Calculate Staffing Requirements
Enter your call center metrics to determine optimal staffing levels
Advertisement
Understanding Call Center Staffing
Proper staffing is the foundation of successful call center operations. Too few agents leads to long wait times, abandoned calls, frustrated customers, and burnt-out staff. Too many agents wastes money on unnecessary labor costs while leaving agents idle and disengaged. Finding the right balance requires understanding call volumes, handle times, service level targets, and workforce management principles.
Call center staffing calculations use queuing theory, particularly the Erlang C formula, to predict how many agents are needed to meet specific service levels. This mathematical approach accounts for the random nature of incoming calls and the probability that all agents will be busy simultaneously. The formula considers call arrival rates, average handle time, and desired service levels to calculate optimal staffing.
Workforce management extends beyond simple calculations to include scheduling, forecasting, real-time adherence, and shrinkage planning. Effective staffing strategies balance customer service goals with operational efficiency and employee satisfaction. This calculator provides baseline staffing requirements, which should be adjusted for your specific operational realities like breaks, training time, meetings, and system issues.
Key Staffing Metrics Explained
Service level is the most common call center metric, typically expressed as a percentage of calls answered within a specific timeframe. The industry standard "80/20" service level means 80% of calls are answered within 20 seconds. Different industries and call types may have different targets—emergency services aim for 90/10, while back-office functions might accept 70/60. Your service level target significantly impacts required staffing levels.
Average Handle Time (AHT) combines talk time and after-call work time to represent the total time an agent invests in each call. AHT directly affects capacity—if AHT increases by one minute, each agent handles fewer calls per hour, requiring more staff for the same volume. Monitoring AHT trends helps identify training needs, process issues, or changing call complexity.
Occupancy measures the percentage of time agents spend handling calls versus being available but idle. While 100% occupancy might seem ideal for efficiency, it's actually problematic. High occupancy (above 85-90%) causes agent stress, burnout, and quality issues. Optimal occupancy ranges from 75-85%, providing breathing room between calls while maximizing productivity.
Shrinkage accounts for all time agents are paid but unavailable to take calls—breaks, training, meetings, coaching, system issues, and absences. Typical shrinkage ranges from 25-35% of scheduled time. If you need 20 agents on the phones, you might need to schedule 27 agents to account for 26% shrinkage. Accurately tracking and managing shrinkage is crucial for effective staffing.
The Erlang C Formula and Workforce Planning
The Erlang C formula, developed by Danish mathematician Agner Krarup Erlang in the early 1900s for telephone networks, remains the foundation of modern call center staffing. It calculates the probability that a caller will need to wait when all agents are busy, based on traffic intensity (measured in Erlangs), number of agents, and target service level. While the mathematics are complex, the concept is straightforward: more calls and longer handle times require more agents.
Traffic intensity, measured in Erlangs, represents the workload on your system. One Erlang equals one circuit or agent being continuously occupied. If you receive 60 calls per hour with 5-minute average handle time, that's 300 call-minutes or 5 Erlangs (300 minutes / 60 minutes). You'd need at least 5 agents just to handle the workload, plus additional agents to maintain service levels during call arrival fluctuations.
Modern workforce management software automates these calculations and adds sophistication through interval-based forecasting (typically 15 or 30-minute intervals), multi-skill routing, and real-time adjustments. However, understanding the fundamentals helps managers make informed decisions and recognize when automated recommendations seem incorrect.
Forecasting Call Volume Accurately
Accurate staffing depends on accurate forecasting. Call volumes vary by hour, day, week, month, and season. Monday mornings differ from Friday afternoons. January differs from December. Understanding these patterns through historical analysis enables better predictions. Most call centers use time-series forecasting methods that identify trends, seasonality, and day-of-week effects.
Special events significantly impact call volume and must be factored into forecasts. Marketing campaigns, product launches, billing cycles, service outages, and external factors (weather, holidays, economic news) can cause substantial volume spikes. Maintaining a forecast calendar that documents these events and their historical impacts improves prediction accuracy over time.
