Call Centre Erlang Calculator: How It Works & When to Trust It
A call centre Erlang calculator is one of the most important tools in workforce planning.
It estimates how many agents you need to meet a target service level based on forecast demand, average handling time and staffing assumptions — turning the question "how many people do we actually need on the phones?" into a structured calculation instead of a gut feel.
But an Erlang calculator is a model, not an oracle. It rests on a set of assumptions that don't always hold in the real world, and it's only ever as good as the forecast, AHT and service-level target you feed it.
Used well it sharpens staffing decisions and budgets; used blindly it manufactures false confidence.
If you want a fast answer now, use the free Online Erlang C Calculator. If you want to understand the logic behind the numbers — what Erlang is, how it works, and where its limits begin — this guide walks you through it.
What it does
Converts a demand forecast, AHT and a service-level target into the number of agents required — and the rostered headcount once shrinkage is applied.
Why it matters
It connects random customer demand to workforce reality, replacing opinion-based staffing with data-led planning you can defend in a budget conversation.
What this guide covers
What Erlang is, the data you need, Erlang A vs B vs C, a worked sensitivity example, the model's limitations, common mistakes, and when it isn't enough on its own.
🧮 Need an answer now?
Use the free Online Erlang C Calculator to estimate the number of agents required, expected service level, occupancy, probability of waiting, and rostered headcount after shrinkage.
What is a Call Centre Erlang Calculator?
A call centre Erlang calculator is a forecasting tool that estimates how many agents are needed to answer customer contacts within a target timeframe.
It's most commonly built on the Erlang C formula — a queueing-theory model widely used to predict waiting behaviour and staffing requirements in contact centres.
It takes a handful of inputs — contact volume, average handling time (AHT), interval length and target service level — and converts them into a staffing requirement.
Instead of guessing whether 12 agents is enough, or debating whether "the phones feel busy", you get a structured estimate of the headcount required to hit your operational targets.
In plain English
An Erlang calculator answers one of the biggest operational questions in any call centre: how many people do we actually need on the phones to deliver the experience we want?
It moves staffing from opinion to evidence — but the evidence is only as good as the assumptions you put in.
✓ What it IS
- A staffing estimator based on queueing theory
- A scenario tool — model volume, AHT and service-level changes
- A way to justify staffing budgets with structure
- A planning input that pairs with judgement and good data
✗ What it is NOT
- A guarantee — it's a model with assumptions
- A fix for poor forecasts or inflated AHT
- A full answer for skills-based routing or blended work
- A reason to chase a universal service-level "benchmark"
Why Call Centres Use Erlang Calculators
Customer demand doesn't arrive in neat, evenly spaced intervals. Calls come in randomly — some periods are quiet, others spike, and handling times vary from one interaction to the next.
That randomness is exactly what makes workforce planning hard, and exactly what queueing theory was built to handle.
A good Erlang calculator lets you model the relationship between forecast demand, AHT, service-level targets, occupancy, speed of answer, and the number of agents required.
For Contact Centre Leaders
It turns a demand forecast into a defensible staffing plan — and lets you test the operational impact of lower staffing before you live through it.
For WFM & Planners
It's the core engine for modelling volume growth, longer AHT, and shrinkage effects on rostered headcount across every interval.
For Finance & Operations
It connects labour cost to service outcomes, so staffing budgets and business cases for systems, process or training rest on numbers, not anecdotes.
The point
An Erlang calculator isn't only a day-to-day planning tool — it's a scenario-modelling tool.
It lets you explain, in advance, the operational consequences of changes in demand, productivity, process design and customer behaviour. That foresight is where most of its value lives.
What Data You Need to Use an Erlang Calculator
Most Erlang C calculators rely on a small set of core inputs, usually available from your phone system, WFM platform or reporting suite. The six below cover almost every model.
Contact volume
The number of inbound contacts expected in a defined period — typically a 15, 30 or 60-minute interval.
Average Handling Time
The average time to handle one interaction — talk time, hold time and after-call work combined.
Service level target
The percentage of contacts you want answered within a target time, e.g. 80% in 30 seconds.
Interval length
The period you're modelling. Most call centres plan in 15, 30 or 60-minute intervals.
Shrinkage
The percentage of paid time lost to breaks, coaching, training, meetings, leave and other non-productive activity.
Optional assumptions
Depending on the tool: occupancy caps, caller patience, abandonment, or multi-scenario comparisons.
Garbage in, garbage out: if your data quality is poor, your staffing outputs will be too.
