measuring work doesn’t take split-second timing, but it does require valid standards
LOOKING OVER a warehouse worker’s shoulder and counting how many minutes he takes to finish a task isn’t nearly as exciting as clocking the lead runner’s progress at the Boston Marathon. But just as shaving a few milliseconds off the previous record sets a new milestone in the race, making small improvements in daily activities can result in huge leaps in productivity and personal efficiency on the job.
The problem is, assessing labor isn’t easy. Although measurement practices in the fulfillment industry have become much more sophisticated in the last two decades, work remains difficult to evaluate because we tend to gravitate to the readily quantifiable when we measure – and as anyone who has tried will tell you, “readily quantifiable” and “labor measurement” do not mix very well.
This is also the reason labor management software isn’t as highly developed as programs in inventory or transportation management, the other two largest cost areas in most operations. For these functions, effective and reasonably priced applications have been available for some time, but until recently, practical solutions for labor were hard to find at any price.
As with bar coding, the grocery industry led the way in implementation of labor productivity improvements decades ago. Today, the mix of a tight labor market, a sharper focus on labor by software developers, and rising distribution costs has resulted in significant gains in work measurement in other industries.
Managers are starting to use tools to optimize labor resources even in highly dynamic work environments, and they are finding that the results are well worth the investment.
Brass-ring approach Not surprisingly, the organizations spearheading this charge are the same innovative industry leaders who have sought competitive advantage elsewhere. They range in size from less than $100 million to multiple billions of dollars in sales, and from single-site operations to multi-tiered distribution networks.
A typical organization might have an average of four to seven sites, have posted recent growth at or above 25% per year, employ hundreds of employees on at least two shifts five days a week, and pay average fully loaded wages of between $12 and $18 an hour, union or non-union.
As a rule, such an organization has several objectives for a productivity initiative that might include:
* improving productivity through labor measurement;
* improving labor resource utilization; and
* controlling or reducing payroll expenses.
To attain those goals, many companies have turned to an approach found throughout the grocery industry – engineered standards, combined with advanced data capture and reporting technology.
As a strategy, the innovative organization looks to develop a combination of team and individual measures best suited to each operation. The company applies these standards throughout the system so that all associates receive fair treatment and “everyone has an equal chance for the brass ring.” Interfaces between the data-reporting function and the resident warehouse management system (WMS) permits real-time reporting, which benefits both management and the individual employee. Work plan execution and current performance are then accessible to those most qualified to affect both for the better.
Tread softly The ideal productivity improvement initiative has two major components: (1) engineered standards to establish goals and improve the ability to predict work content, and (2) labor reporting software to make results available quickly and in the most suitable format for taking action. The first piece requires industrial engineering experience and strong people skills. The second piece usually appears in the form of commercially available software. Because of the specialized calculations needed and the complexity of predicting and tracking labor activity, in-house development of this kind of application is generally not recommended.
Although the situation is changing, few of even the top-tier WMS vendors provide solutions that accommodate both engineered standards calculations and highly flexible software tools to report labor activity in real time. The functionality is specialized and, to date, has not warranted much attention from most WMS developers. The most powerful and capable applications come from other software providers who have focused on the labor aspect of distribution. Professional educational organizations such as the Council of Logistics Management, Warehousing Education and Research Council, or consulting firms are usually good resources for locating and recommending software for this purpose.
The best programs support timely reporting of both performance and utilization, assist in balancing labor across all functions in the warehouse, facilitate labor allocation planning, and enable analysis of historical data housed in the database. “There are very few stand-alone tools to choose from, but they continue to improve in functionality,” says one software consultant. “Developers are listening to users and have gotten some great ideas from clients.”
Standard issue The heart of the issue is always the establishment of a fair and objective means by which to measure work. It is complicated by the fact that neither the work itself nor the individuals who perform it are the same from instance to instance.
The single most effective process for establishing a reliable, objective measurement for performance is the classical engineered standard developed through detailed time study of each task in context. Although specifics may vary, the general method is as follows:
Establish best practices. The process begins by identifying and implementing the optimal methods and procedures for performing the work (putaway, whole or broken case replenishment, case or pallet picking, and so forth). “There’s no point in developing metrics for the inefficient or inconsistent methods – they will not have much value,” says a senior operations executive. “Rather, you want to train people to do things in the best way and plan that way as well.”
Build standards. After implementing best practice improvements, an engineer studies the work, breaks each task into its constituent elements, and develops an allocation of time for each occurrence of each element in that work. For instance, if an order-picking task involves one line and three cases, the unit of work would contain elements and related time for picking up the order, moving to the location, picking each of three cases, and moving those cases to outbound staging, among other things. Some highly varied tasks may use 30 or 40 elements.
A precise calculation of the total “earned” time for this task assumes that the person performing the work is using mutually agreed-upon methods. Managers’ expectations should be that a normal person with normal skills can sustain this pace for the duration of a shift (eight or ten hours) without undue fatigue.
For some tasks (loading, pass-along picking, or unloading a truck), an individual gauge may be especially difficult. In those cases, the most common solution is to create a standard time for the task and relate it to the total time for the entire team.
By studying many iterations of each work task and element, the engineer is able to arrive at time allocations that are highly reliable across a wide sample of the work employees are asked to perform and can accurately predict how long any given work assignment should take. However, facility managers need to validate any standard before applying it to a work situation to assure that everyone is satisfied that it is fair and accurate. The next measurement activity then becomes the process of capturing actual time applied to the task and comparing it with the predicted or calculated “standard” time.
