Analytics
Occupancy
Also called: occupancy rate · agent occupancy
Occupancy is the proportion of an agent's logged-in time spent handling contacts rather than waiting for one. It is offered load divided by the number of agents on the phones.
Occupancy and utilisation are often confused. Occupancy measures only logged-in time, so breaks and training sit outside it — those belong to shrinkage. An agent can be 85% occupied while being available for phone work just 60% of their paid shift.
The commonly cited healthy band is 75% to 85%, with roughly 83% observed as a practical working maximum. Sustained occupancy above 85% is strongly associated with burnout and attrition, which is why occupancy is best treated as a ceiling constraint in staffing rather than a number to maximise.
High occupancy and high service level pull against each other. Answering calls quickly requires idle agents waiting for them, so a queue staffed to 95% service level will necessarily run lower occupancy. Any staffing model that ignores this trades agent wellbeing for a metric.
See it in the product
Frequently asked questions
How is occupancy calculated?
Offered load in Erlangs divided by the number of agents on the phones. Equivalently: time spent handling contacts divided by total logged-in time.
What is a good occupancy rate?
Between 75% and 85%. Below that you are over-staffed; sustained above 85% and you are borrowing against agent attrition to pay for today's service level.
Related terms
See Occupancy in a real call.
Book a 30-minute demo and watch Finn handle inbound and outbound calls end to end — no stack to assemble.