Forecast accuracy should be measured and continuously improved. Compare predicted volumes against actual volumes and analyze variances. Typical call centers achieve 90-95% forecast accuracy for stable operations, though this varies by industry and business volatility. Even small forecast errors compound when calculating staffing—a 5% volume underestimate might require 10% more agents due to service level degradation.
Scheduling Best Practices
Converting staffing requirements into actual schedules involves balancing operational needs with agent preferences and labor regulations. Most call centers use shift bidding, where agents select preferred schedules based on seniority, or automated scheduling software that optimizes coverage while respecting constraints like maximum consecutive days, minimum rest periods, and skill requirements.
Shift start times should be staggered to match call arrival patterns rather than having all agents start simultaneously. If call volume peaks from 10am-2pm, schedule more agents with start times that ensure maximum coverage during those hours. Split shifts, where agents work two shorter periods with a long break between, can effectively cover extended hours without excessive overtime.
Real-time adherence monitoring tracks whether agents are working their scheduled activities (taking calls, on break, in training) or deviating from the schedule. Schedule adherence percentages above 95% are excellent, 90-95% acceptable, and below 90% problematic. Address chronic non-adherence through coaching, as it forces other agents to handle excess volume and degrades service levels.
Managing Variable Call Volume
Call volumes rarely match forecasts exactly. Real-time management involves monitoring actual versus forecasted volumes and making immediate adjustments. If call volume exceeds predictions, options include reassigning agents from back-office work, offering overtime, using overflow agents from other sites, or temporarily accepting lower service levels if capacity is maxed out.
When volume falls below forecast, avoid sending agents home immediately. Small sample sizes early in shifts can mislead—an unexpectedly slow first hour might be followed by a busy afternoon. However, if sustained low volume occurs, offering voluntary time off (VTO) helps control costs while maintaining agent goodwill. Some agents appreciate the flexibility, though consistent VTO can harm morale if agents rely on scheduled hours for income.
Peak period strategies help manage predictable volume spikes. Schedule all-hands-on-deck periods where non-phone work is minimized. Cross-train employees from other departments to assist during peaks. Implement flexible break scheduling that avoids breaks during busy periods. Consider premium pay or incentives for agents willing to work the most challenging shifts.
Multi-Skill and Multi-Channel Considerations
Modern contact centers handle multiple channels (phone, email, chat, social media) and multiple contact types (sales, service, technical support). Multi-skill routing allows agents to handle different contact types based on their training and availability, improving efficiency. However, multi-skilling complicates staffing because different channels have different handle times, service level targets, and concurrency (chat agents can handle multiple simultaneous conversations).
Blended environments require workforce management systems that can calculate requirements across channels and optimize agent allocation dynamically. An agent might take phone calls when available, but automatically receive chat or email assignments during slower phone periods. This flexibility improves occupancy and service levels across all channels simultaneously.
Cost Considerations in Staffing Decisions
Labor typically represents 60-70% of call center operating costs. Staffing decisions have enormous financial impact. However, understaffing to save money often backfires through higher turnover (replacing agents is expensive), lower customer satisfaction (which costs future revenue), and agent burnout (which reduces productivity). The optimal staffing level balances service quality with cost efficiency.
Calculate the cost per contact by dividing total labor costs by contacts handled. This metric helps evaluate the financial impact of staffing changes. If adding 2 agents costs 50,000 annually but improves service level from 70% to 85%, calculate whether the customer satisfaction improvement justifies the cost. Some organizations find that investing in slightly higher staffing reduces downstream costs like complaints, escalations, and customer churn.
Frequently Asked Questions
How many agents do I need for my call center?
The number of agents required depends on several factors: expected call volume, average handle time, target service level, and hours of operation. As a basic calculation, divide your total call minutes per hour by 60 to get Erlangs (traffic intensity), then add 15-25% for service level buffer. For example, 100 calls per hour with 5-minute average handle time equals 500 call-minutes or 8.33 Erlangs, suggesting you need approximately 10-11 agents to maintain good service levels. However, this simplified approach doesn't account for call arrival variability, which is why proper Erlang C calculations or workforce management software provides more accurate results. Your specific requirements also depend on your service level goals—meeting 80/20 requires fewer agents than 90/10. Additionally, don't forget to factor in shrinkage for breaks, training, absences, and other non-phone activities. If your calculation suggests 10 agents on the phones, you might need to schedule 13-14 agents to account for typical shrinkage of 25-30%. Start with these baseline calculations, then monitor actual performance and adjust staffing based on real results. Track metrics like service level, abandonment rate, and average speed of answer to determine if you're staffed appropriately.
