When organizations need a new shift schedule, their managers often begin by searching for either:

• A specific work pattern, e.g., Pitman, DuPont, 6-on-3-off, or 4-on-2-off
• A desired schedule format, e.g., 10 or 12-hour fixed shifts

This approach often will produce inferior results for three reasons:  (1) workload variations, (2) crew requirements, and (3) overlapping shifts. Let's take a look at each of these reasons in more detail.

Workload Variations
Most smaller organizations don't have a steady volume of work. Some have a workload that varies by time of day. For example, police and communications centers are often busier during the evening hours. Customer service centers may have more calls during the daylight hours, such as 4-person coverage from 0800 to 1700 and 2-person coverage from 1700 to 0800. Some organizations don't have a steady work volume throughout the week. For example, they may have lighter workloads on the weekends, such as 4-person coverage Monday to Friday and only 2-person coverage on Saturday and Sunday.

Most 24/7 schedule patterns are designed to provide the same coverage around-the-clock, e.g., 3-person coverage on a 24/7 basis. This are inappropriate for organizations whose workload varies by shift or day of the week. It's important to keep in mind that the primary reason for having a schedule is to match the coverage with the workload. Managers tend to forget this key point when they search for a schedule pattern they hope will improve morale or reduce turnover.

Crew Requirements
Work patterns are designed for a specific number of employees or crews. For example, the Pittman 12- hour shift pattern (also known as an Every Other Weekend Off schedule) is designed for 2 crews per shift. The 6-on-3-off pattern using 8-hour shifts is designed for 9 crews (or employees) on each shift.  Yet organizations use these patterns with the wrong number of crews all the time.

I just finished a job for a client that was using the 6-on-3-off pattern with only 5 employees on each shift. As a result, their coverage varied from day to day in a haphazard fashion. Employees liked the schedule pattern, but it was not an effective solution for the organization. Starting the search for a new schedule with a specific pattern in mind will rarely result in the most effective choice for the organization.

Overlapping Shifts
When organizations choose a desired shift length in an attempt to satisfy employees, they fail to consider the impact on coverage. This usually occurs with 10-hour shifts. Employees want to get an extra day off each week and they dread the idea of 12-hour shifts. So 10-hour shifts seem to be an ideal choice. The problem is that they come at a high price for the organization.

Since the organization is now working 30 hours a day (three 10-hour shifts), this requires 25% more personnel or 25% less coverage. For example, a group of 16 employees on 8-hour shifts had 4-person coverage around-the-clock. After changing to 10s, the coverage dropped to 3 people. In addition, there are six hours every day in which the shifts overlapped. They now had 6 people working for this six-hour period every day when they only needed 4 people.
 

Summary and Recommendations
Have you ever reflected on why you have a schedule? Yes, satisfying the employees is important. But that's not the reason for the schedule. The schedule is there to ensure the organization deploys its resources as effectively as possible. Many managers seem to forget this critical point. They hear about a popular pattern or they think 10-hour shifts will make everyone happy. They don't think about the impact on the organization.

I'm not saying that we ignore employee preferences. I'm simply saying that they're not the starting point in the search for a better schedule. Here is the approach I recommend:

1. Determine the hours of operation. Does the organization operate 24/7 or something less than that? If the hours are less than 24/7, make sure they are based on business requirements, not on shift length preferences. For example, a recent client was looking for 20/7 schedule. He didn't have enough staff to achieve the desired coverage levels, so I questioned him about the hours. He said he chose them in order to justify using two 10-hour shifts. He forgot that working an extra 4 hours a day would require 25% more personnel.
2. Determine whether the workload is constant or variable. If it's variable, two general approaches can be followed. One way is to establish base coverage using either 8-hour or 12-hour shifts and then add an overlay shift (sometimes called a Power shift) to boost coverage during the busy period. The other way is to combine different shift lengths or use 10-hour shifts and match the overlapping shifts with the busy period(s). 
3. Identify the coverage requirements (i.e. the number of employees you need to show up for work - by hour of the day and day of the week). Many managers mistakenly specify the coverage requirements using the current shift length. This simplifies the task, but it may not be the best match for the workload. If you can identify the work volumes and coverage requirements by hour of the day, that is ideal.
4. Decide which shift lengths are the best fit with the coverage requirements. If the workload is constant, 8-hour or 12-hour shifts would be best. If the workload varies, which shift length(s) are the best fit? If more than one shift length would work, which is preferred? Notice that shift length selection is the 4^th step in the process, not the first. 
5. Calculate whether the staff size is adequate to produce this coverage. If not, overtime may have to be built into the schedule or more employees hired. The free staffing calculator on my website can help with this step. Go to:  http://www.shift-schedule-design.com/Staffing_Calculator. If there is more staff than necessary, can these "extra" resources be used to build relief time into the schedule to cover absences?
6. If the resources are adequate, it's time to start the actual shift schedule design. Often you can apply common schedule patterns as a starting point, and modify them to fit your unique circumstances. Suppose you have 5 employees on a 12-hour day shift. This will require a 5-week schedule. You could start with a common 2-week pattern, double it to 4 weeks, and add a new pattern for the 5th week. An important consideration is to use a pattern that matches your organization's pay week. If the original pattern was intended for a different pay week, you could end up with unnecessary overtime and you could lose some of the weekends off. Again, please note that pattern selection is the 6th step in the process - not the first step.
7. We always try to create several different shift schedule options to show a range of possibilities, such as the number of consecutive days worked, different shift lengths, different work patterns, with or without relief coverage, and so on. The goal is to both satisfy client specifications and explore possibilities the client may not have considered.

The bottom line is that the search for a new schedule begins with the business requirements, not with employee preferences. If you reverse this by searching for a schedule that you think employees will be happier with, the organization will suffer the consequences. My recommended approach does not ignore employee preferences; it just doesn't start with them.