Warehouse Management & DC Optimization: Measuring What Matters
Selecting the best metrics at your company's distribution centers requires the right tools, careful study, and constant collaboration.
More to the Story:
Recently, a company that outsources product distribution to Orbit Logistics in Ashland, Va., ran into a problem, but didn't know it. The company sold its products through multiple channels, all pulling from the same inventory housed in one of Orbit's three distribution centers—"effectively cannibalizing their own inventory," says Wendy Hudson, vice president for client solutions for Orbit, which provides e-commerce and retail fulfillment through its three warehouses in Virginia. The company could accept orders on different channels at the same time and surpass the available inventory without realizing it. With insufficient inventory, Orbit, which assembled the products on site, needed to rush to create more. The company did not realize the magnitude of their problem, but fortunately, Orbit did.
"They needed help to see that they had to have their inventory in place before committing to an order, so they wouldn't come up short and potentially run into penalties," Hudson says. "They also had to realize they incurred real cost when we had to remake products for them."
Orbit used measurable data to convince the client how important it was to address the issue. For another client, Orbit used metrics to demonstrate the steep costs they accrued for absorbing an unusually high number of returns. The news was revelatory for the company, which had assumed the impact was minimal. "With quantifiable data, you can identify issues and work on viable solutions," Hudson says.
In today's complex distribution center environment, the right data can mean everything. The rise of e-commerce and the need for fulfillment centers that can warehouse a large variety of products and ship them quickly to customers on demand has dramatically altered the landscape. The heightened emphasis on speed and a diversity of products can put a strain on a distribution center while demanding innovative and ambitious steps to keep up.
"Many companies today are faced with the challenge of how to position inventory to make it readily accessible to a large percentage of customers with varying delivery demands—all without unnecessarily driving up costs by holding too much inventory or spreading it too thin," says Jose V. Gonzalez, head of CLS product management, North America for APL Logistics, a Scottsdale, Ariz.-based company that provides supply chain services worldwide.
"These challenges are having a profound impact on the number and types of distribution centers being built—and increasing demand for larger facilities that allow room for the labor-intensive operations and greater number of workers that e-commerce requires," he adds. "They also are driving the more rapid introduction of technology solutions into warehousing operations."
Distribution center performance has become more important than ever for retailers. If customers expect to order any item they desire, anywhere at any time—and then receive it quickly—the work at the DC plays a critical role in making that expectation a reality.
For that reason, measuring the work done at distribution centers also takes on a particular prominence. Both the retailers that use distribution centers and the logistics professionals who manage them need accurate, readily available data that provides a clear picture of how successfully they are performing their mission.
Data, Data, Data
The rise of increasingly sophisticated technology and the abundance of data those advances make available have opened doors to a granular study of performance that would have been impossible until recently.
"Few logistics professionals will claim that new technology isn't a critical part of increasing distribution center efficiency," Gonzalez says. "In most cases, the more information you're able to collect, transfer, and leverage, the more successful your operation will be."
Not too long ago, Orbit relied on a paper-based picking system with manual inventory entry. That slowed data collection. Today, Orbit uses a digital system with assigned barcodes and license plates.
"The system lets us check things much faster and more accurately," Hudson says. "Data can be available on demand. Even one year ago, our customers didn't have that on-demand expectation. They were fine if their inventory was batched at the end of the day or the end of the week. Today, they want to know what inventory they have right now. Technology has increased their expectation for data availability."
The migration to digital remains an ongoing process in the distribution center sphere, says Matt Miller, CEO of MobileDemand, a Hiawatha, Ill.-based maker of rugged tablet PC systems for business and enterprise operations. He estimates between 10 and 20 percent of MobileDemand's new rugged tablet customers have been using pencil and paper tactics for inventory—companies that are automated on the back end but not on the front.
The effect of making the transition from a paper-based process can be dramatic. For instance, Rimports, a Provo, Utah-based producer of scented wax, wax warmers, and essential oils and diffusers, was using clipboards and handheld scanners to track inventory in the warehouse. The system's inconvenience led employees to scan as few labels as possible, hindering inventory tracking and movement.
MobileDemand worked with Rimports to deploy a Microsoft Surface Pro 4 in a MobileDemand rugged case, mounted on a forklift. The upgrade improved and quickened data capture, allowing for better measurements.
MobileDemand is unusual in that it offers inexpensive rugged tablets, allowing smaller operations to automate data collection. "Technology advances and the consumerization of devices are driving down hardware costs," Miller says.
Despite the vast new wealth of data, one of the biggest mistakes companies make is measuring and monitoring too many attributes at once. When APL introduced labor management systems into its distribution centers, it monitored 44 measurements on its facility scorecard. When it found itself overextended, APL eventually cut back to 28 measurements.
"Ultimately, there's only so much time in a day—and only so much energy companies can devote to evaluating and analyzing performance," Gonzalez says. "Even though it's tempting to measure every aspect of a facility's performance, it's important to remember that trying to measure an overly large number of criteria only results in creating more work for people. And it makes it difficult for those who are reviewing the information to focus on what's important."
Finding Your KPIs
The measurements Orbit's customers most closely monitor can vary, Hudson notes, but some overarching points of emphasis typically apply. In all cases, a clear view of inventory status is crucial. In most, so is speed and accuracy.
"Companies are usually focused on trying to get the most volume in and out of the business with the fewest errors possible," Hudson says. "How fast can inventory move and how accurately? Both of those questions are directly related to how profitable you will be."
