Question 1

What is the Erlang C formula used for in call centers?

Accepted Answer

The Erlang C formula calculates the probability that a call must wait in queue and determines the minimum number of agents needed to meet a service level target (e.g. 80% of calls answered within 20 seconds). It uses traffic intensity in Erlangs, number of agents, average handle time, and target answer time.

Question 2

How do you calculate Erlang C for a chat center with concurrency?

Accepted Answer

For chat centers, divide traffic intensity by the concurrency factor: A = Chats x AHT / (Interval x Concurrency). If agents handle 3 chats simultaneously, effective traffic drops by 3x. Service level and wait time calculations should also use effective AHT = AHT / concurrency.

Question 3

What is the difference between Erlang C and Erlang A?

Accepted Answer

Erlang C assumes no caller abandonment - all contacts wait indefinitely, making it conservative. Erlang A (M/M/N+M model) accounts for callers who leave the queue, using an average patience parameter. Erlang A typically recommends 5-15% fewer agents for the same service level.

Question 4

How do you calculate AI deflection ROI for a contact center?

Accepted Answer

The correct method uses Erlang C before and after deflection to find the true FTE reduction. Run Erlang C on your current volume to get current scheduled headcount, then run it again on post-deflection volume. The difference in scheduled agents is your FTE saving. Multiply by fully loaded hourly rate and annual hours, subtract annual bot platform cost, to get net annual ROI. Simple hour-multiplication methods overstate savings by 20-40% because they ignore queuing dynamics.

Question 5

What is a realistic AI deflection rate for a contact center?

Accepted Answer

Realistic deflection rates vary by query type. Basic IVR self-service (account balance, store hours) typically achieves 20-35% deflection. AI chatbots on high-volume simple queries (password reset, order status) achieve 25-45% containment. Complex queries, complaints, and billing disputes rarely deflect fully. A realistic blended deflection target for a mixed-channel contact center is 20-35%.

Question 6

What is shrinkage in call center workforce management?

Accepted Answer

Shrinkage is the percentage of paid time agents are unavailable to handle contacts - including breaks, training, meetings, absences, and system downtime. Typical range is 25-35%. Formula: Scheduled Agents = Raw Erlang Agents / (1 - Shrinkage%). A 27% shrinkage means you need 1.37x the Erlang-calculated agents.

Question 7

What is a good occupancy rate for a call center?

Accepted Answer

The recommended occupancy range is 80-87%. Below 70% indicates overstaffing. Above 90% leads to agent burnout, quality degradation, and high attrition. Occupancy = Traffic Intensity (Erlangs) / Number of Agents.

Question 8

How do you calculate traffic intensity in Erlangs?

Accepted Answer

Traffic Intensity (Erlangs) = Call Volume x Average Handle Time / Interval Duration. For example: 200 calls in 30 minutes with 180s AHT = 200 x 180 / 1800 = 20 Erlangs. For chat centers, divide by concurrency.

Question 9

What is the 80/20 rule in call centers?

Accepted Answer

The 80/20 rule is a common service level target: 80% of calls answered within 20 seconds. This is used as an input to the Erlang C formula to calculate the minimum number of agents needed to achieve this target.