Most 24/7 schedules are intended for organizations with a constant workload. The schedules use four crews of the same size and produce the same coverage around-the-clock. These are the schedules that people tend to talk about, e.g., Southern Swing, Pittman, DuPont, etc.

 

Many smaller organizations can't use these schedules because their volume of work does not remain the same at all times, even though it changes in a fairly regular pattern. These workload variations tend to be one of three types:

• Monthly variations. Some businesses may have a period every year in which they are especially busy. With retailers, this would be the holiday season in November and December. With hotels, it could be the summer months when people take vacations. This is often called a seasonal workload increase.
   
• Daily variations. Organizations also can be busier on certain days of the week. For some, the weekdays are busiest. For others, the weekends are busiest. An example of the latter is a police department that has the highest volume of activity on Friday and Saturday evenings.
   
• Hourly variations. Some groups have workloads that vary by time of day. A call center, for example, may have a huge increase in calls every evening, and only a few calls between 2 a.m. and 8 a.m.

Let's take a look at some of the strategies used to respond to variable workloads:
 

 

Monthly Variations

The common approaches to seasonal workload variations are:

• Temporary help. This is an option for jobs that don't require extensive training. Temps must be easy to hire and get up to speed. The downside is that organizations using temps often fail to let them go when the busy season is done.
   
• Part-time employees. This is an option for smaller groups. They can have workers who are adequately trained, who can step in and boost the coverage when needed. During the rest of the year they can be used to cover absences.
   
• Staff for peak periods. This is the most costly approach since the organization must employ more people than needed for the rest of the year. The only way this might work is if the organization is able to average the pay-hours over the course of the year. They may have 36-hour work weeks for the slow period and 48-hour work weeks during the busy period. A better solution is the liberal use of compensatory time-off for the extra work incurred during the busy months.
   
• Overtime. This is probably the most popular approach. It ensures adequate coverage by employees who are trained and familiar with the jobs. Sometimes this can be used to supplement the above strategies or to provide training and oversight for temps or part-timers.

When the staffing or overtime increases, it may be beneficial to have a new, temporary schedule so that the added personnel or added hours of work can be evenly allocated. Once the busy season is over, the organization can revert to its normal work schedule.
 

 

Daily Variations

The best way to address daily workload variations is with a custom-designed shift schedule. Unlike most 24/7 shift schedules that follow a standard pattern of days worked and days off (e.g., 4-on-4-off or 6-on-2-off), these schedules must be created from scratch to match the coverage with the work volumes. If the weekends are lighter, more weekday shifts would be scheduled. If the weekends are busier, more weekend shifts would be used. Each situation and work pattern is unique because of the different staff sizes, coverage requirements, and workload distributions.
 

 

Hourly Variations

Organizations with a workload that changes over the course of the day (i.e. hour-by-hour) should try to assemble data that reflects the average volume of work for each hour of the day. For example, call centers may use the number of in-bound or out-bound calls, retailers may use the number of customers entering the store, manufacturers may use units of production, and transportation centers may be the number of trips. This can be enormously helpful in guiding the selection of the best scheduling options, namely shift length, shift start/end times, and staffing allocation to the different shifts.
 

Here is an example from a large city's police department. The data was compiled by the International Association of Chiefs of Police (IACP).

 

Hour	Incidents
12 - 1 am	5,853
1 - 2 am	5,343
2 - 3 am	5,141
3 - 4 am	3,153
4 - 5 am	2,267
5 - 6 am	1,646
6 - 7 am	1,410
7 - 8 am	2,148
8 - 9 am	2,896
9 - 10 am	3,107
10 - 11 am	3,360
11 - 12 pm	3,734
12 - 1 pm	4,215
1 - 2 pm	4,299
2 - 3 pm	3,874
3 - 4 pm	5,009
4 - 5 pm	5,574
5 - 6 pm	5,580
6 - 7 pm	5,649
7 - 8 pm	5,463
8 - 9 pm	6,190
9 - 10 pm	6,460
10 - 11 pm	5,769
11 - 12 am	5,724

The city was running 8-hour fixed shifts. The start times were 8 a.m., 4 p.m., and 12 midnight. These are shown in the graph below. The 8 a.m. shift is shown using yellow bars, the 4 p.m. shift uses green, and the 12 a.m. shift uses blue. What really stands out is the variation in night shift work volumes. From midnight to 3 a.m., there is a very heavy workload. The remainder of the shift has a very light workload. The department allocated its officers to each shift based on the average workload for the 8-hour period. For the night shift, this was 3370 incidents per hour. The first 3 hours of the shift have almost twice that many incidents, so officers were not able to respond in a timely manner during that 3-hour period.

 

With highly variable workloads such as this, the ideal solution would consist of a complex set of shift lengths with different shift start times. This is the only way to vary the coverage so that it matches or comes close to each hour's work volume. For example, the shifts may be a 12-hour shift beginning at 8 a.m. with 42-person coverage, an 8-hour shift beginning at 9 a.m. with 6-person coverage, a 10-hour shift beginning at 11 a.m. with 5-person coverage, an 8-hour shift beginning at 2 p.m. with 21-person coverage, and so on.

 

If you are willing to tolerate some differences between the coverage and the workload, it may be possible to come up with a simplified solution by either changing the shift start times or using different shift lengths. Let's look at three possibilities.

 

One way to fix this would be to change the start times for the three shifts. You could start the shifts at 11 a.m., 7 p.m., and 3 a.m. so that the workload variations for each shift are minimized. This is shown below with the same color coding as the previous graph (yellow for the day shift, green for the evening shift, and blue for the night shift). Yes, these are unusual start times. But isn't it more important to match the coverage with the workload? We certainly think so.

Another approach would be to use 12-hour shifts, with the shifts starting at 3 a.m. and 3 p.m. This would be an even better match than the modified 8-hour start times. This is shown below.

A third approach would be to use 10-hour shifts. This would require two separate schedules that started 6 hours apart. This would give them extra coverage for 12 hours a day while the shifts overlapped (i.e. the blue bars in the graph above). The six shifts would be: 0300 to 1300, 0500 to 1500, 0900 to 1900, 1300 to 2300, 1900 to 0500, and 2300 to 0900. It is doubtful that the employees would ever agree to that many different shifts with odd starting times, but it is the best solution if 10-hour shifts are used.

 

We've gone over three possible solutions to this organization's variable hourly workload. While none of them are particularly attractive for the employees, they would benefit the organization by having the coverage more closely aligned with the workload variations.

 

Does your group face a similar situation? Can you find data that reflects the work volumes? If you can collect historical averages by hour of the day, you can assemble a table like the one shown above. There are many possible work volume distributions, so it's impossible to make any conclusions without looking at the data. Maybe it requires a simple adjustment of the shift start/end times. Maybe it requires a different shift length, e.g., 10 or 12-hour shifts. Or maybe it requires a combination of multiple shift lengths with staggered start times.

 

Send us your data and we will help you design the most efficient set of schedules. Just click on the link below: