In a contact center, service level is the percentage of contacts answered within a set time threshold. It is usually written as two numbers, such as 80/20, meaning 80% of calls answered within 20 seconds. It is one of the main measures used to judge whether a queue is staffed well enough for the demand coming in.
At a glance
- Service level combines a percentage and a time threshold: X% of contacts answered in Y seconds.
- 80/20 is a common example for phone queues, not a standard; targets vary by business, channel and contact type.
- The formula matters: how abandoned and very short calls are counted changes the result.
- It is measured per queue and per interval (often every 15 or 30 minutes), then rolled up.
- It is an operating measure, not a contract; an SLA may reference it but is a separate thing.
What problem it solves
“How long do customers wait?” sounds simple, but an average hides the experience of the customers who waited longest. Service level gives a clearer target: most customers should be answered within a stated time. It turns a vague goal (“answer quickly”) into something that can be planned, staffed and reported.
It also links demand to staffing. Workforce management (WFM) tools use a service level target, forecast volume and expected handle time to calculate how many agents are needed in each interval. That makes service level the bridge between customer experience and labor cost, which is why operations, finance and outsourcers all watch it.
How it works
The basic formula. Contacts answered within the threshold, divided by contacts offered, for a given queue and time period. If 400 calls arrive in an hour and 320 are answered within 20 seconds, service level for that hour is 80%.
Abandoned contacts. This is where formulas differ. Common variations:
- Count abandoned calls as misses (strictest).
- Exclude abandoned calls from the calculation entirely (most generous).
- Exclude only short abandons, such as callers who hang up within a few seconds, and count the rest.
Each variation gives a different number from the same traffic. When comparing platforms, sites or outsourcers, confirm the formula first.
Intervals and roll-ups. Service level is usually tracked in short intervals because staffing and demand change through the day. A daily figure of 80% can hide a morning at 50% and an afternoon at 95%.
Digital channels. For chat and messaging, the same idea applies with longer thresholds, such as first response within a minute or two. For email and asynchronous messages, it is often expressed as a response within hours. Thresholds should reflect what customers expect on each channel.
Related measures. Average speed of answer (ASA) reports the average wait for answered contacts, and abandonment rate reports how many gave up. Service level, ASA and abandonment together give a fuller picture than any one alone. Pushing service level higher usually raises staffing costs and lowers agent occupancy, because more agents must be idle and ready when contacts arrive.
When it matters for buyers
- When outsourcing. A BPO contract usually includes service level targets; the formula, intervals and penalties must be written down precisely.
- When changing contact center platforms, because the new system’s default formula may not match the old one, making before-and-after comparisons misleading.
- When setting staffing budgets. Moving from one target to a stricter one can require noticeably more agents; model it before committing.
- When adding channels, which need their own thresholds and reports rather than a single blended number.
Most contact center as a service platforms report service level out of the box, with configurable thresholds.
Questions to ask vendors
- How does your platform calculate service level by default, and can we change how abandoned and short-abandoned contacts are treated?
- Can we set different thresholds by queue and channel?
- At what interval is service level reported, and is it available in real time for supervisors?
- How does your workforce management tool use service level targets to build schedules?
- For outsourcers: which formula will the contract use, at what interval is compliance measured, and what credits or penalties apply?
- How are contacts that move between queues (for example after a transfer or overflow) counted?
How it differs from an SLA
A service level agreement (SLA) is a contract term that sets what a provider commits to and what happens if it misses, such as uptime for a cloud service or repair times for a circuit. Contact center service level is an operational measure of answer speed. The two meet when a contract with an outsourcer or a managed contact center provider includes a service level target as one of its SLAs. Even then, the contract should define the formula, because “service level” alone does not say how abandoned contacts or measurement intervals are handled.