Capture actual performance data. This process need not be complex, but to be valid, it must be complete. Beginning and ending times for all tasks and all users are essential. Radio frequency (RF) devices can capture data automatically, or bar code scanning or keyboard entry can record it manually. The key is to minimize the time required to capture the information while assuring that it is captured properly.
Calculate performance. This step simply relates standard time to the actual time that a person takes to accomplish a task. If, in the course of the task, time elapses that does not relate to the work and is beyond the individual’s control (such as a fire drill or a battery change), an adjustment is made to remove that portion of the total time from consideration (exception time).
By dividing standard time (360 minutes) by the actual time (400 minutes), the system returns a performance rating of 90% (or in a zero-based form, 10%) of the expectation or standard.
Calculate utilization. Performance measures the employee’s work, whereas utilization measures management, or the application of available time (labor) to open work, and it indicates how well managers have used the labor resource. For example, there are 450 minutes in an eight-hour shift (with two 15-minute breaks; lunch is not included). If 400 minutes of that time were applied to relevant work, that person’s utilization rate would be 89% (see chart on page 38).
Calculate total productivity. This measure is derived from relating the two results discussed above to show how efficiently work is accomplished. Multiplying performance for a whole period (90%) by the utilization factor (89%) shows that total performance for the facility is 80%.
Peak performance So far, nothing has been improved except work methods – we have only established measurements. Labor standards and reporting software have value only when they facilitate change for the better. The whole point is to position users and managers to be able to make improvements. Here’s how to put information into action:
Implement. Following the validation of the standards for each area, managers need to work with all employees to ensure that they understand how the standards were developed, how the reporting process works, and how to use the right methods. The implementation of the productivity enhancement process takes place over several weeks, with the expectation that workers will attain 100% performance level by the end of that period.
The illustration on page 38 depicts a typical learning curve for implementation of a standards program. The dashed line represents an approximate rate of expected improvement across an eight-week period. The precise rate of improvement will vary for each individual (the other lines), but needs to be planned and communicated to everyone involved.
Observe, offer feedback, and train. Supervisors and managers are frequently better prepared to manage inventory, transportation, and information than they are to work with people. With few exceptions, leading firms in the direct-to-customer business have made a strong commitment to providing their management the tools and training needed to do the job properly. Supervisors receive training to identify those who need help in attaining the graduated rate of improvement during the implementation period. The primary vehicle for accomplishing this task is personal observation and interaction with each person during the course of his work.
By learning how to interpret the data generated by the process, supervisors can quickly pinpoint those who need more training and counseling. It’s important to establish a high level of feedback for all parties at the outset. This includes daily and weekly reports by employee, supervisor, department, shift, and facility, as well as an option for workers to inquire about their status at their convenience (see chart on page 40).
This is the stage where full realization of productivity enhancement benefits is most at risk. Productivity improvement is as much about training managers to use the techniques and develop their teams as it is about capturing performance data. Ironically, it is also the area of greatest opportunity. Supervisors who are actively involved in helping employees attain reasonable goals often achieve results well beyond initial expectations.
Plan and analyze. Managers can now begin to use performance data to build schedules for future periods and to monitor work as the plan unfolds during a shift. Now is also the time to adjust labor levels by department, zone, wave, or function to optimize production, minimizing both downtime and overtime.
Maintain data validity. Many other refinements and nuances can be added to the process. By integrating reporting data with the payroll time clock function, for example, managers can account for all time being paid for operations work. With an effective report writing program, managers can create or customize reports to suit individual needs, in addition to keeping corporate management informed about the status of the project.
Swift returns One the most appealing aspects of a productivity effort is that it is usually quite cost-effective and begins generating meaningful benefits well before the project is completed. It is not unusual to have a return on investment for a project that is under six months. Most are less than one year.
The precise degree of improvement is obviously a function of the individual company, and often, the particular facility. One way to locate your operation on the spectrum is to rate the quality of your supervisory group.
Workforce demographics also affect results. The employee performance chart on page 42 illustrates a common distribution of performance levels across a representative group of warehouse workers. The least productive 25% are often the group with the highest rate of turnover, as well as a greater training burden.
Considering these factors, you should be able to approximate the level of improvement possible in your facility. By relating productivity improvement to your payroll, you can estimate the annual dollar savings available.
Acme Endeavours, a food supplier, had 15 facilities organized into four divisions with more than 1,300 employees. The sites varied widely as to size, volume, number of employees, and wage rates. Management chose a pilot site for productivity improvement that promptly outperformed every other location by a significant margin. Despite having the second-highest hourly wage rate of all divisions ($15.05), the pilot location came in next to lowest in cost per case shipped ($0.0172) and had the highest cases-per-hour rate in the entire company (113.9). Based on savings in the pilot site, savings for the whole organization were estimated at more than $5 million in the first year. The ROI for every division was under one year, with the smallest (the least potential for improvement) taking the longest time – ten months (see charts on pages 43 and 44).
And that ain’t all! During the course of implementing productivity improvement programs, some organizations have made other important discoveries that enable them to enhance performance.
* It is important to conduct multiple observations with each employee and in each work area during project implementation and beyond.
* Workers and managers must be trained more thoroughly than is usually expected, especially in managing, interpreting and using performance data; the benefits of such training are exponential.
* More emphasis on frequent communication of all kinds is vital, including real-time feedback on performance, more frequent counseling for improvement, and wider participation in decision making about productivity.
* As employees attend more closely to their performance, quality rises along with productivity.
“People focus more on their jobs,” says one labor expert. “When people know what’s expected, how they’re doing, and that their contribution is fairly and objectively documented and recognized, they tend to become better, happier workers – the kind everyone is looking for. And they tend to stay longer.”