Tool

Free Erlang Call CenterStaffing Calculator

Calculate exactly how many agents you need to hit your target service level. Powered by the industry-standard Erlang C formula — used by call centers, BPOs and help desks worldwide.

Whether you run an inbound or outbound call center, this free calculator forecasts agents, occupancy, service level, and the percentage of calls that wait — instantly. With hourly staffing breakdowns and CSV export, straight into your WFM workflow.

Calculate Staffing Now
100% Free No Sign-up Required CSV Export Live Recompute Erlang C Math

Inputs

Try:
30 min

Time window for this calculation. 15 or 30 min is typical.

100 → 200/hr

Total calls expected to arrive in the reporting period.

4 min 0 sec

Talk time + hold time + after-call work, in seconds.

20 sec

The 't' in 80/20 SL — how quickly you want to answer.

80%

Industry standard is 80%. Financial/emergency services target 90%+.

85%

Recommended ceiling: 85%. Higher rates burn agents out.

30%

Breaks, training, sick days, meetings — typically 25–35%.

Results

Real-time staffing estimate

Required agents: 25. Service level: 81.0%. Occupancy: 78.4%.
Agents Required
25

Base: 17 · with shrinkage: 25

Occupancy
Healthy
78.4%

Healthy range

Service Level
On target
81.0%

Target: 80% in 20s

Calls That Wait
25.7%

Erlangs (traffic): 13.33

Service level vs target81.0% / 80%

With 17 agents you'll answer 81% of calls within 20s, at 78% occupancy. Add shrinkage and you need 25 on the schedule.

The math

The Erlang C formula, explained

The calculator above runs the standard Erlang C queueing model. Here's every step, in plain language.

1

Traffic intensity (Erlangs)

Convert call volume + AHT into a single 'workload' number called Erlangs.

u = (calls × AHT) ÷ (period_minutes × 60)
2

Pick a candidate agent count

Start at the smallest integer N greater than u. N must exceed u or the queue grows forever.

N = ⌈u⌉ + 1
3

Probability of waiting (Erlang C)

Compute P_w — the chance an arriving call has to wait at all.

P_w = (uᴺ/N!)·(N/(N−u)) ÷ [Σ uᵏ/k! + same numerator]
4

Service level at N agents

The fraction of calls answered within target time t (with AHT A).

SL = 1 − P_w · e^(−(N−u)·t/A)
5

Increment N until SL hits target

Keep adding agents one at a time. The smallest N where SL ≥ your target is your base staffing requirement.

N* = min{ N | SL(N) ≥ target }
6

Apply shrinkage

Gross up the base count to account for breaks, training, and absence. This is the number you actually schedule.

schedule = ⌈ N* ÷ (1 − shrinkage) ⌉

Implementation note

Our calculator runs the math in log space using a Lanczos lgamma approximation, so it stays numerically stable for very large agent counts (BPO-scale workloads with 1,000+ agents). No factorial overflow.

Hourly staffing requirements for your inbound & outbound call center

Plan your full shift hour-by-hour. Drop your forecasted call volume into each interval and instantly see agents required and the service level you'll hit for every time slot. All settings inherit from the calculator above.

Interval: 30 min18 slots · adjust on the calculator above
Hourly staffing requirements by time slot.
Time slotIncoming callsAgents requiredService levelOccupancy
9:00 AM – 9:30 AM00.0%0.0%
9:30 AM – 10:00 AM00.0%0.0%
10:00 AM – 10:30 AM00.0%0.0%
10:30 AM – 11:00 AM00.0%0.0%
11:00 AM – 11:30 AM00.0%0.0%
11:30 AM – 12:00 PM00.0%0.0%
12:00 PM – 12:30 PM00.0%0.0%
12:30 PM – 1:00 PM00.0%0.0%
1:00 PM – 1:30 PM00.0%0.0%
1:30 PM – 2:00 PM00.0%0.0%
2:00 PM – 2:30 PM00.0%0.0%
2:30 PM – 3:00 PM00.0%0.0%
3:00 PM – 3:30 PM00.0%0.0%
3:30 PM – 4:00 PM00.0%0.0%
4:00 PM – 4:30 PM00.0%0.0%
4:30 PM – 5:00 PM00.0%0.0%
5:00 PM – 5:30 PM00.0%0.0%
5:30 PM – 6:00 PM00.0%0.0%
Totals / Averages00 avg · 0 peak0.0%—

Step by step

How To Use an Erlang Call Center Staffing Calculator?

