Workforce planners say shrinkage is “around 30%” as casually as people say the monsoon arrives in June. If you have not planned a roster, the figure can feel oddly large. Where would nearly a third of paid time go? The best way to find out is to follow one eight-hour paid shift and one hundred seats through a single example. Every number here is made up for the walk-through, so use your own process data in real life.
Step one: define it
Shrinkage is the share of paid, scheduled time during which agents are not available to handle customer contacts. Formally: (paid time minus time available for contacts) divided by paid time. Definitions vary between companies, which is why two centres quoting “30%” may not be comparable. Some count breaks as shrinkage, some treat them as part of the schedule. Some include unplanned leave, some track it separately. Before comparing, check the definition.
Here, we will count everything that takes a paid agent away from the queue.
Step two: list the pieces for one shift
An eight-hour paid shift is 480 minutes. Our imaginary process has these average losses per agent per day:
| Item | Minutes per day | Share of 480 |
|---|---|---|
| Breaks and lunch | 45 | 9.4% |
| Training and coaching | 20 | 4.2% |
| Team meetings and huddles | 15 | 3.1% |
| System downtime | 10 | 2.1% |
| Unplanned absence (7% of days) | 33.6 | 7.0% |
| Planned leave (5% of days) | 24 | 5.0% |
| Total | 147.6 | 30.75% |
The final row is what gets called “about 30%”. Notice that the absence rows are averages. On a day when 7% of agents are absent, the lost time is not spread evenly: it is a whole shift missing for some people. That is why planners treat it differently from breaks, which are scheduled and predictable.
Step three: split planned from unplanned
In our table, breaks, training, meetings and planned leave are known in advance, so the schedule can arrange them: stagger breaks, put coaching in quiet periods. Unplanned absence and system downtime cannot be scheduled. This split matters in daily management. You can plan around the planned items, and the unplanned ones are the ones that damage service level on the day. Reducing planned shrinkage means looking at whether every meeting is needed. Reducing unplanned shrinkage means looking at the root causes of absence, which is a people question more than a scheduling one.
Step four: turn it into headcount
Suppose the demand forecast says you need 100 agents actually available at the desks. If shrinkage is 30.75%, each hired agent is available for 69.25% of paid time. Required headcount is 100 divided by 0.6925, which is about 144.4, so you need 145 people on the roster. The common mistake is to compute 100 plus 30.75%, giving 131, which comes up short. Shrinkage is a share of the paid time, not an addition to the needed number.
Now see how sensitive this is. If true shrinkage is 35% and you planned at 30.75%, you needed 100 / 0.65, or about 154 people. You are nine short, and the gap shows up as long queues and abandoned calls in the afternoon. If shrinkage is lower, say 25%, you needed about 133, and you have been paying for around a dozen people to do nothing at the margin. A five-point error in shrinkage is a large cost either way, and a good reason to measure rather than copy a number from another process.
Occupancy is separate. Even if all 100 seats are filled, agents will not be on calls every second. A target occupancy of, say, 85% is a different assumption, and the two combine in the full staffing calculation. If 100 desks at 85% occupancy and 69.25% availability are in play, the productive share of paid time is well under 60%, which is a number that surprises managers.
Step five: watch how it behaves
Shrinkage is not constant. It rises around festivals, exam seasons, during monsoon commutes, and in the weeks of a new product launch when training increases. It is higher in the first weeks of a new hire’s life, when training and nesting absorb time. Planners therefore track it weekly and by interval, and build seasonal assumptions: a Diwali week is not planned with the same figure as an ordinary February week.
Finally, treat the figure honestly when it is high. A centre with 40% shrinkage is not suffering from bad luck. Something is going on: heavy training load, unrealistic leave rules, high absenteeism, unreliable systems. The number is a prompt to ask which of the rows in your own table is biggest, and why.
Figures and names in the examples are illustrative.
Leave a Reply