Erlang is powerful, but it cannot rescue a bad forecast, an inflated AHT assumption, or a service target nobody funded the labour to hit.
Erlang A vs Erlang B vs Erlang C
There are three common Erlang models, each used for a different purpose. The differences come down to one question: what does the model assume a blocked or queued contact will do?
Erlang A
Includes customer abandonment. Useful when caller patience materially affects outcomes — it models the reality that some callers hang up rather than wait.
See abandonment rate for why this matters.
Erlang B
Used to size the number of lines, trunks or channels required, where a blocked contact is simply lost (gets a busy signal) rather than queued.
Erlang C
The most widely used model for call centre staffing. It assumes queued calls wait rather than abandon — which is why it's the default in most workforce planning tools, and also its biggest simplification.
Which one do people mean?
When people search for an "Erlang calculator", they almost always mean an Erlang C calculator. Just keep its core assumption in mind: it assumes nobody abandons the queue.
In operations with real abandonment, Erlang C tends to slightly over-estimate the agents you need — which is usually a safe direction to be wrong in, but worth knowing.
How Erlang C Works in a Call Centre
Erlang C models the probability that a contact will have to wait for an available agent, based on traffic load and staffing levels.
From that, it estimates expected service level, average speed of answer, and the staffing required. You don't need to calculate the formula by hand — but you do need to understand the commercial logic behind it.
Forecast the demand
Your forecast estimates how many contacts will arrive in a defined interval.
Convert demand into workload
Your AHT estimate turns that volume into workload — the total agent time the demand represents.
Test staffing against the target
The Erlang C model tests how many agents are needed to absorb that workload while still hitting your target service level.
Apply shrinkage for rostered headcount
Shrinkage is applied to convert "agents required on the phones" into the rostered headcount you actually need to schedule.
The flow, at a glance
Worked Example: Why AHT Matters So Much
One of the biggest benefits of an Erlang calculator is showing leaders how sensitive service levels are to small changes in key inputs.
Hold call volume and agent numbers steady but let average handling time creep up, and your service level can fall sharply.
That means a new product issue, more complex queries, longer wrap time, inexperienced agents or poor systems can all open a significant staffing gap — even when demand itself hasn't changed at all.
Example: how a rise in average handling time can significantly reduce expected service level when staffing stays the same.
The takeaway
Erlang isn't just a staffing calculator — it's a way to explain the operational consequences of changes in demand, productivity, process design and customer behaviour before they hit your service level.
A 30-second AHT change is rarely "just 30 seconds".
A Brief History of Erlang
The name Erlang comes from Agner Krarup Erlang, a Danish mathematician whose work in the early 1900s laid the foundation for modern queueing theory.
At the time, telephone networks needed a way to estimate how many operators and circuits were required to handle demand efficiently — and his work solved it.
More than a century later, the same principles still underpin many workforce management and forecasting tools used in contact centres today.
That staying power tells you something: even in modern omnichannel environments, the core challenge of matching random demand to finite capacity hasn't gone away.
The Limitations of Erlang C
As useful as Erlang C is, it isn't magic. It makes assumptions that don't always hold in real-world operations — and knowing them is the difference between using the model well and trusting it blindly.
What Erlang C assumes
- Callers will wait rather than abandon (often untrue — that's what Erlang A addresses)
- Contacts arrive in a broadly random, stable pattern
- All agents are equally skilled and can handle every work type
- Staffing is continuously available within the interval
- It's less reliable in very low-volume environments
- On its own, it isn't well suited to complex skills-based routing
None of this makes Erlang C "wrong". It means the model should be used intelligently — with an understanding of your channel mix, routing design, operating model and data quality.
Treat its output as a well-reasoned estimate to interrogate, not a number to obey.
Common Mistakes When Using an Erlang Calculator
Many leaders use the formula correctly but still reach poor decisions, because the assumptions feeding the model are weak. These are the mistakes that quietly undermine the output.
❌ Ignoring shrinkage
"Agents on the phones" is not the same as rostered headcount. Skip shrinkage and you'll consistently roster too few people.
❌ Using a blended AHT badly
If handling times vary heavily by contact type, a single average distorts the result. Model the segments separately where it matters.
❌ Chasing an arbitrary service level
Targeting premium responsiveness without funding the labour creates chronic, designed-in failure. The "right" target is a commercial decision, not a benchmark to copy.
❌ Using hourly averages for volatile demand
Big spikes inside the hour can collapse service even when the hourly average looks comfortable. Plan at the interval level.
