Tool · Erlang C
Call center staffing calculator
Erlang C for agents on contacts, shrinkage for agents scheduled, and the number most Erlang calculators skip: how many agents AI actually removes once you count resolution, not containment.
Last updated
Use resolved contacts, not contained ones: a contact that ends in the bot and comes back through another channel still needs an agent. See containment vs resolution, or check what resolution rate is realistic with the industry benchmark.
Erlang C assumes contacts queue until answered and one contact per agent. For concurrent chat, divide AHT by realistic concurrency first; for email, use a longer answer target.
How the calculator works
- Workload in Erlangs is contacts per hour × average handle time in seconds ÷ 3,600.
- Erlang C finds the smallest agent count that meets your service level and keeps occupancy under the cap.
- Agents scheduled divides that count by (1 − shrinkage). Dividing, not adding, is the step most spreadsheets get wrong.
- The AI run removes the resolved share of contacts, raises handle time on what remains, and repeats steps 1–3.
The formulas, two worked examples and sourced shrinkage benchmarks are in our call center staffing guide.
Want this modelled on your real traffic?
Book a working session — we will map realistic AI resolution onto your interval data and show the headcount it actually frees.