erlang calculator

What is an Erlang calculator?

An Erlang calculator is a workforce planning tool used in call centers and customer support operations to estimate how many agents are needed to handle incoming demand. The term Erlang refers to a unit of traffic intensity: one Erlang is one hour of work arriving in one hour.

Most workforce teams use the Erlang C model to answer practical questions like:

• How many agents are needed to hit an 80/20 service level?
• What occupancy will my team run at if volume spikes?
• How long will customers wait before reaching an agent?

How this Erlang C calculator works

This calculator takes your volume forecast and handle time, converts them into offered load, and then searches for the smallest number of agents that meets both:

• Your service level target (for example, 80% answered within 20 seconds)
• Your maximum occupancy target (for example, 85%)

It also applies shrinkage to convert net required agents (productive time) into scheduled agents (rostered headcount).

Core equations used

Offered load (Erlangs) = Contacts per hour × AHT (hours)

Occupancy = Offered load ÷ Agents

Erlang C probability of waiting and service level are then computed with standard queueing formulas for an M/M/N queue with waiting allowed.

Step-by-step example

Suppose your forecast is 120 contacts per hour and your average handle time is 4.5 minutes.

• Offered load = 120 × (4.5/60) = 9 Erlangs
• That means you receive the equivalent of 9 hours of work every hour
• You need more than 9 agents logged in, or queues will grow indefinitely

If your target is 80% in 20 seconds with max occupancy 85%, the calculator will find the minimum staffing level that satisfies both conditions and then inflate that number by shrinkage.

Why shrinkage matters

Shrinkage includes breaks, coaching, meetings, training, absenteeism, and other non-queue activities. If you need 10 agents on the phones and your shrinkage is 30%, you cannot schedule only 10 people.

Scheduled agents are calculated as:

Scheduled = Net required ÷ (1 − shrinkage)

With 30% shrinkage, 10 net agents becomes 14.3, which rounds up to 15 scheduled agents.

Best practices for accurate Erlang forecasting

1) Use interval-level forecasts

Daily averages hide peaks. Erlang math works best when done in short intervals (typically 15 or 30 minutes), then aggregated into a staffing plan.

2) Keep AHT clean and current

AHT drift can materially change staffing recommendations. Split by queue or skill where possible.

3) Pair service level with occupancy guardrails

A plan that technically hits service level at 98% occupancy will burn out your team. Occupancy constraints are essential for sustainability.

4) Reforecast intraday

Even strong plans need intraday adjustments. Actuals diverge from forecast due to weather, incidents, marketing activity, or outages.

Limitations of Erlang C

Erlang C assumes callers wait indefinitely and no one abandons. Real queues have abandonment, callbacks, asynchronous channels, and multi-skill routing. So use this as a planning baseline, then refine with real operational behavior.

• For blocked systems with no queue, use Erlang B.
• For abandonment-aware modeling, consider Erlang A or simulation.
• For blended channels, use channel-specific assumptions and concurrency models.

Quick interpretation guide

• High occupancy + low service level: understaffed; add coverage or reduce demand.
• Low occupancy + high service level: overstaffed; consider cross-skilling or reallocation.
• High probability of wait: customers queue often; monitor ASA and abandonment closely.

Final thoughts

An Erlang calculator is one of the fastest ways to turn demand forecasts into staffing decisions. It gives planners a transparent, repeatable framework for balancing customer experience, cost, and agent workload.

Use it frequently, validate it against actual performance, and treat it as one part of a broader workforce management process that includes forecasting, scheduling, intraday control, and continuous improvement.