Service Level: Definition, Formula, and the 80/20 Rule
Service level is the percentage of inbound contacts answered within a defined threshold, most commonly 80 percent within 20 seconds, and it is the accessibility target that every contact center staffing plan is built to meet.
Service level is the percentage of contacts answered within a chosen number of seconds. It is written as two numbers, such as 80/20, meaning 80 percent of calls answered within 20 seconds. It describes how accessible a contact center is to the people trying to reach it, and it is the objective that forecasting, staffing and intraday management all serve.
What Is Service Level?
Every queue makes a promise about waiting. Service level is that promise stated as a measurable target. When a center adopts 80/20, it is saying that in any given period, four out of five callers will hear an agent within 20 seconds. The remaining fifth will wait longer, and how much longer is described by other metrics.
Unlike handle time or occupancy, service level is a target chosen by management rather than a fact observed in the operation. It is a business decision that balances what customers expect against what the company will pay in staffing. A premium banking line might choose 90/10. An outbound sales team's inbound overflow might accept 70/60. Neither is wrong; they reflect different economics.
Service level is also the input the Erlang C model is solved for. Given call volume, handle time and a service level target, the model returns how many agents must be available. Change the target from 80/20 to 90/20 and the agent requirement rises, sometimes sharply, because the last few percentage points of accessibility demand disproportionate slack in the queue.
The 80/20 Convention
Nobody owns 80/20 and no standards body set it. Call Centre Helper's industry standards review describes 80 percent of calls answered in 20 seconds as the traditional industry standard, and most contact center platforms ship with it as the default. Its persistence has less to do with customer research than with the fact that it sits at a reasonable point on the cost curve: going from 80/20 to 90/20 typically requires several more agents on a mid-sized queue, while dropping to 70/30 saves less than it costs in abandonment.
Sensible practice is to treat 80/20 as the starting point and then test it. If abandonment stays low at 75/30 for a particular queue, that queue may be over-staffed. If customers begin abandoning at 15 seconds on a sales line, 80/20 is too loose for that line.
How to Calculate Service Level
The base formula is:
Service level (%) = Calls answered within threshold ÷ Calls offered × 100
The complication is abandoned calls. A caller who hangs up at 12 seconds was never answered, but they also did not wait beyond the threshold. Centers treat them in one of three ways, and the choice can move the reported figure by several points:
| Variant | Formula | Effect |
|---|---|---|
| Abandoned calls count against you | Answered within threshold ÷ (Answered + Abandoned) | Most conservative; the usual default |
| Abandoned calls are excluded | Answered within threshold ÷ Answered | Most flattering; hides queue problems |
| Only calls abandoned after the threshold count against you | Answered within threshold ÷ (Answered + Abandoned after threshold) | Middle ground; treats short abandons as not the center's fault |
Whichever variant is used, it must be applied consistently across periods and matched to the variant the ACD uses in its own reports, or the numbers will never reconcile.
Worked example: a 50-agent inbound team, one interval
In the 10:00 to 10:30 interval, the queue was offered 412 calls. Of those, 331 were answered within 20 seconds, 58 were answered after 20 seconds and 23 were abandoned (9 within 20 seconds, 14 after).
- Conservative variant: 331 ÷ 412 = 80.3 percent
- Excluding abandons: 331 ÷ 389 = 85.1 percent
- Counting only late abandons: 331 ÷ 403 = 82.1 percent
Same interval, three answers spanning five points. The conservative figure is the one to report upward, because it is the one that reflects what the caller experienced.
Service Level vs Average Speed of Answer
Service level and average speed of answer (ASA) both describe waiting, but they describe it differently and can move in opposite directions.
| Service level | Average speed of answer | |
|---|---|---|
| What it reports | Share of calls answered inside a threshold | Mean wait for calls that were answered |
| Sensitive to | The threshold chosen | Long-tail waits |
| Hides | How long the unlucky 20% waited | Whether most callers waited briefly |
| Good at | Setting a staffing target | Describing typical experience |
| Typical figure at 80/20 | 80% | 15 to 30 seconds |
A queue can hit 80/20 while ASA climbs, because the 20 percent who missed the threshold waited three minutes instead of one. Reporting both is standard practice, and adding the abandonment rate makes the picture complete.
