Workforce management is one of the highest-leverage disciplines in contact center operations. Done well, it aligns staffing precisely to demand — keeping service levels consistent without over-hiring. Done poorly, it produces the two most expensive outcomes in call center management: agents sitting idle during low-volume periods, and callers abandoning during understaffed peaks. This guide covers every stage of the WFM cycle from forecasting through adherence monitoring, the metrics that matter, and how modern cloud contact centers handle WFM.
What Is Workforce Management in a Call Center
Workforce management in a call center context is the process of matching agent supply to contact demand, interval by interval, across every operating hour. The goal is to have the right number of agents — with the right skills — available at the right times to deliver a consistent service level while keeping labour costs in line with the business.
WFM is not a single tool or a single department function. It is a continuous operational cycle that runs through four stages: forecasting (predicting how many contacts will arrive and when), scheduling (building shifts that meet the forecast), intraday management (monitoring real-time adherence and making adjustments during the day), and post-period analysis (comparing predicted versus actual to improve future forecasts).
In small contact centers, WFM is often handled manually using spreadsheets. As headcount grows, the complexity of scheduling constraints — shift preferences, skill groups, part-time vs. full-time ratios, training time, break schedules — makes manual scheduling increasingly error-prone. This is the point where purpose-built WFM software typically delivers the most visible return.
See: MCC Contact Center for a cloud contact center platform with integrated real-time reporting that feeds WFM decisions.
Forecasting Call Volume
Forecasting is the foundation of the entire WFM cycle. If the forecast is wrong, the schedule built on top of it will be wrong, and no amount of intraday management fully compensates for a systematically inaccurate forecast.
The standard approach to call volume forecasting uses historical call data as its base. Most contact centers work with at least 13 months of historical volume by interval — enough to capture seasonal patterns, day-of-week variation, and year-over-year trends. The forecast model identifies repeating patterns and projects them forward, then adjusts for known upcoming events: marketing campaigns, product launches, billing cycle dates, public holidays, or any other factor expected to shift volume above or below the baseline trend.
Forecast granularity matters. A daily forecast tells you roughly how many calls to expect on a given day, but a staffing schedule needs to know the distribution within that day — typically by 15- or 30-minute intervals. A contact center that averages 400 calls per day but has 60% of them arrive between 10:00 and 14:00 needs a very different staffing shape than one with flat arrival throughout the day.
Forecast accuracy is measured by comparing predicted volume to actual volume after the fact, expressed as a percentage variance per interval. No forecast is perfectly accurate, so WFM planning always accounts for variability through intraday adjustments. What matters is reducing systematic bias — forecasts that are consistently high or consistently low in particular intervals, which compound into structural overstaffing or understaffing.
Scheduling and Shift Planning
Once the volume forecast is complete, the scheduling process translates it into agent schedules. The core question is: how many agents do you need in each interval to meet your service level target? The answer comes from applying a staffing model — most commonly Erlang C — to the forecast volume, average handle time, and target service level for each interval.
Erlang C calculates the minimum number of agents needed to handle a given arrival rate with a given average handle time, such that a target percentage of calls are answered within a target number of seconds. The result is a staffing requirement curve across the day — the minimum agent headcount needed in each interval. The scheduling process builds shifts that cover that curve as closely as possible, subject to real-world constraints: shift length minimums and maximums, contracted hours, break regulations, part-time and full-time ratios, and agent skill requirements.
Shrinkage is a critical input to the scheduling calculation. Shrinkage represents the portion of scheduled time during which agents are unavailable to handle contacts — breaks, meals, training, team meetings, coaching sessions, and unplanned absences. If your shrinkage rate is 30%, you need to schedule roughly 1.43 agents for every agent you need on the phones, to end up with the required staffed headcount after shrinkage. Underestimating shrinkage is one of the most common causes of chronic understaffing against forecast.
Shift planning also involves decisions about flexibility: how many agents are on fixed shifts versus flexible or split shifts, how much overlap to build in at shift boundaries, and whether to maintain a reserve of part-time agents who can be called in on high-volume days. These decisions reflect a trade-off between scheduling efficiency and agent preference, and they vary considerably by contact center type and workforce composition.
Real-Time Adherence Monitoring
Schedule adherence monitoring compares what agents are actually doing in real time against what their schedule says they should be doing. The goal is not punitive oversight — it is operational visibility. If adherence drops materially during a period, it means actual staffed headcount is below the scheduled level, which typically leads to queue build-up, rising wait times, and deteriorating service levels.
Modern WFM platforms pull agent state data from the contact center platform in real time: available, on a call, in after-call work, on break, in training, logged out. This is compared against the agent's scheduled state for each interval. Adherence is expressed as a percentage of scheduled time during which the agent was in the expected state.
Supervisors use real-time adherence dashboards to identify when agents are off-task and to make intraday adjustments: moving scheduled breaks to cover unexpected volume spikes, pulling agents from non-phone activities back to the queue, or requesting voluntary time off during unexpectedly low-volume periods. The faster a supervisor can identify an adherence problem and respond, the smaller its impact on service level.
It is important to distinguish between adherence (following the schedule) and compliance (being in the right place at the right time). Some WFM systems track both, noting not just whether an agent was in the right state but whether they were in the right state at the right time within their schedule. Both metrics are useful for different purposes — aggregate adherence for staffing analysis, compliance for individual coaching.