The calculator above is most accurate when you feed it grounded inputs. Here's how WFM teams configure each field — and the rules of thumb behind each one.

  1. Step 1

    Set your reporting period

    Pick the time window your forecast applies to. 15- or 30-minute intervals are the standard for accurate intraday planning — long enough to be statistically stable, short enough to capture peaks.

  2. Step 2

    Enter your forecasted call volume

    Pull historical data from the last 4 weeks for the same weekday and time. Average the four data points to smooth out anomalies. For trending lines, add a small growth factor.

  3. Step 3

    Input your Average Handle Time (AHT)

    AHT = talk time + hold time + after-call work. For inbound support this is straightforward. For outbound, include dial wait and call disposition time in your AHT.

  4. Step 4

    Set your target answer time

    This is the 't' in your service-level target — the speed by which you want to answer. The classic '80/20' SL means 80% of calls answered within 20 seconds.

  5. Step 5

    Choose your target service level

    80% is the industry default for general support. Premium and emergency services target 90%+. Self-service overflow lines might run 70%. Higher SL = more agents = higher cost.

  6. Step 6

    Configure max occupancy & shrinkage

    Cap occupancy at 85% to prevent burnout. Set shrinkage between 25–35% for most contact centers. Ignoring shrinkage typically understaffs you by 30%+.

  7. Step 7

    Review your results & build your schedule

    Use the hourly breakdown below to produce a staffing curve for the full shift. Export to CSV and drop into your WFM tool to layer in shift patterns.

Under the hood

How Does the Call Center Staffing Calculator Work?

The Erlang C model was developed by Danish mathematician A. K. Erlang in 1917 to size telephone switchboards at the Copenhagen Telephone Company. A century later it's still the gold standard for inbound contact center staffing — because the underlying queueing theory translates one-to-one onto modern call centers.

The model rests on three assumptions: calls arrive randomly following a Poisson distribution, handle times are exponentially distributed around your AHT, and callers wait in an infinite queue without abandoning. These are simplifications — real callers abandon, real arrivals cluster around bursts — but they hold well enough that Erlang C lands within ±1 agent of more sophisticated simulation models for typical workloads.

Internally the calculator works by iteration. It computes traffic intensity (Erlangs) from your call volume and AHT, then starts at the smallest viable agent count (one above traffic intensity) and asks: at N agents, what fraction of calls would be answered within target? If the answer beats your target SL, that's your base agent count. If not, N goes up by one and the calculation repeats. The math runs in log space so it stays accurate even for BPOs with thousands of agents.

Shrinkage is then applied as a final multiplier: ceil(base / (1 − shrinkage)). If your base requirement is 16 agents and shrinkage is 30%, you actually need to schedule 23 — because the other 7 will be on breaks, in training, or out sick on any given day. Ignoring shrinkage is the single most common WFM mistake; it typically understaffs a contact center by 30–40%.

A note on the limits: pure Erlang C doesn't model abandonment (the Erlang A variant does), multi-skill routing, or omnichannel blending (voice + chat + email). For most planning purposes Erlang C is more than enough, but if your team handles 4+ skills with strict routing rules, layer in a separate simulation. For omnichannel, treat chat and email as separate Erlang C problems with their own AHT and SL targets.

Use cases

Inbound vs Outbound Call Center Staffing Calculations

Erlang C handles both — with one tweak.

Inbound

Driver
Demand-driven (calls arrive randomly)
Goal
Hit a target service level
AHT components
Talk + hold + after-call work
Model fit
Erlang C is the gold standard
Typical settings
80/20 SL, 85% max occupancy
Example
Support, help desk, customer service

Outbound

Driver
Supply-driven (agents place calls)
Goal
Hit a contact/conversion rate
AHT components
Dial wait + talk + ACW
Model fit
Erlang C with adjusted AHT + dial ratio
Typical settings
Treat 'calls' as completed dials
Example
Telesales, surveys, collections

For outbound, simply set your AHT to include dial wait and call disposition, then treat "forecasted calls" as your target completed dials per period. The calculator's output gives you the agent count needed to hit that contact rate at your target SL.

