Staffing and roster calculator
Turn the agents you need on the queue into a roster: weekly productive hours, paid hours, full-time equivalents and the most people you must schedule at once.
Roster size
78
78.0 full-time equivalents
Most people scheduled at once
29
Busy hours, after shrinkage
Productive hours per week
2,184
Paid hours per week
3,120
From requirement to roster
An Erlang C result tells you how many agents each interval needs. A roster is a different question: how many people do you employ so those intervals are covered every week? The answer depends on how many hours you are open, how many days a week, how much of paid time is lost to shrinkage and how many paid hours each person works.
This calculator uses a simple shape of the day: one level of agents for busy hours and another for quiet hours, placed across your opening hours. It works in 30-minute intervals, adds up the productive agent-hours the queue needs each week, grosses them up for shrinkage, and divides by the paid hours per person.
The formula
Productive hours per week = sum(agents needed in each interval x interval hours) x days open per week
Paid hours per week = productive hours / (1 - shrinkage)
FTE = paid hours per week / paid hours per person per week
Roster size = FTE, rounded up
People scheduled in an interval = agents needed / (1 - shrinkage), rounded upWorked example
A queue is open 24 hours a day, seven days a week. It needs 20 agents for 12 busy hours and 6 agents for the other 12 hours. That is 20 x 12 + 6 x 12 = 312 productive agent-hours a day, or 2,184 a week. With 30% shrinkage, paid hours needed are 2,184 / 0.70 = 3,120 a week. At 40 paid hours per person, that is exactly 78 full-time equivalents.
At the busiest time you need 20 / 0.70 = 28.6, so 29 people scheduled at once. Note that the roster of 78 is larger than 29 because a seven-day, 24-hour operation needs several people for each seat over a week: nobody works every day, and nobody works every hour.
What this does not do
A roster size is a headcount estimate, not a schedule. It assumes shifts can be arranged to cover each interval, that people can work the contract hours you enter, and that nothing limits how shifts are combined. Real schedules also have to respect rules such as rest between shifts, days in a row, night work, meal timing, public holidays and skills, and the headcount you need rises when those rules bite.
It also takes the requirement as given. If your requirement is a guess, the roster is a guess. Build the requirement from forecast volume and handle time by interval, using Erlang C, and check it against actual results.
Common mistakes
- Dividing weekly hours by 40 without adding shrinkage first.
- Using headcount where full-time equivalents are meant. Part-time staff count for the share of the hours they work.
- Planning for the average hour instead of the peak, so the busy interval is always short.
- Forgetting that a 24-hour operation has night and weekend shifts that nobody wants, which affects attrition and overtime.
- Ignoring ramp-up. New hires do not reach full productivity at once, so hiring waves need to start before the gap opens.
Using the result for hiring and shift planning
Treat the roster size as the headcount you need working, not the number you have to hire. Subtract the people you already employ, then allow for attrition and the weeks it takes a new hire to finish training before they count toward coverage. In operations with high turnover, the hiring plan is often bigger than the gap.
Then check the shape. A roster that is large enough in total can still be short in the busy hours if too many people are rostered into quiet ones. Compare the peak number of people you must schedule at once with how your shifts overlap, and look for intervals where the requirement is higher than the people you can field. That interval-level view is where schedules actually fail.
Typical ranges to check against your own operation
Full-time contracts in many countries sit around 40 to 48 paid hours a week, but your labor rules and contracts decide. Shrinkage is often discussed in the range of about 25% to 35%. Typical ranges, verify them for your operation and jurisdiction before you plan hiring on them.