Key WFM Metrics
WFM generates a set of operational metrics that together describe how well staffing is matching demand. Understanding what each measures and how they interact is essential for managing the WFM cycle effectively. See: call center metrics reference for a broader list of contact center KPIs.
| Metric | What it measures | Why it matters |
|---|---|---|
| Service Level | % of calls answered within target time (e.g., 80% in 20 seconds) | Primary customer experience indicator; the target the schedule is built to achieve |
| Occupancy | % of logged-in time agents spend on calls or ACW | High occupancy reduces idle cost but increases agent fatigue; very high occupancy leaves no buffer for volume spikes |
| Schedule Adherence | % of scheduled time agents were in the expected state | Closing the gap between planned and actual staffing; needed to validate schedule assumptions |
| Shrinkage | % of scheduled time agents are unavailable for contacts | Drives the gross staffing multiplier applied to net staffing requirements |
| Forecast Accuracy | Variance between predicted and actual call volume by interval | Identifies systematic bias in the forecast model; drives improvements to forecasting methodology |
| Average Handle Time | Average talk time + after-call work per contact | Direct input to Erlang C staffing calculations; changes in AHT shift staffing requirements |
See also: first call resolution and average handle time — both metrics that interact directly with WFM staffing calculations.
WFM Software vs Manual Scheduling
Many contact centers start with manual scheduling — typically spreadsheets — and move to purpose-built WFM software at some point as headcount and complexity grow. Understanding where the boundary lies helps operations leaders time the transition appropriately.
Manual scheduling is tractable for small, relatively homogeneous teams with simple shift patterns. If you have 10 to 20 agents, one or two skill groups, standard shift lengths, and predictable volume, a well-designed spreadsheet with an embedded Erlang C calculator can produce adequate schedules. The limitation is not computational — it is the ability to handle multiple intersecting constraints simultaneously and to update schedules rapidly as conditions change.
WFM software adds value in proportion to complexity. A contact center with 50 or more agents, multiple skill groups, blended inbound and outbound, significant part-time staff, and variable shift patterns will find that manual scheduling consumes disproportionate planner time, produces more errors, and responds more slowly to intraday changes. WFM software automates the constraint-solving that makes multi-skill, multi-shift scheduling tractable at scale.
Beyond scheduling generation, WFM software typically includes automated adherence monitoring, intraday management tools, long-range capacity planning, and historical reporting on forecast accuracy and schedule effectiveness. These capabilities are hard to replicate in spreadsheets because they require real-time integration with the contact center platform to pull live agent state data and compare it to schedule data continuously.
How Cloud Contact Centers Handle WFM
Cloud contact center platforms have changed WFM in two important ways. First, they make the underlying data — call volume by interval, agent states, handle times, queue metrics — available in real time through APIs and reporting tools, which simplifies the integration work required to feed WFM systems. Second, many modern cloud platforms offer native WFM modules or certified integrations with leading WFM tools, reducing the custom integration work that was historically required.
In a cloud contact center environment, real-time adherence monitoring is typically handled by pulling agent state data from the platform via API or native integration. The WFM system compares this to the schedule in real time and surfaces exceptions to supervisors through dashboards or alerts. Because cloud platforms expose this data through standardised interfaces, the integration is generally more reliable and easier to maintain than legacy on-premise integrations.
Cloud platforms also simplify WFM for distributed and remote workforces. Because agent state data flows through the same platform regardless of where the agent is physically located, the WFM system sees remote agents and office agents identically. This matters for contact centers that operate across multiple sites or run hybrid work models — the scheduling and adherence process does not change based on agent location.
See: MCC Contact Center for a cloud platform with real-time reporting and agent state visibility that integrates with WFM workflows.
Looking for a cloud contact center platform with the real-time reporting your WFM process needs?
Explore MCC Contact CenterFAQ
What is workforce management in a call center?
Workforce management (WFM) in a call center is the discipline of ensuring the right number of agents with the right skills are available at the right times to meet anticipated call volume while keeping costs under control. It covers four main activities: forecasting how many calls or contacts will arrive by interval, scheduling agents to match that forecast, monitoring adherence in real time to ensure agents are following their schedules, and analysing outcomes to improve future forecasts. WFM is a continuous cycle, not a one-time planning exercise.
What metrics does WFM track?
The core WFM metrics are service level (the percentage of calls answered within a target time), occupancy (the proportion of time agents spend on calls or after-call work versus waiting), schedule adherence (how closely agents follow their assigned schedules), shrinkage (time lost to breaks, training, meetings, and unplanned absences), and forecast accuracy (how closely predicted volume matched actual volume). Most WFM platforms also track average handle time and after-call work time, since both feed directly into staffing calculations.
How does WFM software improve service levels?
WFM software improves service levels by replacing manual guesswork with data-driven forecasting and scheduling. It applies historical volume patterns, trend analysis, and seasonality adjustments to predict call arrival by 15- or 30-minute interval. It then uses an Erlang C or similar staffing model to calculate exactly how many agents are needed in each interval to hit a target service level. Real-time adherence monitoring alerts supervisors when actual staffing deviates from the schedule, so corrective action can happen during the day rather than after the fact. The cumulative effect is fewer understaffed intervals, lower abandon rates, and more consistent service levels.