Taking Measure of Picking

We don't have to tell you that improving profitability is the number-one driving force in today's operating environment. Honing your picking productivity is one way to reduce labor costs, extend facility capacity (buying time until your logistics network is expanded), and improve employee retention — all of which contribute to profitability.

But improving your picking efficiency has subtler benefits as well. It can help boost service levels and worker morale, reduce your dependence on temporary workers, and provide management with better tools and metrics, among other pluses. So how do you improve picking productivity?

If it is true that what gets tracked gets done, the question is: “Is the picking function in the distribution center getting the right things done?” To maximize your picking operations, it pays to track the productivity and accuracy of individuals, work groups, and processes. Establishing goals for the facility and its workers builds an internal culture that will never be content with status quo.

Pickers often operate at a high level, but management is frequently ill equipped to manage and track the productivity. By measuring productivity you may find that your DC layout needs improvement and that your picking methods need modifications.

Performance
vs. UPH

Two methods are commonly used to determine the productivity of individuals and teams. The first is performance, which is calculated by comparing the time it actually takes to complete an activity with an engineered standard. The second, and more popular, method is to compare the associate's hourly production, expressed as units (for instance, orders/lines/cartons) per hour (UPH), with the activity's historical UPH.

Let's say Mike is a picker in the outbound picking operation. During 60 minutes Mike picks 200 units from 80 locations (see chart below). Mike's UPH is 200, which was calculated by dividing the number of units picked by the number of hours worked in the operation. His performance is 100%, which was determined by dividing the total time Mike earned by the total number of minutes considered standard for the operation.

Now let's look at Sally, another picker in the outbound picking operation. During her 60 minutes, Sally picks 130 units from 100 locations. Her UPH is 130. But her performance is 105%, which takes into account that she traveled to more locations than Mike did.
So although Mike had a higher UPH, he did not perform as highly as Sally. That's because UPH is not always the best measure of productivity. It does not take into account variations within an operation that may make a task easier for one associate and more difficult for another. Given this complexity, an unfair productivity measure will often leave associates exhibiting illogical behavior. Therefore, if you want to change the behavior of your picking staff, you must use the appropriate metrics.

Case in point: In February 2004 a $1.1 billion manufacturer/marketer of footwear wanted to improve its distribution operations by implementing a pay-for-performance incentive program within its 300,000-sq.-ft. distribution center. But the engineering team had trouble accurately measuring performance among the pickers. The facility had four cart-pick modules and three pick-to-carton modules in which the cartons were delivered to the picker via conveyor. Each of the pick-to-carton models contained sections of case-flow racks, pallet-flow racks, and full-pallet static racks. Many of the picking modules were similar in layout, process, or product mix, but few were similar enough to accommodate a common productivity measurement.

Because of the differences, the engineering team had to determine separate productivity measures for nearly every module in the DC, taking into account multiple variables rather than a single measure of UPH. Though the variables were similar in definition from operation to operation, the pickers earned varying time credits for each.

Across the operation, three common variables were used to measure productivity: number of units picked, active locations visited, and shipping cartons filled. Prior to the start of the program, management committed to a rigorous schedule to ensure that all DC workers took part in a one-on-one coaching and feedback session.

The result of the program: The picking operations improved productivity more than 60%.

Customizing Your Metrics

Straightforward UPH measurements may not measure the true productivity within a picking operation because of the variety of factors that affect picking operations, including picking method, product mix, and picking accuracy. For that reason, you need to tailor your metrics to your operation.

Let's start with variables regarding picking methods. Most distribution facilities have a variety of picking methods. This makes a universal picking productivity value difficult within a warehouse. To determine the mix of methods used in your facility, try a “with/to/from” analysis:

• With what are you picking: forklift, guided vehicle, on-foot worker?
• To what are you placing the picked item: carton, pallet, tote?
• From what are you picking: pallet, bin, carousel, static rack, flow rack?

Next, consider the mix of product being picked. In a furniture warehouse, for instance, there are large physical differences among types of furniture. An employee who has to pick sofas and armoires all day will certainly not complete as many picks as an employee picking ottomans and dining chairs. If you use a simple UPH measurement as your productivity indicator, you are creating an environment where associates will cherry-pick their work to ensure desired productivity results. This results in disbelief in the fairness and equity of the program, in addition to animosity toward those workers who seek opportunities to cheat the system.

And of course, you should factor picking accuracy into productivity measurement. When an operation begins undergoing productivity measurement, the workers typically make a point of working faster to meet the standard. Unfortunately this often results in a higher rate of mistakes. You can curtail this behavior by deducting points, in a sense, for picking mistakes, thereby lowering their overall productivity scores. This concept is known as factored performance.

