What is call center occupancy?

Call center occupancy is the percentage of a team member's logged-in time that is spent actively handling customer contacts, including talk time, hold time, and after-contact work. The remainder is time spent available and waiting for the next contact.

Occupancy answers a narrow question: when people were at their desks and ready to work, how much of that time did the volume actually keep them busy? It is a demand metric more than a performance metric. A low occupancy figure usually means the center was staffed for more volume than arrived, not that anybody was slacking.

This page covers what occupancy measures, how to calculate it, what a healthy range looks like, why very high occupancy is a problem rather than an achievement, how occupancy differs from utilization and adherence, and how to bring it into range.

Call center occupancy in one sentence

Occupancy is the share of available time that got used.

What occupancy actually measures

Occupancy divides logged-in time into two states: handling a contact, and waiting for one. Time spent handling counts toward occupancy. Time spent available and idle does not.

Occupancy only counts time the person was logged in and available to receive contacts. Breaks, lunch, training, coaching, and team meetings sit outside the calculation entirely. That is what separates occupancy from utilization, and it is the most common source of confusion between the two.

The thing occupancy is most often misread as measuring is effort. It is not. Occupancy is set almost entirely by the relationship between staffing and arriving volume, both of which are decided above the individual. A team member can work flat out for eight hours and post 62% occupancy because the forecast over-staffed their interval. Managing individuals on occupancy is one of the more reliable ways to demoralize a team while changing nothing.

How to calculate call center occupancy

Occupancy = total handle time ÷ (total handle time + total available time) × 100

Handle time is talk time plus hold time plus after-contact work. Available time is time logged in, ready, and waiting.

A worked example. Across a shift, a team member spends 300 minutes handling contacts and 120 minutes available and waiting. Their break and lunch total 60 minutes and do not enter the calculation.

Occupancy = 300 ÷ (300 + 120) × 100 = 71.4%

What decides the number

Whether after-contact work counts. Most definitions include it. Some centers that use a separate wrap-up state exclude it, which drops occupancy by several points and makes the figure incomparable with anyone else's. Pick one and stay with it.

The interval length. Occupancy measured across a full day looks calm. The same day measured in half-hour intervals often shows a stretch at 94% during the morning peak and a long tail at 40% after 4pm. The daily average hides both problems.

Channel mix. Occupancy was designed for voice, where a person handles one contact at a time. On chat, where someone may hold three concurrent conversations, the arithmetic breaks down unless the platform accounts for concurrency. Applying a voice occupancy target to a blended team produces a number that means nothing.

What is a good call center occupancy rate?

The range most contact centers work toward is 75% to 85%. The most concrete figure available comes from Call Centre Helper, which reports an average of 83.3% across more than 160,000 entries into its online Erlang calculator. Read that precisely: it is the average maximum-occupancy cap that planners set, not measured occupancy across the industry.

The two ends of that range fail in opposite directions:

Below 70% generally means over-staffing. People are logged in and waiting, the schedule is carrying more heads than the volume needs, and the cost per contact rises accordingly. Occasionally it means something more interesting, such as a routing rule sending contacts to a smaller pool than intended.

Above 90% is the one worth taking seriously. ICMI has long put the burnout threshold between 88% and 92% sustained across several consecutive half-hours, and Brad Cleveland, who edited that guidance, now states it more simply as around 90%. Both are expert judgment rather than a controlled study, and worth treating as such. At that level there is almost no gap between one contact ending and the next arriving. Sustained, it produces exactly what you would expect: rising handle times as people get tired, falling quality, more errors, and attrition. The productivity gain is illusory because it is being paid for out of the team's capacity to keep doing the job. Above 95%, people have effectively no recovery time between conversations at all.

The useful way to read occupancy is as a range check rather than a target to maximize. It is one of the few contact center metrics where the best result is a middling number.

Occupancy vs. utilization vs. adherence vs. shrinkage

Four workforce management metrics that get conflated constantly. They measure different denominators.

The distinction that matters most: occupancy excludes breaks, utilization includes them. Utilization is always the lower number. If somebody quotes an occupancy figure that looks suspiciously low, check whether they have actually calculated utilization.

Why occupancy matters

It validates the forecast. Occupancy is the clearest read on whether the staffing plan matched reality. Persistent low occupancy means the forecast is running hot. Persistent high occupancy means it is running cold and the team is absorbing the difference.

It connects staffing to cost. Every point of occupancy below target is paid time spent waiting. On a large team that arithmetic gets expensive quickly, which is why occupancy tends to be the metric finance asks about.

It is an early warning on burnout. Occupancy running above 90% for weeks is visible in the data long before it shows up in attrition or quality scores. Of the standard contact center metrics, it is one of the few that gives advance notice of a people problem.

Where occupancy misleads

It rewards slow work. A team member who takes longer on every contact posts higher occupancy than a faster colleague handling the same volume. Read in isolation, occupancy makes inefficiency look like productivity. It only makes sense next to handle time and resolution rates.

Aggregate figures hide the peaks. A comfortable 78% daily average can contain a two-hour stretch at 95%. The interval view is the one that tells you whether the team had a bad day.

It does not work unmodified for concurrent channels. See the channel mix note above. Chat and messaging need a concurrency-adjusted calculation.

It is not an individual performance metric. Worth repeating because it is the most common misuse. Occupancy is set by staffing and volume. Holding individuals accountable for it asks them to control something they do not control.

How to bring occupancy into range

If occupancy is too high, the honest answers are more people, better forecasting at the interval level, or less volume. Shifting break times can smooth a peak but cannot fix a structural shortfall. Reducing repeat contacts is the option that helps most and takes longest, because a contact that never had to happen is worth more than a contact handled faster.

If occupancy is too low, the schedule is carrying slack. Options in rough order of how easy they are: adjust shift patterns to match the arrival curve, blend in deferrable work such as email or outbound follow-up during quiet intervals, and revisit the forecast assumptions that produced the over-staffing.

Either way, look at shrinkage first. Occupancy problems are frequently shrinkage problems in disguise. If uncoded training and meetings are quietly removing people from the floor, the schedule looks adequate and the intervals are not covered.

Occupancy when AI handles part of the volume

AI resolution changes occupancy in a way that is easy to misread, and the misreading is expensive.

When AI handles the simple, short, repetitive contacts, the volume that reaches a person shifts toward the complex and the long. Handle time per contact rises, contact count per person falls, and occupancy tends to rise because there is less dead air between longer conversations. None of that means the team got busier in a way that is bad, but it does mean the pre-AI occupancy target no longer describes a healthy state. Targets set against a Tier 1 mix need resetting once the mix changes.

The second effect is on the idle time itself. Some of the gap between contacts in a traditional center is genuine recovery, and recovery matters more when every conversation is a hard one. A team handling only complex work at 88% occupancy is under more strain than a team handling a mixed queue at the same number.

The practical guidance: recalculate the target when the mix changes, keep occupancy for human-handled volume only, and read it next to handle time rather than alone.