FAQ

Frequently Asked Questions

A call center staffing model translates forecasted call volume into the number of agents you need on the floor at any given time. It blends three inputs — call arrivals, average handle time, and the service level you're targeting — and accounts for shrinkage (breaks, training, sick days, meetings). Most modern staffing models use the Erlang C queueing formula at their core, sometimes with adjustments for abandonment or multi-skill routing.
An Erlang call center staffing calculator applies the Erlang C formula — developed by Danish mathematician A. K. Erlang in 1917 for the Copenhagen Telephone Company — to estimate the smallest number of agents that can meet a target service level. You feed in forecasted calls, average handle time, target answer time, and target SL, and the calculator iterates upward until the smallest integer N of agents satisfies your target. It's the industry standard for inbound contact center staffing.
Workforce management (WFM) teams use Erlang calculators in two main ways. First, for capacity planning: how many agents do we need to hire to handle next quarter's forecast? Second, for intraday scheduling: how many agents need to be on the phones at 10:00 a.m. on Wednesday? They run the calculator once per interval (typically 15 or 30 minutes) across the operating day to produce a staffing curve, then layer shift patterns and shrinkage on top.
Yes — with one adjustment. Pure Erlang C models inbound queueing where calls arrive randomly. For outbound, the volume is supply-driven (agents place calls), so you treat the calculator differently: use AHT to include dial time + talk + wrap-up, and treat your 'calls' input as the number of completed connections per period. Many WFM teams combine the calculator with a dial-ratio (calls per agent-hour) to size their outbound team for a target contact rate.
This calculator implements the canonical Erlang C formula in log space to stay numerically stable at any scale. Results cross-check within ±1 agent of published industry calculators (callcentercalculator.com, callcenterhelper.com). Like any queueing model, Erlang C assumes Poisson arrivals, exponential handle times, and no abandonment — so real-world results may differ slightly if your traffic is bursty or you see high abandonment. For most planning, ±1 agent of accuracy is well within tolerance.
Three things drive every staffing decision: call volume (your forecast), average handle time (talk + hold + after-call work), and the service level you want to deliver. On top of that, shrinkage — the time your scheduled agents are unavailable for calls — is the silent multiplier. Typical shrinkage runs 25–35%. Ignoring it can understaff a contact center by a third. Other adjustments include multi-skill routing, omnichannel blending (chat + email + voice), and agent occupancy ceilings to protect against burnout.
An inbound calculator is demand-driven: calls arrive randomly and you size to meet a service level (e.g., 80% answered in 20 seconds). An outbound calculator is supply-driven: agents place calls and you size to hit a contact or conversion target. Inbound uses Erlang C directly. Outbound typically uses Erlang C with an adjusted AHT (including dial wait and after-call work) plus a separate dial-ratio model. This page handles both — for outbound, set your AHT to include dial + talk + wrap and treat 'calls' as completed dials.
The industry-standard target is 80/20 — answer 80% of calls within 20 seconds. Many B2C support lines run 80/20 or 80/30. Premium and emergency services (financial advisors, medical triage, roadside assistance) typically target 90/15 or even 95/10. Low-stakes self-service overflow lines might run 70/30. The right target balances customer experience against headcount cost — every 5-point bump in SL roughly translates to one or two extra agents for every 15–20 base agents.
Shrinkage = (total scheduled time − productive on-phone time) ÷ total scheduled time. Add up every non-productive minute over a representative period: paid breaks, lunch, training, team meetings, 1:1s, system downtime, paid time off, sickness, coaching. Divide by total scheduled hours and you have your shrinkage rate. Typical contact centers run 25–35%. Once you have a baseline, this calculator's shrinkage slider grosses up your base agent count: agents on the schedule = ceil(base / (1 − shrinkage)).

Keep exploring

More free Calilio tools

Vanity Number Generator

Turn your number into a memorable word or phrase.

Open

VoIP Speed Test

Test your connection's readiness for VoIP calls.

Open

Phone Number Checker

Check the validity and carrier of any phone number.

Open

Random Phone Number Generator

Generate randomized phone numbers for testing.

Open

Transform Your Business Telephony with Calilio

Step into the future of seamless connectivity with Calilio. Choose our advanced business phone system and revolutionize your business telephony today.

4.95200+ Reviews16+ Badges