Call Center Staffing Calculator

Call Center Staffing Calculator - Erlang C Formula

📞 Calculate required call center operators using Erlang C formula. Optimize staffing based on call volume, average handle time, service level targets, and shrinkage factors.

What is Erlang C?

Erlang C is a mathematical formula used in call center workforce management to calculate the number of agents needed to handle a given call volume while maintaining a target service level. It accounts for queuing theory and assumes calls that cannot be immediately answered wait in a queue.

Key Metrics Explained

• Calls per Hour: Total incoming calls during peak hour
• Average Handle Time (AHT): Talk time + after-call work (wrap-up)
• Service Level: % of calls answered within target time (e.g., 80/20)
• Occupancy: % of time agents are actively on calls (not idle)
• Shrinkage: Breaks, training, meetings, absences (non-productive time)

Traffic Intensity (Erlang)

Formula: E = (Calls per Hour × AHT in seconds) / 3600

Traffic intensity measures the workload in Erlangs. One Erlang = one agent continuously busy for one hour.

• Example: 100 calls/hr × 180 sec AHT = 18,000 / 3600 = 5 Erlangs
• Means 5 agents working non-stop could handle the load (theoretical minimum)

Required Agents Calculation

Step 1 - Base Agents (from Traffic Intensity):

Minimum Agents = Traffic Intensity (E)

Step 2 - Add for Service Level (Erlang C):

Erlang C formula determines additional agents needed to achieve service level target. This is iterative - we test different agent counts until service level is met.

Step 3 - Adjust for Occupancy:

Agents Needed = E / Target Occupancy

Step 4 - Add Shrinkage:

Total Staff = Agents / (1 - Shrinkage %)

Calculation Example

Given:

• 100 calls/hour
• 180 seconds AHT (3 minutes)
• 80% service level in 20 seconds
• 85% occupancy target
• 30% shrinkage

Calculation:

• Traffic Intensity: (100 × 180) / 3600 = 5 Erlangs
• Base agents: 5 / 0.85 = 5.88 ≈ 6 agents (minimum)
• For 80/20 SLA, need additional buffer → ~12-15 agents
• With 30% shrinkage: 15 / 0.70 = 21.4 ≈ 22 total staff

Service Level Standards

• 80/20: 80% of calls answered in 20 seconds (industry standard)
• 90/20: 90% in 20 seconds (premium service)
• 70/30: 70% in 30 seconds (basic service)
• 80/30: Common for technical support

Occupancy Rate Guidelines

• 80-85%: Optimal for agent well-being
• 85-90%: Acceptable, cost-efficient
• 90%+: Too high, leads to burnout
• <70%: Inefficient, excess capacity

Shrinkage Components

• Breaks: Lunch, coffee breaks (10-15%)
• Training: Ongoing development (5-10%)
• Meetings: Team meetings, 1-on-1s (3-5%)
• Absences: Sick leave, vacation (5-8%)
• System issues: Technical problems (2-5%)
• Total typical: 25-35%

Call Distribution Patterns

Call volume varies throughout the day. Typical pattern:

• 9-11 AM: Peak morning (100% of average)
• 11 AM-1 PM: Lunch dip (70-80%)
• 2-4 PM: Peak afternoon (90-100%)
• After 5 PM: Lower volume (40-60%)

Improving Call Center Efficiency

• Reduce AHT: Better training, scripts, tools
• Skill-based routing: Right agent for right call
• Self-service: IVR, chatbots for simple queries
• Flexible scheduling: Match staffing to call patterns
• Forecasting: Predict volume spikes (seasons, campaigns)

Common Mistakes

• Ignoring shrinkage: Always account for non-productive time
• Using average volume: Staff for peak hour, not daily average
• Forgetting AHT includes wrap-up: Not just talk time
• Setting occupancy too high: Leads to agent burnout
• No buffer for variance: Real calls don't arrive evenly

Advanced Considerations

• Multi-skill agents: Agents handling multiple queues
• Abandonment rate: % of callers who hang up while waiting
• Call-back options: Reduce queue wait time perception
• Interval planning: 15-30 minute intervals vs hourly
• Real-time adherence: Agents following schedule

💡 Pro Tip: Don't just calculate for average conditions! Run scenarios for your busiest hour of the busiest day of the week. Most call centers experience 20-30% variance in volume. Build in a safety buffer of 1-2 extra agents beyond what Erlang C suggests, especially if you have unpredictable call spikes. It's cheaper to have slightly excess capacity than to lose customers to long wait times!