What is a good service level for a call center?
The most common service level target in call centers is 80/20—answering 80% of calls within 20 seconds. This has become an industry standard because it balances customer satisfaction with operational efficiency. However, "good" service levels vary significantly by industry, call type, and customer expectations. Emergency services and urgent support lines often target 90/10 or even higher, because immediate response is critical. Sales organizations might accept 70/30 because callers are highly motivated and willing to wait. Back-office or tier-2 support might use 80/60 since these contacts are often scheduled callbacks or less time-sensitive. B2B call centers serving business customers often need higher service levels (85/15 or 90/20) than B2C operations. Consider your specific customer base—premium customers or high-value accounts might warrant tighter service levels than general inquiries. Also consider the competitive landscape and customer alternatives. If competitors answer instantly, 80/20 might lose you business. Monitor the relationship between service level and customer satisfaction in your operation. Use post-call surveys or NPS scores to determine if your current service level meets customer expectations. Don't set unrealistic targets that require excessive staffing—90/15 sounds great but might cost 40% more in labor than 80/20 for minimal satisfaction improvement.
What is shrinkage in call center staffing?
Shrinkage represents all time that agents are scheduled and paid but unavailable to handle customer contacts. It includes scheduled activities like breaks, lunches, training, coaching, team meetings, and one-on-ones, plus unscheduled activities like absences, tardiness, extended breaks, and system downtime. Typical call center shrinkage ranges from 25-35%, meaning if you need 20 agents actually taking calls, you must schedule approximately 27-30 agents to account for shrinkage. Calculating shrinkage accurately is essential for effective staffing. Track each component: if agents work 8-hour shifts with two 15-minute breaks and a 30-minute lunch, that's 1 hour or 12.5% scheduled shrinkage just for breaks. Add 1 hour weekly per agent for training (2.5% for a 40-hour week), plus team meetings, coaching, and other scheduled activities. Unscheduled shrinkage depends on your organization—some have 2-3% absenteeism, others experience 8-10%. System issues, extended breaks, and other factors add more shrinkage. Total it up and you'll understand your true shrinkage percentage. Managing shrinkage involves balancing business needs with agent wellbeing. You can't eliminate breaks and training, nor should you—these are necessary. However, you can minimize unproductive shrinkage through clear policies, attendance management, reliable technology, and efficient meeting practices. Monitor shrinkage trends and investigate sudden increases, which often indicate morale issues, system problems, or policy gaps requiring attention.
How do I calculate occupancy rate for call center agents?
Occupancy rate measures the percentage of time agents spend actively handling contacts versus being available but idle. Calculate it by dividing total handle time (talk time plus after-call work) by total available time (logged-in time minus breaks), then multiply by 100 for a percentage. For example, if an agent is available for 7 hours (420 minutes) after deducting breaks, and spends 340 minutes in talk time and after-call work, occupancy is (340/420) × 100 = 81%. This indicates the agent was productively engaged 81% of available time, with 19% idle between calls. While higher occupancy seems better for efficiency, occupancy above 85-90% becomes problematic. Agents need brief moments between calls to mentally reset, update systems, and prepare for the next contact. Sustained high occupancy causes stress, burnout, errors, and turnover. Most experts recommend target occupancy of 75-85% for optimal balance between productivity and agent wellbeing. Occupancy varies inversely with service level—if you improve service level by adding agents, occupancy decreases because the same workload is distributed across more people. This is expected and acceptable. Low occupancy (below 70%) suggests overstaffing, which wastes resources and can demotivate agents who feel underutilized. However, temporary low occupancy during training periods, slow seasons, or after adding new agents is normal. Monitor occupancy trends rather than fixating on single-day measurements, as call volume variability causes occupancy fluctuations even with proper staffing.
