A practical guide to running a call center: forecasting and staffing, building schedules agents can live with, the metrics that matter, coaching and quality, and the wage and hour rules that catch call centers out.
Call centers combine unpredictable demand, tight service targets and repetitive, high-pressure work. Volume can change by hour, day and season; a marketing campaign or outage can double calls without warning. Agents need to be in their seats at the right times, which makes attendance and schedule adherence critical. And turnover tends to be high, so managers are always training new people. See call center agent motivation for the people side.
To record clock-ins, breaks and overtime automatically, see EZTime time and attendance software.
Staffing starts with a forecast of contacts per interval, usually 15 or 30 minutes. Use historical data by day of week and time of day, then adjust for known events such as promotions, billing cycles, holidays and product launches. Track forecast accuracy and refine it each week. Even a simple spreadsheet based on the last several weeks of data is far better than staffing by instinct.
Convert the forecast into the number of agents needed per interval. You need the expected number of contacts, the average handle time (talk time plus after-call work) and your service target. Many centers use the Erlang C model or workforce management software for this calculation; our staffing calculator helps with the basics.
Then add shrinkage: time when scheduled agents are not available to take contacts because of breaks, meetings, training, coaching, absences and time off. If shrinkage is 30 percent and you need 14 agents on the phones, you need to schedule 20 (14 ÷ 0.70).
Call center schedules must cover peaks while staying fair and predictable for agents. Practical approaches include:
Predictable schedules reduce absenteeism, which in turn protects service levels.
| Metric | What it measures | Watch out for |
|---|---|---|
| Service level | Share of contacts answered within a target time | Hitting the target by over-staffing at high cost |
| Abandonment rate | Share of callers who hang up before reaching an agent | Spikes during understaffed intervals |
| Average handle time (AHT) | Talk time plus hold plus after-call work | Pushing AHT down can hurt resolution and quality |
| First contact resolution (FCR) | Issues resolved without a follow-up contact | Hard to measure; use surveys or repeat-contact tracking |
| Customer satisfaction (CSAT) | Customer ratings after contact | Low response rates can skew results |
| Schedule adherence | Whether agents are available when scheduled | Rigid enforcement can damage morale |
| Occupancy | Share of logged-in time spent handling contacts | Sustained very high occupancy leads to burnout |
| Absenteeism | Unplanned absence rate | Often a sign of scheduling or morale problems |
Pick a handful that reflect your goals and review them together; optimizing one at the expense of others usually backfires. Our absenteeism rate calculator helps track attendance.
A support line forecasts 120 calls between 10:00 and 10:30 on Mondays, with an average handle time of 6 minutes. That is 720 minutes of work in a 30-minute interval, or 24 agents fully busy. Because agents cannot be 100 percent occupied without long queues, a staffing model for a target such as answering most calls within a short wait might call for several more, say 27. With shrinkage of 30 percent, the center needs to schedule about 39 agents for that interval (27 ÷ 0.70). Repeat this for each interval and you can see where to start and end shifts and where breaks can go without hurting service.
Monitoring a sample of calls or chats against a clear scorecard, and coaching agents on what you find, is the most direct way to improve both quality and efficiency. Keep scorecards short and focused on behaviors that matter to customers: understanding the issue, accuracy, resolution and courtesy. Coach regularly, in short sessions, with specific examples, and recognize improvement. See positive feedback examples.
Two DOL statements cover most call center pay questions about breaks and overtime:
A frequent call center issue is the ‘unpaid’ meal break that is interrupted. If an agent must stay logged in or answer escalations during lunch, they are not fully relieved of duty, and the time is generally paid. Make sure the phone system and the schedule agree on when agents are truly off.
Call center work involves several activities that are easy to leave unpaid, which is why call centers are a frequent subject of wage claims. For non-exempt agents:
Management checklist complete.
Rounding punches can also be risky if it consistently favors the employer. See time clock rounding.
Agent turnover is costly: each new agent needs recruiting, training and time to reach full productivity. The biggest levers are predictable schedules, fair treatment of adherence, realistic targets, regular coaching and recognition and a visible path to senior agent, team lead or quality roles. See how to reduce employee turnover and the cost of employee turnover.
EZTime schedules agents by interval and role, publishes schedules to phones, handles shift swaps and time off, and records clock-ins from the web or app so agents can clock in before booting up. Break tracking with alerts and labor rule alerts help keep pay compliant, and attendance reports show lateness and absence by agent. See EZTime for call centers.
Commonly used metrics include service level, abandonment rate, average handle time, first contact resolution, customer satisfaction, schedule adherence, occupancy and absenteeism. Review them together rather than optimizing one alone.
Forecast contacts per interval, multiply by average handle time to get workload, use a model such as Erlang C to find the agents needed for your service target and then add shrinkage for breaks, training and absence.
Generally yes. Time non-exempt agents spend starting computers and logging into required systems before taking calls is usually compensable work time.
The share of scheduled time when agents are not available to handle contacts, because of breaks, meetings, training, coaching, absence and time off. It must be added on top of the agents needed on the phones.
Offer predictable, fair schedules, realistic targets, regular coaching and recognition, flexibility through shift swaps and clear career paths.
Federal law does not require breaks, but short breaks of about 5 to 20 minutes that you do offer count as paid work time, including toward overtime. Bona fide meal periods, typically 30 minutes or more with the agent fully relieved of duty, can be unpaid. Some states require breaks.
A common target is answering a set share of calls within a set number of seconds, for example 80 percent in 20 seconds, but the right number depends on your customers and costs. Pick a target that matches what customers expect, then staff to it rather than copying another industry’s figure.
EZTime’s AI auto-scheduling on Advanced plans and up helps build shifts around demand, and agents see their schedules and swap shifts in the app.