Staffing a service desk or back office to an SLA in hours or days
Tickets and cases do not queue like calls. How to staff work with a turnaround SLA: throughput first, then the backlog the deadline allows, then the peak that breaks it.
Published
Service desks, claims teams and back offices share one property that separates them from a phone queue: the customer is not waiting on the line. Work can wait, within a deadline, and that changes the staffing maths completely. This guide is how to staff to a turnaround SLA, using the email calculator, which is built for exactly this shape of work.
Why Erlang is the wrong model
Erlang C staffs so that a customer who is holding does not hold long. With a deadline of hours, holding is not the problem; the backlog is. Put a ticket queue through Erlang C with an eight-hour target and it collapses to the throughput answer, which is right but says nothing about whether the deadline holds through the morning peak. The right model is throughput plus a backlog limit.
Step 1: throughput
Tickets per day times handle time per ticket is the work. Divided by the staffed hours in the day and the occupancy you plan to, it is the agents on task. Back-office occupancy can run higher than voice, 85 to 90 per cent, because there is always work waiting; it cannot run at 100 because switching, systems and questions exist.
Step 2: the backlog the deadline allows
If arrivals were flat, throughput alone would meet any deadline. They are not: there is a busy part of the day, and staffing at the average lets a backlog build during it. The ticket at the back of that backlog waits longest, and the turnaround caps that wait. The email calculator models the peak as a factor and a length and finds the capacity that keeps the longest wait inside the SLA. A long SLA lets you staff near the average; a short one pushes you to the peak.
Step 3: shrinkage and the day
Agents on task become agents to schedule after shrinkage, and a daily figure becomes FTE over staffed days and contracted hours, exactly as for voice. Back-office shrinkage is often lower than voice because there is less coaching and no ACD to be logged out of, but it is rarely below 20 per cent.
Multi-touch cases and priorities
A case that needs three touches is three tickets with their own handle times, or one ticket with three times the handle time and a turnaround that applies to each touch. Priority tiers are separate queues with separate turnarounds, and they pool badly, because the urgent tier cannot borrow from the routine one’s backlog. Staff the urgent tier to its own peak.
Try it with your own numbers
Each staffed day, after shrinkage.
- Agents on task
- 13.9
- Capacity needed
- 62.5/h
- Agents if arrivals were flat
- 13.9
- Longest wait
- 1.0 h
- Backlog at the end of the peak
- 63
- Handling work per day
- 100.0 h
- Agent-hours per day
- 111.1 h
Concurrent agents actually working email.
Set by daily throughput.
For the email at the back of the peak backlog.
Workload ÷ occupancy.
Show the working
- Workload = 500 × 12 min ÷ 60 = 100.0 hours of handling a day.
- Throughput: 500 emails over 8 staffed hours is 62.5 an hour on average; the peak is 1.5 × that, 93.8 an hour, for 2 h.
- Turnaround: with capacity c, the backlog after the peak is (93.8 − c) × 2 h and takes that ÷ c to clear, so c ≥ 93.8 × 2 ÷ (2 + 4) = 31.3 an hour keeps every wait under 4 h.
- Capacity needed = the larger of throughput and turnaround = 62.5 an hour (throughput binds).
- An agent clears 60 ÷ 12 × 0.90 = 4.50 emails an hour, so 62.5 ÷ 4.50 = 13.89 agents on task.
- Shrinkage: 13.89 ÷ (1 − 0.10) rounds up to 16 to schedule.
Doing this for every interval of the week? Pebble WFM computes the requirement from your forecast and builds the roster. Free month, no card needed.
Common mistakes
- Staffing to the daily average and reporting SLA monthly, so the daily breaches are invisible until a customer complains.
- Clock-hour SLAs on business-hour desks. See above.
- Counting resolution as one touch when the case history says three.
- Running occupancy at 100 per cent because there is always work. There is also always a reason a ticket takes longer than planned.
Where next
- How to staff an email queue with a turnaround time: the model in detail, with a worked example.
- How to calculate FTE from contact volume: from the daily figure to FTE.
Frequently asked questions
- Our SLA is 95 per cent of tickets resolved within 8 working hours. Which number goes in the calculator?
- Eight hours is the turnaround. The 95 per cent is the tolerance for misses, which this model does not staff for explicitly; it staffs so that the longest wait in a normal day stays under eight hours, which delivers well above 95 per cent on normal days and less on abnormal ones. If the SLA is measured over a month, a bad day is affordable; if it is measured daily, staff to the peak factor of the worst day.
- Does it matter whether the SLA counts working hours or clock hours?
- Enormously. An 8-working-hour SLA on a 9-to-5 desk is one day; an 8-clock-hour SLA with the same desk is breached by every ticket that arrives after 09:00. Write SLAs in working hours or staff the night; there is no third option.
Stop doing this one interval at a time
Pebble WFM forecasts your demand, computes the staffing requirement for every interval, builds the roster and publishes it, with self-service for agents and a copilot that can do what a planner can. Explore a sample organisation on day one.