Should I use in-house staff or outsource my call center?
The decision between in-house and outsourced call centers depends on multiple factors including cost, control, quality requirements, volume, and strategic importance. In-house operations provide maximum control over agent training, quality standards, processes, and customer experience. Your agents are part of your organization, understand your culture, and can be deeply trained on your products and values. This typically produces higher quality but at higher cost—you bear all infrastructure, technology, management, and labor expenses. In-house makes sense when customer interaction is a key brand differentiator, when you handle sensitive information requiring tight security, when call volume justifies the fixed costs, or when you need rapid process changes that outsourcers couldn't accommodate quickly. Outsourcing reduces fixed costs by converting them to variable costs—you pay per minute, per call, or per agent hour without bearing overhead. Outsourcers provide flexibility to scale up or down quickly for seasonal variations or business changes. They offer expertise in workforce management, quality assurance, and call center best practices. However, you sacrifice some control, and agent turnover is typically higher in outsourced environments, affecting quality and customer relationships. Many organizations use hybrid approaches: keeping premium customer segments, complex issues, or sales in-house while outsourcing high-volume, lower-complexity contacts. This balances cost with quality. Some use outsourcers for overflow during peaks while maintaining core in-house operations. Evaluate total cost of ownership, not just hourly rates—include training, quality, and the impact on customer satisfaction and retention.
What is the Erlang C formula and why is it important?
The Erlang C formula is a mathematical model used to calculate the probability of call queuing in systems with multiple servers (agents) where callers wait if all servers are busy. Developed by A.K. Erlang in the early 1900s for telephone networks, it remains the foundation of call center workforce management today. The formula accounts for call arrival rates following a Poisson distribution (random arrivals), exponential service times (call durations), and a specified number of agents. It calculates the probability that a caller must wait when all agents are occupied, which then determines how many agents are needed to achieve service level targets. The importance of Erlang C lies in its ability to optimize staffing scientifically rather than through guesswork. Without it, managers either overstaff (wasting money) or understaff (providing poor service). The formula reveals that staffing requirements don't scale linearly—doubling call volume doesn't simply require doubling agents because of statistical pooling effects. Larger centers are more efficient per agent than smaller ones. The formula also demonstrates why service level improvements require disproportionate staffing increases—going from 70% to 80% service level might add 2 agents, but 80% to 90% might add 4 more agents. Modern workforce management software embeds Erlang C calculations and adds sophistication through interval-based forecasting, multi-skill routing, and real-time optimization. However, understanding Erlang C fundamentals helps managers interpret recommendations, recognize when results seem wrong, and make informed staffing decisions during unusual circumstances.
How do I forecast call center staffing for seasonal variations?
Forecasting seasonal staffing requires analyzing historical patterns to predict future volume fluctuations, then calculating staffing requirements for each period. Start by gathering at least two years of historical call volume data at the interval level (15 or 30-minute intervals throughout each day). Two years helps identify consistent patterns versus anomalies. Plot this data to visualize seasonal trends—many call centers see December holiday peaks, January post-holiday surges, summer slowdowns, or patterns tied to fiscal cycles like tax season or billing periods. Use time-series forecasting methods that decompose volume into trend, seasonality, day-of-week effects, and random variation. Many workforce management systems automate this using algorithms like Holt-Winters or ARIMA. Alternatively, simple approaches like calculating the percentage variance from average for each period works reasonably well. For example, if December call volume averages 130% of the annual average while July averages 75%, apply these factors to your base staffing plan. Don't forget to document and factor in special events beyond regular seasonality—marketing campaigns, product launches, system changes, and external factors that drove unusual volume in the past. Maintain a forecast calendar noting these events and their impact for future reference. For seasonal peaks, begin recruiting and training well in advance—most call center agents need 3-4 weeks of training before reaching full productivity. Consider using temporary staff, overtime, outsource partners, or work-from-home agents specifically for peak periods. Build relationships with staffing agencies or seasonal workforce pools you can tap regularly.