distribution to the team is common.
Strategic / monthly view
Primary metrics
- ▸CSAT / NPS (rolling 4-week average)
- ▸FCR % (monthly trend)
- ▸Cost per contact (monthly)
- ▸Agent attrition rate (monthly rolling)
- ▸Headcount vs. plan
- ▸WFM forecast accuracy (WAPE, monthly)
- ▸Complaints and escalations (volume and rate)
- ▸FCA / regulatory breaches (if applicable)
Not for this view
Interval-level data (too granular for senior audience). Real-time or intraday metrics (not actionable at this level). Operational detail that leadership cannot act on — they need trends, not transactions.
Tooling note
Senior leadership dashboards are typically built in a BI tool (Power BI, Tableau, Google Looker Studio) pulling from ACD, WFM, CRM, and QA systems. Refresh weekly or monthly depending on business rhythm.
Leading vs. lagging indicators
Leading indicators
Change before the outcome they predict. Act on these to prevent a problem, not react to one.
Volume vs. forecast (current interval)
Predicts service level in 30–60 min
Agents in not-ready state
Predicts service level in 15–30 min
AHT vs. plan (current interval)
Predicts capacity gap in next 30 min
Absence call-in rate (today, early shifts)
Predicts staffing shortfall in 2–4 hours
Lagging indicators
Confirm what has happened. Use for trend analysis and root cause investigation, not real-time action.
Service level % (yesterday / last week)
Confirms whether SL was met
CSAT / NPS score
Survey lag: hours to days
FCR % (daily/weekly)
Requires repeat-contact match (24-hr+ lag)
Agent attrition rate (monthly)
Lags the causes (occupancy, wellbeing) by 3–6 months
Five commonly misused contact centre metrics
AHT as a hard KPI
Why it goes wrong
When AHT is a performance target with consequences, agents optimise for the number by cutting calls short. FCR falls, repeat contacts rise, CSAT drops. The cost of the repeat contacts almost always exceeds any saving from shorter AHT.
Better approach
Use AHT as a diagnostic indicator, not a target. Monitor AHT variance (not the absolute) to identify knowledge gaps, inefficient call structures, or system issues. Investigate outliers; do not penalise.
Average Speed of Answer (ASA) as a substitute for service level
Why it goes wrong
ASA is the mean wait time across all contacts. It is heavily distorted by a small number of very long waits. An ASA of 30 seconds can coexist with 40% of contacts waiting over 2 minutes — the long waits are hidden in the average. ASA tells you nothing about whether SL targets were met.
Better approach
Use service level % (% of contacts answered within X seconds) as the primary queue metric. ASA can be a secondary indicator but should never replace SL %.
Occupancy as an agent efficiency metric
Why it goes wrong
High occupancy (above 85–88%) is not a sign of efficiency — it is a sign of understaffing. Occupancy is a consequence of volume and headcount, not agent effort. Using occupancy as an individual performance KPI leads to the wrong conclusion.
Better approach
Occupancy is a workforce planning input (used to assess burnout risk and staffing efficiency), not an agent performance metric. Monitor at team and operation level, not individual. See the occupancy guide for thresholds.
Calls handled / contacts handled count
Why it goes wrong
Volume handled by an agent is partly a function of AHT (short AHT = more contacts per hour), partly a function of schedule (more hours worked = more contacts), and partly a function of queue depth (high volume periods = more contacts). It measures none of these independently.
Better approach
Use contacts handled only in aggregate (for capacity and cost modelling), not as an individual performance KPI. For individual productivity, use adherence % and AHT trend, both adjusted for contact mix.
Schedule adherence % as the primary performance metric
Why it goes wrong
Adherence measures whether agents followed their scheduled activities — it does not measure quality, customer outcomes, or genuine effort. An agent with 98% adherence who provides wrong information on every call looks excellent by this measure.
Better approach
Adherence is a compliance metric, not a performance metric. It should be tracked and managed (deviations investigated), but performance assessment must include quality, FCR, and CSAT alongside adherence.
Dashboard design questions
What metrics should be on a contact centre real-time dashboard?
Current service level %, contacts in queue/longest wait, agents available vs. scheduled, agents not-ready count, abandon rate (current interval), and volume vs. forecast. What should NOT be on the real-time wall board: AHT, FCR, CSAT, NPS, or any metric agents cannot influence in the next 15 minutes.
What is the difference between a KPI and a metric in a contact centre?
A metric is any measured value. A KPI is a metric with an active target being managed toward. Contact centres typically track 30–50 metrics but manage 8–12 KPIs. Over-KPI-isation causes gaming: agents optimise the measured target at the expense of unmeasured outcomes (classic example: hard AHT KPI drives call-cutting, FCR falls, repeat contacts rise).
Related guides
CC metrics guide
Full metrics reference
Reporting guide
Report structure and cadence
Intraday management
Using real-time data to re-plan
Real-time adherence
RTA tools and thresholds
CC benchmarks 2025
Industry benchmark reference
WFM ROI guide
Quantifying the value of good WFM data
Forecast accuracy calculator
WAPE — the primary metric for the WFM section of the dashboard
Schedule adherence calculator
Adherence rate and its SL contribution for the dashboard