When work is not getting done, or is being performed poorly, it is tempting to claim “lack of people”, or resources, as the cause. In today’s world with high staff turnover, job vacancies and scarcity of qualified staff, this seems like a plausible excuse. But is it really a lack of resources, or might it be a mis-allocation of resources? With the constant introduction of technology, outsourcing, and flatter organizational structures, workload composition changes. If the work is changing, then a chain reaction involving work flows, procedures, job training, position descriptions, and performance management and compensation systems should be keeping pace.

When faced with the problems of poor service levels and low customer satisfaction, I always ask five questions:

1. What is the staff turnover rate?
2. What percent of your authorized positions are vacant?
3. When were procedures for key processes last updated?
4. When was the training for those procedures last updated?
5. When were position descriptions for those processes last updated?

More often than not, the above factors are not in alignment with each other, or in synch with current operating conditions. But if all are good, then where might one look for opportunity?

The Composition of Work

Along with the required knowledge, abilities, and skills necessary to do a job, a position description should list the required duties performed by the incumbent, and how much time is allocated to each. For example, 25% allocated to data entry, 20% to client inquiries, 10% to email, etc. Presumably, these allocations were based on a systematic job/task analysis but, most likely, were based on estimates by people of unknown qualifications.

One technique I have found useful for determining actual time spent to perform required tasks and re-allocating staff to important duties is work sampling. Developed by L.H.C. Tippet in 1934, work sampling is a technique used to determine the proportion of time people spend performing various job tasks – productive and non-productive. If a position description states that the incumbent 10 staff members spend 40% of their time on customer inquiries, but the work sample results show statistically, with a 5% margin of error, that they actually spend 10%, then the difference of 30% X 8 hours/day X 10 people = 24 hours per day mis-allocated.

The other work being performed by these people may be important, but who made that decision, and how was that work prioritized? A properly designed and applied work sampling study can help you baseline existing workload allocations. Other techniques including time studies and standard data systems, like MSD, MOST, and MTM, can help you determine the standard times required to perform specific tasks.

How to Perform A Work Sample Study

Following are the general steps for conducting a work sampling study:

1. Define the population to be studied. This could be an entire organization, department, or work unit, or specific functions across the organization.
2. Determine the positions to be studied, such as analysts, clerks, mechanics, nurses, etc.
3. Define the tasks that should be performed by each position, and the proportion of time shown in the position description. These should come from the most recent position descriptions.
4. Define the key task elements. These are the definitive steps or actions of the task that will be observed during the study. Additional ancillary categories should also be added, such as “walking empty-handed”, “carrying materials and supplies”, “discussing something with another worker”, "being idle", "waiting for work", "being absent", and so on. They may not be defined in the position description, but may be natural supporting elements of key tasks. Be careful to focus on the major chunks of work and not try to create more categories than you can observe and code.
5. Design the study. This includes designing the checksheets and forms that will be used to record the observations, determining how many observations will be required, deciding on the number of days or shifts to be included in the study, scheduling the observations,
and finally determining the number of observers needed. As with any sample, strive for representativeness and no bias. The required number of observations is a straight-forward calculation and driven mostly by the number of people being observed and desired margin of error.
6. Train and calibrate the observers. This is to ensure observers are consistent when classifying what they observe. Failing to calibrate can undermine the credibility of your conclusions.
7. Announce the study, its purpose, and how the findings will be used. If the study is to be conducted in a unionized workplace, make sure the representatives are informed and comfortable with the approach.
8. Conduct several dry runs to test your methodology and consistency of the observers. Make the predetermined number of random visits throughout the workplace and collect the required number of observations. A few dry runs will help the observed become acclimated to the observers and reduce potential bias.
9. After completing the study, summarize the results by position type and tasks. This includes the proportion of time actually spent, and the converted person hours spent for the period of time studied. These values can be further projected to represent hours per week, month, and year, and total FTEs (Full Time Equivalents) people.
10. Compare the actuals to the required standard or expected times. Work sampling is beneficial for determining the proportion of time spent to perform tasks. It will not help you determine the effectiveness of the time spent; however, performance measures will. Utilizing the results of work sampling studies in conjunction with performance measures for the work performed can improve the quality of staffing and time allocation decisions.


Jobs have a tendency to “drift” over time because of many factors. Technology is certainly a factor, as are regulations, culture changes, staff turnover, work ethics, customer requirements, and management demands on the work environment. When customer satisfaction is slipping, but procedures and training are sound, consider the composition of the work being performed. Short of a full-blown organizational capacity study, work sampling, in conjunction with performance measures, can provide quick insight for an effective reallocation of resources more in alignment with your needs.