Let's reconsider the performance of Sally, assuming that her accuracy was measured as 94% during her hour of work. To determine her factored performance, we would multiply her performance percentage and her accuracy percentage: 105.0% × 94.0% = 98.7% factored performance.

Rather than being measured by her normal performance, Sally will be judged by her factored performance, which is significantly lower and illustrates the need for an improvement in picking accuracy. Additionally, in performance incentive programs, you can establish minimum accuracy levels to prevent associates from earning bonuses unless they meet and maintain a minimum accuracy rate.

Managing Change

To ensure that your program is effective, you must guarantee that the productivity measurements are not only accurate but also fair and equitable. The best way to prove this is through on-the-floor coaching sessions, which ensure that positive work practices are recognized and negative work practices are identified and avoided.

Generally, four factors affect an employee's level of performance:

• skill — talent and proficiency
• technique — efficiency and effectiveness of methods
• activity — time spent staying on task
• rate — level of effort expended

Only after you've coached your workers so that they can maximize these four factors should you put in place the surrounding reward structures and recognition programs.

Timeliness of feedback is also imperative. Performance data should be posted immediately instead of allowing time between effort and reporting. This allows workers to more easily understand the link between work practices and performance.

Take the art of the “long view” when developing the measurement program. It should be a lasting program, not a one-time project. The standards should be modified only when changes in process, equipment, layout, or systems require it. If you do need to change a standard, do so immediately to avoid losing credibility. The program should include all employees — supervisors and management as well as the pickers themselves. Try to make it entrepreneurial in nature, so that the employees feel as if they are in business for themselves.
Recognize that this is not an engineering or technology solution but rather a shift to a performance-driven culture. Employees will alter their mindsets only if they see the point of the change and agree with it — at least enough to give it a try.

The Formula
for Success

Successful project implementations come down to several important factors:

• multiple, specific productivity variables customized to each operation
• supervisor involvement with employees on the floor
• coaching of associates and feedback in the areas of skill, rate, technique, and methods

A good measurement formula therefore would be:
Objective standards + timely performance feedback + coaching and skills improvement = success in measuring.

Many of the companies heralded for their supply-chain best practices tend to be fairly open with benchmarking and facility tours, which showcase their capabilities. Such best-in-class firms continually look to further streamline, standardize, and simplify.

H.J. Heinz of H.J. Heinz Co., one of the country's leading food condiment companies, has been credited with inventing the public factory tour. One of his founding principles, dating from 1876, stated, “To do a common thing uncommonly well brings success.” His statement captures the very essence of the important subtleties you need to manage when tackling a picking productivity measurement project.

Factoring in Tracking Technology

In distribution facilities that use a warehouse management system (WMS) with radio frequency (RF), data-capture capabilities allow management to establish a detailed production history for every associate. In warehouses with little or no capability for data capture, the level of detail in the automated productivity measurement is minimized.

In such facilities you can nonetheless effectively record time and production manually. Data entry clerks, however, are generally required to complete the sensitive task of summarizing time and productivity data.

Depending on production tracking capabilities, you have several options for calculating productivity:

• use customized spreadsheet and database applications
• capitalize on the WMS functionality of existing enterprise resource planning (ERP) or niche applications
• procure best-of-breed bolt-on applications for labor management
• leverage the warehouse control system (WCS) already managing picking tasks and automation equipment

As a note of caution, our experience tells us that industry buzz phrases such as LMS (labor management system) tend to take on a life of their own on as a sole operating strategy. But the companies heralded as supply-chain “best performers” do not think in terms of IT projects.

Rather, their philosophies start with tough management solutions, not a new technology that promises salvation. To them, technology is only a specific enabler of management solutions. A meaningful solution can exist at some level with little or no technology just as well as with the most modern systems in place; the results are in the implementation.
— EW/JD

Eric Watterson is a project manager for XCD Performance Consulting, an Atlanta-based logistics, supply-chain management, and operations consultancy. Jeremy Davidson is an account executive in the Nashville, TN, office of Fortna, a provider of distribution software, material-handling systems, DC design, and operations-planning services.

What Have You Done for Me Lately?

You no longer have to rely on your workers for that information. Today's sophisticated software can tell you everything you want to know about your staffers' productivity.