❌ Treating Erlang as perfect
It's a model, not a guarantee. The output is a starting point for judgement, not the end of the conversation.
❌ Ignoring occupancy
Service level can look fine while agent occupancy becomes operationally unsustainable — a leading indicator of burnout and attrition.
The benchmark trap: there is no universal "correct" service level — 80% in 20 seconds is a common convention, not a law.
The right target depends on customer expectations, demand patterns, commercial constraints and channel strategy. Feeding a copied benchmark into Erlang just gives you a precise answer to the wrong question.
When Erlang Isn't Enough on Its Own
An Erlang calculator is helpful in most voice operations, but it isn't always sufficient by itself. Be cautious — and reach for more advanced modelling — in environments like these:
- Very low-volume operations, where the maths gets noisy
- Highly variable or "lumpy" arrival patterns
- Multi-skill or specialist routing environments
- Blended teams handling calls, email, chat and back-office work together
- Operations where schedule adherence is poor or intraday volatility is high
Go further than a basic calculator
For these cases you need workload modelling and scenario planning beyond a single Erlang C result.
The ACXPA WFM Hub brings together staffing simulators, interval planning and shrinkage tools designed for exactly these real-world conditions.
Frequently Asked Questions About Erlang Calculators
What is an Erlang calculator used for?
It estimates the number of agents required to meet a target service level, based on forecast demand and handling-time assumptions.
It's also widely used for scenario modelling — testing how changes in volume, AHT or staffing affect service outcomes.
What is Erlang C in a call centre?
Erlang C is the most common queueing formula used to calculate staffing in call centres.
It estimates the probability of waiting, the resulting service level, and the agent numbers required — under the assumption that queued callers wait rather than abandon.
Is Erlang C accurate?
It can be very useful when its assumptions broadly match your operating environment, but it remains a model.
Accuracy depends on good forecast data, realistic AHT, a sensible service target, and an understanding of its limitations.
Because it assumes no abandonment, it tends to slightly over-estimate staffing in operations with real abandonment.
What service level should a call centre use?
There's no universal answer, and you should be wary of anyone who gives you one. Many centres use 80% of calls answered in 20 or 30 seconds, but that's a convention, not best practice.
The right target depends on customer expectations, demand patterns, commercial constraints and channel strategy. Set it deliberately — then model the staffing and occupancy it implies before you commit.
What is shrinkage in workforce planning?
Shrinkage is the percentage of paid time agents are unavailable to handle contacts — breaks, coaching, meetings, leave, training, system issues and other non-productive activity.
Erlang gives you agents required on the phones; shrinkage converts that into the rostered headcount you actually need to schedule.
Can I use Erlang for live chat or other channels?
Sometimes, with caution. The closer the work behaves like random, queued demand handled by interchangeable resources, the more useful Erlang is.
Chat (where agents handle concurrent sessions), email and blended work break those assumptions, so they usually need additional modelling rather than a straight Erlang C calculation.
Should I aim for 80% of calls answered in 20 seconds?
Only if your customers, your demand and your commercials say so.
"80/20" became popular as a default decades ago and stuck — but it isn't a standard, and copying it can mean over-spending on responsiveness customers don't value, or under-funding a service promise you've actually made.
Decide the target from your context, then use Erlang to cost it.
What's the difference between an Erlang calculator and a WFM system?
An Erlang calculator answers a single staffing question for a single set of inputs.
A workforce management (WFM) system uses Erlang-style maths as one engine inside a much wider toolkit — forecasting, interval-level planning, scheduling, shrinkage, intraday management and scenario simulation.
Erlang is the formula; WFM is the discipline around it.
Where to Next
Summary: Call Centre Erlang Calculator
A call centre Erlang calculator remains one of the most practical tools for estimating staffing, modelling service levels and making better workforce planning decisions.
It connects demand, handling time and service expectations to the number of agents required — and lets you test scenarios before they become operational reality.
But it's a model, not a guarantee. Used well — with good forecasts, realistic AHT, a deliberately chosen service target and an eye on its assumptions — it strengthens forecasting, staffing conversations and budget cases.
Used blindly, it manufactures false confidence and a precise answer to the wrong question. The value comes from understanding both the formula and the environment you're applying it to.
If you want a fast staffing answer, use the free Online Erlang C Calculator. If you want to build real workforce planning capability, explore the WFM Hub and the CX Skills Workforce Optimisation training courses.
