What Is an Erlang? VICIdial Forecasting Explained Simply

When you open VICIdial's forecasting reports you will see a column called Erlangs, and it is the number that quietly drives every staffing estimate. The word looks intimidating, but the idea is simple once you see it. This post defines it in plain English with a worked example.

For the bigger picture of where forecasting fits among the reports, the VICIdial reports guide is a good starting point.

The one-sentence definition

An Erlang is a unit of telephone traffic. One Erlang equals one line kept busy for one full hour. It is not a count of calls and it is not a rate of speed. It is a measure of load: how much continuous talk-and-wait activity is happening over a span of time.

Two calls that each last 30 minutes in an hour add up to the same one Erlang as one call that lasts the full 60 minutes. Same load, different shape. The unit lets you compare an hour of many short calls against an hour of a few long ones on equal footing, which is exactly what staffing math needs.

A simple example

Say 100 inbound calls land on your Ingroup in one hour and the average call lasts 6 minutes. Six minutes is one tenth of an hour, or 0.1. Multiply: 100 calls times 0.1 hour gives 10 Erlangs of traffic for that hour. That is exactly how VICIdial fills the Erlangs column.

flowchart LR
  A[100 calls in the hour] --> C[Multiply]
  B[6 min avg = 0.1 hour] --> C
  C --> D[10 Erlangs of traffic]

Notice both pieces matter. More calls raise the Erlangs, and longer calls raise them too. Trim your Average handle time (AHT) and the same call count produces fewer Erlangs, which means you need fewer agents to carry it. One subtlety in how VICIdial counts: a call that begins in one hour and ends in the next has its time split across both intervals, so its traffic shows up in each. That keeps the per-hour Erlang figure honest even for long calls that straddle the clock.

Why VICIdial cares

Erlangs are the input to the staffing math. Once the report knows the traffic load in Erlangs and you have entered a target drop rate, it can work out how many agents keep your Service level where you want it. Without converting calls and durations into Erlangs first, there is no clean way to compare a busy noon hour against a slow morning.

It also explains a common surprise: a Call queue with fewer calls can still need more agents than a busier one if its calls run much longer. The Erlang figure captures that, where a raw call count would not. This is why two teams with the same number of calls per day can need very different headcounts, and why staffing off call counts alone tends to under-serve the longer-call queues.

Where you will see it

Erlangs appear in both the Inbound Forecasting Report and the Advanced Forecasting Report. The Advanced version layers cost and revenue math on top, which is covered in the inclusive SLA per day report and its neighbors. Either way, the Erlang figure is the shared foundation.

Get comfortable reading Erlangs as plain traffic load and the rest of the forecasting columns stop looking like algebra. If you want a managed VICIdial server with these reports ready to run, take a look at VICIfast pricing.