When engine and generator maker Briggs & Stratton began to implement a labor management system (LMS) in its DC, David Zuern thought he knew what to expect. The company already had a warehouse management system (WMS) in place, so he assumed he'd find that the DC was operating at pretty close to maximum efficiency. As it turned out, there was a surprise in store for him. "We discovered that the way we were slotting product was very, very haphazard and created a lot of wasted time for the pickers," explains Zuern, who is the company's director of distribution operations. As a result, Briggs & Stratton reslotted its product to boost efficiency.

Discoveries like this are typical for companies that implement an LMS. Labor management systems build on warehousing systems (many WMS vendors have developed LMS modules), but they approach the process from an entirely different perspective. Whereas a WMS manages order and inventory, an LMS tracks the activities of people, based on input from the WMS about tasks that must be performed.

"People have started to max out what they could do with inventory and are turning to the next big project, which is labor," says Peter Schnorbach, senior director of product management, labor and slotting at software specialist Manhattan Associates. He notes that labor costs typically account for close to half the cost of distribution, which makes them an obvious place to look for savings.

When Briggs & Stratton implemented its LMS, RedPrairie's DLx Labor, Zuern saw productivity increases in the 20- to 25-percent range. That's typical, says Greg Aimi of AMR Research. Implementation of a labor management program typically results in a 10- to 30-percent gain in productivity, he reports. That may not sound like much, but for most operations, it actually translates into significant savings. Labor costs can make the difference between profit and loss, especially in low-margin businesses. And payback for these systems can be quick: A study by ARC Advisory Group found that most companies saw a return on their investment in less than a year.

Even those eye-popping savings haven't made the LMS standard equipment in the modern DC, however. Though common in industries like grocery distribution, labor management systems have yet to be widely adopted in many business segments.

No Quick Fix

Popular perception notwithstanding, the benefits of labor management don't come from just installing a piece of software—far from it. "If you think labor management is software you can take out of a box and plug in, you won't see the full results," Zuern warns. "This is not a quick-fix, out of- the-box solution." Troy VanWormer, a founding partner of XCD Performance Consulting in Rancho Santa Margarita, Calif., agrees. "One of the reasons a lot of these programs fail is [that people] think it's a systems or technology project. It's not. It's a people project."

Rather, the technology should be seen as an enabler for a complete labor management program, a significant undertaking requiring thousands of observations of every task performed in the DC. Those observations provide the basis for the development of engineered labor standards—best practices for each job. The idea is to determine how long it should take to do each specific task and then use the software to compare workers' performance against those standards.

In the end, says Zuern,"labor management programs are about working smarter, not harder—getting more done in the same amount of time." Once Briggs & Stratton embarked on the observation phase, he says, it became obvious that "the barriers to productivity were more prevalent than we had imagined. Discovering this forced us to look at and re-engineer processes—and this was a good thing."

Labor Leaders

These observations must be done for each facility where the LMS is being implemented, even if each handles the same products as a sister site. "That's the only way to do it, because each DC has different characteristics," explains Lillian Warrington, engineered labor standards project manager for food-service distributor Perlman- Rocque, which implemented LMS in all four of its warehouses in 2005. "They use different equipment. Some handle different product." And each warehouse's layout is a little different, too.

Observing and quantifying tasks can take several months, but VanWormer explains why it's necessary: "You can base your data on historical averages, but if you're historically bad, the number is not very high." You could also set standards based on what seems like a reasonable expectation: "You pick 120 units per hour, so we think you can do 150." But that standard is subjective. And a blanket units-per-hour measurement is not accurate, either. If one worker is picking items that weigh 60 pounds each while another worker picks items that weigh half a pound, they obviously won't be able to pick the same number of units per hour.

Chris Smith, director of process improvement for pharmaceutical distributor McKesson, explains how engineered standards are helping his company. "Under the old productivity metric, you only had one overall score on how someone did. You had no visibility if they were doing well in one area and not in another." Now, he says, you can see if they need help in a particular area, and that has helped employees succeed at meeting labor standards—as well as helping the program succeed as a whole.

A Shift in Culture

Perhaps the biggest change that resulted from Briggs & Stratton's labor management program was the culture change within the DC. In the old days, says Zuern, supervisors had no way of knowing whether people just looked busy or were actually being productive. "But now," he says, "employees have to meet the [productivity] standard every day. They're responsible for their own performance. That means employees come to us very quickly with problems that are getting in the way of productivity."

In fact,he says, employees brought to light a number of procedural holdups that had long gone unreported. "Once we started to hear about them, we realized that we really did need to fix these issues so we could be more productive, and that really got us rolling," he says. "Labor management has changed the culture. ... Supervisors and associates are working together to maximize productivity, rather than against each other. ... Supervisors are now problem solvers more than enforcers. ... So what we ended up with was a much more productive, smarter-working workforce, higher throughput, and a culture of process improvement."