Service Level Benchmarks
80/20 remains the benchmark most centers quote. Variation by sector is commonly described in these ranges:
- Emergency and critical support lines: 90/10 or tighter
- Financial services, premium customer tiers: 80/20 to 90/20
- General customer service and retail: 80/20 to 80/30
- Technical support with callback options: 70/30 to 80/60
- Chat: often stated as 80 percent within 30 to 60 seconds to first response
- Email: stated as a percentage within hours, such as 90 percent within 4 hours
A target is only meaningful with a measurement period attached. Hitting 80/20 for the day while missing it in every morning interval is not success; it means the afternoon was overstaffed to compensate. Most centers set the target at the interval level and report both interval compliance and the daily aggregate.
Why Service Level Matters
Service level is the point where staffing cost and customer experience meet. It is the first metric a client asks for in a BPO contract, the figure penalties are attached to and the number an operations manager watches on the wallboard all day. Missing it consistently has a compounding effect: abandonment rises, callbacks rise, the same customers call twice and the extra volume makes the next interval harder to hit.
For the workforce planner, service level is the constraint that all other decisions are solved against. The forecast, the shrinkage allowance, the shift patterns and the intraday moves all exist to deliver a chosen service level at the lowest sustainable occupancy.
How to Improve Service Level
- Fix the interval forecast before adding people. Most service level misses are timing errors, with the right number of agents for the day in the wrong half-hours. Compare forecast to actual by interval for four weeks and look for a pattern.
- Tighten schedule adherence. Agents logging in five minutes late across a 50-person team removes four agent-hours from the morning peak. The schedule adherence entry explains why this hits service level harder than it hits any other metric.
- Schedule shrinkage away from peaks. Training and meetings during the 10:00 to 12:00 window are the single most common self-inflicted service level failure.
- Manage real time. Voluntary time off in quiet intervals and overtime offers in busy ones are cheaper than permanent headcount.
- Reduce handle time waste. Every 15 seconds off AHT at fixed staffing raises service level in the intervals that were marginal.
- Offer callbacks in overflow. A virtual queue removes calls from the wait pool without abandoning them, which improves both service level and customer experience.
- Revisit the target. If the center is chronically five points short of 80/20 and abandonment is under 3 percent, the target may be tighter than the business needs.
Service level is a promise, ASA is the typical experience, abandonment is the consequence of breaking the promise. Report all three, by interval, and manage the schedule to the intervals rather than the day.
How to Track Service Level
Service level comes from the ACD or contact center platform, which timestamps every call from arrival to answer or abandonment. HiveDesk does not compute it.
HiveDesk's role is on the supply side of the service level equation. The shift schedule defines how many agents are supposed to be available in each interval. Clock-in and clock-out records from the desktop, mobile and browser apps show how many were. When an interval misses target, the timesheet view answers the first diagnostic question, which is whether the planned agents were present, before anyone looks at the forecast. Leave records explain planned gaps; attendance records expose unplanned ones. Everything is included at $5 per user per month. The call center capacity planner on this site shows how to turn a service level target into a staffing plan that HiveDesk can then schedule and track.
Related Terms
- Occupancy rate: the efficiency cost of a tighter service level
- Average handle time: the workload input to the staffing model
- Shrinkage: the gap between scheduled and available agents
- Schedule adherence: the intraday discipline that protects the target
- Workforce management: the function accountable for hitting it
The Erlang C calculator converts any service level target into agents required, and the BPO KPI guide explains how service level is written into outsourcing contracts.
Frequently Asked Questions
What does 80/20 service level mean?
It means 80 percent of inbound calls are answered within 20 seconds. The first number is the percentage, the second is the threshold in seconds. It is the most common target in the industry but not a regulatory or formal standard.
How is service level calculated in a call center?
Divide the number of calls answered within the threshold by the number of calls offered and multiply by 100. The treatment of abandoned calls changes the result, so confirm whether abandons are included in the denominator.
What is the difference between service level and ASA?
Service level reports the share of calls answered within a threshold. Average speed of answer reports the mean wait for answered calls. A center can meet service level while ASA rises if the callers who missed the threshold waited a long time.
Why is service level important?
It is the accessibility promise the center makes to customers and, in outsourcing, to clients. It drives the staffing requirement directly, and missing it raises abandonment and repeat contacts, which increase workload further.
Is a 90/10 service level realistic?
For small queues, rarely, because the occupancy required to hit it would be far too low to afford. For large pooled queues on critical lines it is achievable but expensive. Most centers reserve targets that tight for emergency or premium lines.
Should service level be measured daily or by interval?
By interval, with a daily summary. A daily figure can hide morning intervals that missed badly and afternoon intervals that were overstaffed to compensate. The interval view is what drives schedule changes.
Browse more workforce and contact center terms in the glossary.