Original Publication: Customer Management Insight - October 2008
View Online
Ever wonder aloud where your agents are? Rostered staff factor (RSF) or shrinkage calculations can help.
If you’ve ever looked at a supervisor monitor or conducted a headcount on the floor and wondered, “Where is everybody?” ... you’re not alone. I have a hunch that every call center manager has asked that question at one time or another.
Yes, part of the answer is schedule adherence. But another part is realistically anticipating the things that can occupy agents’ time.
One time-tested approach for predicting the things that will keep agents from the work at hand is to come up with factors that reflect these issues. Rostered staff factor (RSF), alternatively called an “overlay,” is a numerical factor that leads to the minimum staff needed on schedule over and above base staff required to achieve your service level and response time objectives. It is calculated after base staffing is determined and before schedules are organized.
This Is Forecasting
Calculating RSF is a form of forecasting. The major assumption is that the proportion of staff off the phones will be similar to what is happening now. In other words, if one person is on break in a group of 10, 10 people will be on break in a group of 100.
The mechanics of RSF are straightforward:
1. Enter the base staff required by half-hour. What base staff includes will depend on the structure of your groups. For example, if you have separate agent groups for transactions that must be handled when they arrive (i.e., phone calls) and transactions that can be handled at a later time (i.e., email), the base staff entered represents one of those groups.
You will need RSF calculations for each group. On the other hand, if you set up groups that handle both types of transactions, base staff is first calculated for both types of work separately and then added together. You would then calculate RSF for the combined group.
2. Identify the things that routinely keep agents from the workload. The next three columns reflect the numbers of staff absent, on break and in training, as they now occur. These categories are just examples, and you can include research, outbound calls (those that are not part of talk time or after-call work, nor otherwise included in base staff calculations) and other activities. You may also want to further subdivide the categories. For example, absenteeism can be divided into planned absenteeism, such as vacations, and unplanned absenteeism, such as sick leave.
3. Add base staff to the number of agents who will be away from the workload, for each half-hour. The “on-schedule” column is the sum of the entries in previous columns, by half-hour.
4. Calculate RSF. The last column is derived by dividing the staff required on schedule by base staff required for each half-hour. The proportions are the mechanism you will use to project future schedule requirements.
5. Use the factors when organizing future schedules. The result of these calculations is a set of factors reflecting expected requirements by half-hour. You multiply them against the base staff you will need when assembling future schedules. For example, if you are putting together a schedule for four weeks from now, and you need 32 base staff between 8:30 and 9:00, you will need to schedule 40 agents (32 x 1.25) for that half-hour — plus any staff required to be working on projects, in meetings or anything else not included in the calculation.
Are RSF And Shrinkage The Same?
For years, many have used the terms rostered staff factor (RSF) and shrinkage interchangeably to refer to the reality of needing more staff on schedule than the base staff required to handle customer contacts. No problem there, and in that context the terms refer to the overall concept of needing more staff on schedule than base staff required to handle the workload.
Others use the terms more specifically — nothing wrong with that either, as long as the context is understood. When differentiated (and it’s subtle in concept) shrinkage refers to how much loss there is between scheduled staff and base staff, while rostered staff factor looks at how much needs to be added to base staff required to reach schedule requirements.
What’s not subtle is how the calculations for each are used when they are diffentiated in this way. Consider an example where base staff is 40, 10 agents will be involved in other activities, so the schedule requirement is 50 agents:
The RSF is 50/40, or 1.25
Application: 40 x 1.25 = 50 agents
(or)
The shrinkage (loss) = 10/50, or 20%
Application: 40 / (1 - 20%) = 50 agents
Uh, oh. Here’s the rub. Some mistakenly apply shrinkage in place of what should be RSF, and end up with too few scheduled agents. And others mistakenly apply RSF in place of what should be shrinkage, and end up with too many scheduled agents.
Investment analysts often have to remind clients of a similar principle in the world of finance. If a $5,000 investment drops by 20 percent, the new balance is $4,000; $4,000 would need to grow by 25 percent (4,000 x 1.25) to again reach $5,000. Many call center managers have found themselves in a similar education process: “Yes, scheduled staff shrinks by 20 percent, which is why we need 25 percent additional staff above base-staff requirements for that increment!”
In that sense, RSF has some advantages: it’s easier to explain. Correctly applying either method at the interval level will lead you to the same answers — however, you have to take shrinkage a step further to calculate staff required, while RSF provides requirements more directly. The RSF methodology also tends to fit nicely into the logical progression of planning: calculate staff and supporting resources, determine and apply RSF, then organize schedules. Shrinkage tends to jump from schedules, to determining loss or shrinkage, to applying the formula to base staff, in order to get back to schedule requirements. It’s a small point that doesn’t matter if either approach is used correctly, but I prefer the RSF approach when explaining these principles to others.
Shrinkage has some advantages, as well. It’s the term that, more often than not, has been adapted by workforce management suppliers (consequently, I run into people who have been in the industry for years and have never heard of RSF). And it’s used more broadly in other environments — e.g., in retailing environments, inventory shrinkage can be caused by loss, damage or theft. (That can be a mixed blessing - if you do an Internet search on “shrinkage” you will find pages of definitions and sites having nothing to do with call centers.)
Finally, the time horizon you are considering may impact the approach you use. As described here, RSF and shrinkage are for scheduling (near and medium term) purposes. When you look at a longer planning horizon, e.g., to determine full-time equivalent requirements, you will likely want to use a shrinkage-like approach in the methodology. But caution is in order: Any mismatch of time — e.g., taking a calculation for shrinkage or RSF determined over a broader timeframe, and applying it to smaller increments — simply doesn’t work.
In short, there are different ways to get from point A to point B. When used correctly and in the right context, both shrinkage and RSF will lead to the same answer.
Update As Needed
We recommend that you initially produce a table of factors for each day of the week and for each agent group you will be scheduling. Then, adjust the calculations as circumstances dictate (i.e., for vacation season or major changes in training schedules).
Also, larger call centers will pick up some accuracy by going through this planning step in 15-minute increments. Like any aspect of planning, examine how accurate your predictions are compared to actual results, and adjust accordingly.
As the environment becomes more complex, this step is becoming more important. As a prerequisite to a stable and well-run call center, it’s worth the effort many times over.