APL Logistics developed a scorecard for its facilities with key performance indicators (KPIs) that include cycle time, employee turnover rate, on-time inbound deliveries, inventory accuracy, and on-time shipping, among others. Similarly, DHL Supply Chain, which provides customized supply chain solutions across industry sectors, looks at numerous operating KPIs for insight into operational performance, says Will Heywood, the company's vice president for regional resource planning and management and operations development.
"At the headline level, we focus on a few numbers that together provide a reliable view to how well a site is performing," he says. "These measurements include direct vs. indirect labor hours, labor cost per unit, inventory location accuracy, safety, associate turnover, and overtime."
No matter which KPIs a logistics provider uses, the companies that employ them will also target their own. Often, those areas of focus overlap with those that other companies emphasize, but some vary sharply. For example, speed may not be a chief concern for some companies.
"For some companies, the data captured is more important," Hudson says. "For instance, if they're tracking products for warranty information, they want to capture the serial number. They might need to gather a lot more information in receiving and it's worth it for them to focus on being more accurate and ready for outbound use. It depends on each market."
Nicole DeHoratius, adjunct professor of operations management at the University of Chicago's Booth School of Business, says her research with collaborators identified two key areas in distribution centers that people believe are important but not adequately measured: consistency—the ability to fulfill orders repeatedly on time and in full, and recovery—the ability to bounce back after a service disruption.
Retailers can improve profitability by determining orders via consistency and recovery measurements rather than "common metrics such as in-stock," argue DeHoratius and her fellow researchers.
"People talk about consistency and recovery but haven't found good ways to measure them," she says. "We argue that consistency and recovery should be included on scorecards. Both measurements clearly matter to supply chain performance, but they have been overlooked because they are hard to measure."
Seeking a Collaborative Approach
Key components of any relationship between a supplier and a distribution center operation are open communication and a collaborative approach. Companies should take ownership of the measurement being done in the distribution centers that manage their products and ensure that their logistics providers are collecting the most useful data possible.
Orbit, for example, works with its clients to target the measurements that fit them best. "Each customer may have either conflicting or complementary KPIs," Hudson says. "One may want more data collected or more sorting in the receiving area than others. It's about finding a balance between overall KPI and individual customer preferences."
New technology makes it much easier for companies to develop a partnership with those that are managing their inventory in distribution centers—and vice versa.
"Service providers need to create transparency and visibility for customers," Hudson says. "We have customers who have never set foot in the warehouse. Technology has allowed us to create a connection with them so they see us as a key part of their business structure."
APL aims for frequent communications with clients via the channels and methods they prefer, which could be regular reports, routine scheduled meetings, or even an on-site client representative. The key is for companies to understand they have the wherewithal to play a part in the process—they should not simply cede decision-making about measurement to their logistics providers.
"To make a 3PL operation all that it can be, a company must be clear about what it expects from its provider and that location down to the nth degree—and, if it doesn't, it's the provider's job to extract that specificity," Gonzalez says. "Ideally, the parties should discuss these expectations upfront and prior to signing a contract because that's the best time to align goals. But it's never too late for a company to spell out what it wants and how it wants it."
Strong DC collaborations remain the exception rather than the rule. DeHoratius was part of a research team that studied fulfillment errors at a major retailer's DC. The research found that the majority of fulfillment errors were due to documentation, barcoding, and retail ticketing mistakes. Some errors originated with suppliers, who then frequently received financial penalties in the form of chargeback deductions, which cut into their revenue.
"Better collaboration among supply chain partners, including buyers, vendors, and warehouse operators, could have limited those errors," DeHoratius says. Ultimately, the study notes, "fulfillment errors reduce the efficiency of the supply chain as a whole and impair implementation of modern retail tools and techniques like automation, pack-by-store, and omni-channel."
To solve persistent fulfillment errors, it is important to take a total cost perspective and consider all the different ways inefficiencies can arise.
"Supply chain partners who focus on collaboration also tend to look at the total cost of the supply chain," DeHoratius says. "They don't look just at the price to acquire or deliver a product, but at the total cost to deliver a product from end to end. This perspective is rare, but much needed."
Getting it Right
In the supply chain, data integrity issues and fulfillment errors can lead to inventory management problems and, consequently, reduced demand.
"One distribution center challenge is that there's often a discrepancy between the inventory on the shelf and the inventory recorded in the system," DeHoratius says. As a result, some critical questions loom: Do the companies know how big that discrepancy is? Do they have methods in place to fix those discrepancies? How can they prevent them?
DeHoratius and her team conducted an experiment in a virtual lab mimicking the experience of pickers. In one scenario, pickers chose between two products with easily distinctive packaging. In a second scenario, the packaging was similar. Researchers observed a productivity loss of approximately 23 percent in the case without distinctive packaging.
Technology can help reduce picking errors, for example by providing pickers with detailed images of the products they are selecting. Alternatively, companies can be more thoughtful about offering pickers visual cues through packaging that differentiates one item from another.
"Most distribution centers , however, still have legacy systems in place that don't help pickers identify when they make a mistake at the point of occurrence," DeHoratius notes.
An Ongoing Experiment
It is easy to fall into the trap of thinking that you are measuring what you need to measure, especially if that data provides you with good news. However, it is important to test your KPIs—not only the basic measurements, but also the methods behind those measurements.
"There are a lot of ways to prove you're right and a lot of ways to prove you're wrong," says Hudson.
Companies that challenge their assumptions and continuously search for the best way to measure performance will gain an advantage. It is rarely as simple as identifying a measurement and sticking with it.
"There are many different ways to measure KPIs, and you tend to have to try them on for a while," Hudson says. "Look at the KPIs that you are using and ask: Do they drive the right behaviors in my workforce and my bottom line? If they don't, then it might be time to try some different ones. The measurement process should constantly evolve."