Help desk staffing is the question of how many agents you need to answer the tickets you get. The help desk staffing calculator turns ticket volume, average handle time, and your service target into a headcount, so you can staff the desk with numbers instead of guesswork.

Most desks are staffed by feel: last year we had six agents, so this year we need six. That works until volume jumps 30% after a new product launch, and then the backlog grows for months before anyone hires. A calculator makes the gap visible before it becomes a queue.

How staffing is calculated

The basic math is simple. Multiply the number of tickets in a period by the average handle time to get total workload in minutes. Divide by the minutes one agent is actually available for tickets in that period, and you get the number of agents needed.

The details are where desks go wrong. Average handle time must include the whole ticket: reading, fixing, documenting, and following up — not just the time on the phone. And an agent is not available for 100% of their shift. Breaks, training, meetings, and leave take a share, usually 25 to 35%, which is called shrinkage.

For phone and chat, where customers wait in a queue, the Erlang C formula adds one more factor: randomness. Tickets do not arrive evenly, so a desk needs spare capacity to absorb bursts. The formula takes your target — for example, 80% of calls answered within 20 seconds — and works out the headcount that delivers it.

Worked example

A desk receives 2,000 tickets a month. The average handle time is 12 minutes, so the workload is 24,000 minutes, or 400 hours, a month.

One agent works 160 hours a month, but with 30% shrinkage only 112 of those hours go to tickets. Dividing 400 by 112 gives 3.6 agents, so the desk needs 4 agents for ticket work alone.

That is the average, though. Monday mornings bring twice the normal volume, and the desk also answers the phone. Adding a peak factor and a queue target usually pushes the answer to 5 agents for this desk — and that is the number to hire against, not 4.

How to use the result

Use the headcount to build the hiring case. A shortfall of one agent is not a small problem: it means every ticket waits longer, the backlog grows, and the remaining agents burn out. Showing the math turns we are busy into we are 25% understaffed.

Use the sensitivity to see what else helps. Cutting average handle time from 12 minutes to 10, through better knowledge articles or a self-service portal, removes the same workload as hiring another agent. Sometimes the cheapest hire is a better article.

Plan for peaks separately from the average. Look at your busiest hour of the busiest day and staff that window, even if it means staggered shifts. A desk that is fine on average but drowns every Monday morning has a scheduling problem, not a headcount problem.

Tip: Re-run the calculation every quarter, and after any event that changes volume — a product launch, a migration, or a new client. Staffing is a moving target, and a number from last year is already out of date.

Staffing and cost per ticket

Headcount is only half the budget question. The other half is what each ticket costs, and that depends on the same inputs: volume, handle time, and agent cost. The cost per ticket calculator takes the staffing answer and turns it into a cost you can compare against outsourcing or automation.

Frequently asked questions

How many agents does a help desk need?

It depends on ticket volume, average handle time, and the service level you want. A desk handling 1,000 tickets a month at 15 minutes each needs about 250 agent-hours a month before you allow for breaks, training, and leave.

What is agent utilization?

Utilization is the share of an agent's working time spent handling tickets. Most desks plan for 70 to 80 percent utilization, leaving the rest for training, admin, and the gaps between tickets.

What is the Erlang C model?

Erlang C is a queuing formula that estimates how many agents are needed to answer a given share of contacts within a target time. It is the standard method for phone and chat staffing, and it accounts for random peaks in demand.

Should I staff for the average day or the busiest day?

Staff for the busy periods, not the average. Tickets arrive in bursts — Monday mornings, after releases, after outages. A desk staffed for the average will drown every Monday.

How do I account for leave and training?

Add a shrinkage factor of 25 to 35 percent on top of the raw requirement. That covers annual leave, sick leave, training, and meetings. A desk that ignores shrinkage is always one absence away from a backlog.