Realistic standards are crucial to worker acceptance of the project—and to its overall success, Zuern says. He reports that Briggs & Stratton encountered very little resistance when it went to expand the LMS beyond its pick, pack and ship operation to its kitting and packaging operation because workers could see that the standards were achievable.

It's more than a matter of employee morale, however. Companies that set unreasonably high standards in hopes of promoting a little workplace hustle risk compromising both accuracy and safety. Ultimately, the costs associated with quality problems and accidents could end up erasing any savings resulting from productivity gains.

Of course, some positions lend themselves more readily to the development of credible standards than others. "Those positions that have more variability in the tasks—shipping and customer return—provided more challenge," says Chris Smith. "Our philosophy was never to force a standard where it didn't make sense." The company chose to count tasks that couldn't be measured easily as "indirect time," which was weighted differently than direct time but nonetheless recorded. "The system counts the indirect time," he says, "so we can see what they're doing."

Selling
the System

Aside from the amount of work that goes into engineering the labor standards, one of the main challenges of implementing an LMS can be handling the transition. "One of my biggest messages is 'Don't underestimate the change management aspect of the program,'" says McKesson's Chris Smith. "Make sure there is strong support from senior executives. Make sure there is tight alignment with field operators and with human resources." At McKesson, field operators were shown P&L statements indicating the savings that could be achieved from the LMS program, which helped secure their cooperation.

Managing change also entails making sure pickers and packers know why the system is being implemented: "not just to lower costs, but to remain competitive," explains Zuern. And Warrington says that at Perlman-Rocque, where three out of the four DCs are unionized, "we not only solicited the union's involvement, but had them work with us on it, which I think helped the process tremendously. We had an open-book project. Anything they wanted to know was available to them." Workers even did some work-process observations. "Addressing their concerns was paramount. I think that made the project a success."

Ongoing external changes are a factor, too. "Our business is dynamic, so our engineers are supporting ongoing training, and our DCs go through process improvements—in part to improve productivity, but also to accommodate legislative changes," notes Smith of McKesson. The company has a dedicated human resources person to manage the personnel aspects of the program, including an incentive program that's based on the productivity standards.

Once standards have been established, there will be a transitional period as employees learn to work to the standard. At Briggs & Stratton, employees were given 60 days after the system went live to gradually work up to full productivity.

Similarly, there's a learning curve in implementation as a whole. When McKesson started implementing a WMS a little over two years ago in two pilot DCs, it took six months to get each DC up and running. "Today our rollout is three months," says Smith. The company has implemented LMS in 26 of its 31 DCs to date, and the remaining five are expected to be online by April 2007.

Worth the Effort

Now that they know what's involved, would the managers who've been through an LMS implementation do it again? "There are a lot of benefits to this, but implementing an LMS is [a] very detailed [process] and it's very hard work," says Zuern of Briggs & Stratton. Still, he characterizes it as a worthwhile effort. "Today, we can truly operate with fewer people, and the greater throughput is evident in the DC." Productivity increased roughly 20 to 25 percent across both operations. The company was able to reduce pick, pack and ship headcount by about 18 percent right away. A few employees left because they didn't want to work under the new standards, he says, but most found it easy to meet the standard after learning how to eliminate the non-valueadded activities in their daily jobs.

In the end, he believes, the program has been a positive experience. "It took a lot of pressure off everybody. Managing employee expectations is a lot easier when everyone knows exactly what those expectations are and feedback is readily available." And the benefits didn't end there, he says. "The culture change is the big improvement—it lets us focus on the things that matter most to our business and our customers."

At McKesson, Chris Smith has no trouble ticking off a list of benefits he's seen from the LMS: "The enhanced productivity within our DCs, reduction of overtime, service-level improvements, the visibility regarding performance." This visibility allows supervisors to continually improve their coaching and feedback to associates, and thus to keep improving performance over time.

Perlman-Rocque also reports good results from its LMS installation. "We saw improvements in productivity and reduction of cost of up to 20-plus percent per DC," says Warrington. There was no workforce reduction, but the company did reduce overtime. And, she says, it has made the job of the front-line supervisors much easier. "They're probably the happiest folks here, because now they know how long work should take. They can manage better. They're less under the gun because there's less ambiguity." Though it required a lot of time and effort to make sure the labor standards were fair, it was worth it, she says. "I'm thrilled with the results we accomplished over the last year and a half."