Work Sampling: A Method for Assessing and Monitoring Productivity

Profitability is driven by productivity. You want to increase profits, so your task is to maximize productivity, which is basically a measure of output per dollar of capital deployed. It does not matter what percentage of your dollars (aka capital1) is spent on labor or equipment; the task is the same — to maximize yield.

Two elements drive productivity:

Use rate — total time units a person or machine is in productive use, divided by total time units of observation time.

Example 1: Bulldozer #224 was used 52 hours over the past 14 days, a use rate of 15.5% (52 divided by 336), assuming that a bulldozer, in theory, could be used 24 hours a day.

Example 2: Tim Tomer was observed 100 times over the past three months; 61 times he was working productively. His productivity rate was 61%.

Output rate — units of output produced by a person or machine, divided by time of work or operation.

Example: Bulldozer #224 moved 87 cubic tons of material during the five days ending September 22. Workdays were eight hours each, so #224 averaged 2.175 cubic tons of material per hour of uptime during this week.

The focus of this article is assessment of use rate. If a business assesses and tracks its use rate and takes steps to increase use, productivity increases and profits improve.

The way to assess use rate is work sampling. It lets you identify, through observation, the percentage of time employees and/or equipment work productively. If it’s equipment, “working productively” depends on what the piece was acquired to do — mold parts, pump water, take pictures or be rented out. If it’s a person, “working productively” means “doing the task the person was hired to do.” It could also be defined as “doing a task that generates revenue or leads to generation of revenue.”

Note: Herein we use the terms productivity and use rate synonymously, to a degree. When we assess people, we talk of productivity. When we assess equipment, we talk of use or usage.

When a company uses work sampling to assess its productivity it is referred to as a work sampling study. Employees and/or equipment are observed repeatedly, over a defined period and, at each observance, the productivity of each person or piece of equipment is recorded. It’s simple in concept, but to garner meaningful and reliable data, the methodology must be soundly developed and executed.

Work Sampling in Practice

2009 Article Series

Joe Smith, owner of ABC Construction, was not happy with his firm’s unprofitability. His efforts to identify the cause and therein fashion a remedy led him to look into his employees’ productivity. Do they work efficiently? Do they stay productive throughout the workday?

He decided to make an objective assessment. He located and consulted with a work sampling expert to design the program and consult on implementation. Once the plan was developed, Joe met with his employees to explain the initiative. He explained what would be measured, who would do the measuring and in which situations. Joe emphasized to the employees that this study was to improve company-wide productivity and profits and that a more productive, profitable company would mean more job security and higher profit distribution bonuses. Assessment of individual employee performance was not the objective.

He also explained to his employees:

Work sampling process. Work sampling is a series of instantaneous observations of work-in-progress done randomly over a specified period. These observations, known as samples, are compiled to show the percentage of the day workers perform productive work.

Identification and definition of work categories. Each observation or sample would be recorded in one of three major categories: productive, supportive or recoverable.

  • Productive: Defined as “getting the job done.” For example, if the observed is a draftsman, he or she is drafting. If it’s a doctor, it’s working with a patient. If it’s a salesman, it’s talking to a customer or prospect.
  • Supportive: Defined as any activity that is required to maximize the time that can be spent getting the job done. For example, if the observed is a draftsman, he or she may be cleaning the drafting table, updating the software, talking to a client or taking continuing education classes. If it’s a doctor, it may be inputting client information, taking continuing education classes or talking with his/her nurses about a patient plan. If it’s a salesman, it’s writing notes from a call, writing a letter or proposal to a client, or studying competitor data.
  • Recoverable: Defined as “not adding any value at all or contributing in any way to getting the job done.” For example, whether the observed is a draftsman, doctor or salesperson, the observation should be recorded as recoverable if the observed is talking to someone on the phone about non-work matters, using the restroom, taking a “smoke break,” talking to co-workers about non-work matters, playing solitaire, or waiting unproductively because “the computers are down.”

Methodology. Employees are told they will be observed but not when. Observers follow preplanned routes, and the study continues until the predetermined number of samples is collected. Each time an employee is observed is one sample. It is the observer’s responsibility to document what the employee is doing when the observation takes place. The observer must document the activity based on the predetermined work categories as he or she proceeds through the job site tour. Was the action productive, supportive or recoverable? Each employee may be viewed multiple times, depending on the length of the study and how many observations are needed.

How many samples are needed? It depends on the level of confidence desired, the allowable margin of error and the current ratio of productive to unproductive time. You can easily figure out how many samples you need to get results with a 95% confidence level and a low 3.5% degree of error by:

  1. Taking an initial sample of 30 observations and determining the percentage of time occupied in productive use for the sample.
  2. Looking in the table below for the number of samples needed, based on results of your preliminary sample.

For example, Joe took 30 samples (i.e., 30 routes around his facility) and calculated the productivity rate at 7% (i.e., productive time plus supportive time). Using the table (top right column), he needs his full study to have at least 578 observations.

Evaluating Results

Computer software can analyze work sampling results. In most cases, observers record the work by hand on worksheets and input the data into a Microsoft Excel or similar spreadsheet software. There are also some new programs that let you record the data directly into a smart phone.

But the study results do very little for you. It’s up to you to use the data to your advantage. How?

First, share the results with your employees. Simply helping them understand how “little distractions” add up should help improve productivity.

Second, talk to your observers about the types of distractions they saw. Then develop ways to eliminate them.

Third, conduct follow-up sampling studies and compare results. Consider devising company-wide goals and rewards for improvements in productivity.

1 Capital also can be parsed into two types: debt and equity.


For more information or help with a work sampling project, contact one of the following:

Jack Greene, Jackson Productivity, 843-422-1298

Nelson Lee, Laubrass, 513-624-6629

This article originally appeared in The Business Owner Journal, the periodical of choice for owners of small and midsize private businesses. All rights reserved, D.L. Perkins LLC. © 2014.

This publication is intended to provide general information on the subject matters covered. It is sold and distributed with the understanding that neither the publisher nor any distributor or advertiser is engaged in providing legal, tax, insurance, investment or other professional advice. The advice of a qualified professional should be sought before any reader applies a concept presented herein to his or her particular situation or business.

D.L. Perkins, LLC is solely responsible for this content.


Leave a Reply