What is service level?
Service level is a measure of how quickly a contact centre answers its customers. It is expressed as a percentage of contacts answered within a target time, written as a pair: X% in Y seconds. A target of 80/20 means 80% of calls should be answered within 20 seconds.
A simple version of the formula is service level = calls answered within threshold / calls offered. In practice, contact centres vary in how they treat calls that are abandoned before the threshold, so it’s important to agree a definition and stick to it.
Service level is mostly used for real-time channels such as phone and webchat. Email and messaging usually have a response-time target measured in hours rather than seconds.
Why service level matters
Service level is the clearest expression of how accessible your organisation is. When it slips, customers wait longer, more of them hang up, and many of them try again later. That makes it a leading indicator of abandonment rate and of customer frustration.
It is also the target that drives staffing. Workforce management teams use forecast volume, average handle time and a model such as Erlang C to calculate how many agents are needed in each interval to achieve the chosen service level. Small changes in the target can mean large changes in staff, especially at the top end: going from 80% to 90% in 20 seconds usually needs noticeably more agents than going from 70% to 80%.
Choosing the right target
The 80/20 convention is a sensible starting point, but the right target depends on the purpose of the line:
- Urgent or safety lines, such as electricity fault lines or hazard calls, usually justify tighter targets, because a delayed answer can carry real risk.
- General enquiries are often fine at 80/20 or similar.
- Low-urgency channels such as email might target a response within one business day.
Many regulated organisations, and government agencies with published service standards, also report service level publicly, so the target should be realistic as well as ambitious.
Measuring service level well
A single daily number can hide a lot. Good practice includes:
- Report by interval. A day at 80% can include a morning at 95% and an afternoon at 40%. Half-hourly or hourly reporting shows when customers actually struggled.
- Track it with related metrics. Pair service level with average speed of answer, abandonment and longest wait, so a reasonable average doesn’t hide a few very long waits.
- Watch the trade-off with occupancy. Very high service levels usually mean lower agent occupancy, because agents need to be available when calls arrive. Pushing both to the maximum at once isn’t realistic.
- Plan for the unplanned. Forecasts handle normal peaks, but a storm or system outage can create a call avalanche far beyond any roster. Overflow call handling and surge capacity protect service level when that happens.