Why Erlang C Matters
If you've ever wondered why adding one more agent makes such a big difference to your SLA — or why removing one agent causes it to collapse — Erlang C is the answer.
It's the mathematical model that describes how queues behave when customers arrive randomly and agents serve them one at a time. It was developed in 1917 by a Danish engineer named Agner Krarup Erlang for telephone exchanges. A hundred years later, it's still the foundation of contact center staffing.
The Intuition
Imagine a coffee shop with one barista. When customers arrive faster than the barista can serve them, a queue forms. The queue grows until either customers leave or the rush ends. Erlang C tells you: given an arrival rate and a service time, how many baristas do you need to keep the queue short?
The key insight is non-linearity. Going from 90% to 95% SLA doesn't require 5% more agents. It might require 20% more. That's because the last few percentage points of SLA are the most expensive — you're staffing for the peak of the peak.
The relationship between agents and SLA is not linear. That's the most important thing to understand about staffing.
What This Means in Practice
It means your SLA target has a cost. And the cost increases exponentially as you push toward 100%. Understanding this helps you have honest conversations with stakeholders about what SLA levels are actually achievable at a given budget.