Anyone in the supply chain world who's attended a conference or leafed through a supply chain journal this year is aware that the consumer goods industry is obsessed with "omnichannel fulfillment." The concept of omnichannel fulfillment encompasses many things, but at the top of the list is the ability for a retailer to use store inventory to fill e-commerce orders in certain situations.
I don't recall the industry being this focused on a single idea since the days of the Wal-Mart-driven radio frequency identification (RFID) mandates in the mid-2000s. This current trend is, of course, different from the RFID craze in a key way: RFID in its initial iterations was a technology without a business case, while for many retailers, omnichannel fulfillment initially was a business concept that lacked a supporting technology set.
The good news, however, is the emergence of two key technologies that are now being used by a number of pioneering retailers with success. The first, distributed order management (DOM) software, handles the complex task of determining which orders to fill from distribution center (DC) inventory versus store inventory. The second, a modified version of warehouse management software (WMS) called "in-store WMS," enables the execution of those orders, which includes picking, packing, and shipping from the store. Figure 1 outlines the capabilities of the two solutions.
Distributed order management
DOM technology itself is not new. It was developed at least a decade ago and, for a long time, was primarily used as a way to allocate in-transit inventory to customer orders before it actually arrived at the warehouse. This technology has been very successfully adapted to the task of deciding when to use store inventory to fill e-commerce orders. Manhattan Associates, one of the early pioneers of DOM, has reported that numerous large apparel retailers have rolled it out to support their omnichannel strategies. Most of these retailers declined to be mentioned in this article, as they view their use of distributed order management as a competitive differentiator, but having seen the list, I can attest that it's an impressive one. One retailer that has gone public with DOM deployment is Lilly Pulitzer, which has used that software to expand its market reach and drive online traffic to its stores.
Another early pioneer of DOM technology was Yantra, first acquired by Sterling Commerce in 2004 and then by IBM in 2010. Although IBM has not been visibly promoting DOM recently, a number of high-profile retailers use the application. One retailer, Cabela's, has been using the solution since at least 2007.
There also are examples of companies that have achieved distributed order management capabilities without using an "off-the-shelf" application. For example, Stage Stores, parent company of Bealls and Peebles department stores, deployed Oracle in 2003 to perform a number of supply chain functions. The company has configured Oracle to help make decisions about when to fill e-commerce orders from the DC versus using store employees and inventory.
Gough Grubbs, senior vice president of distribution and logistics for Stage Stores, stressed that the company's omnichannel strategy is shifting rapidly. "While there is a high level of satisfaction with the fact that our systems provide us a choice in whether to fulfill orders from the stores or the DC, the preference is changing as our online business grows and profitability track records provide better direction," he said in an interview. "Contrary to a recent article publicizing another retailer's shift to increase store fulfillment of online orders, Stage Stores is shifting more toward DC fulfillment. We don't believe there is a common one-size-fits-all solution across retailers. The answer varies by retailer based on store size, depth of product, and location."
Sears has also been leading the charge with a robust technology set to support its omnichannel distribution strategy. The company uses an internally developed system to perform distributed order management. It has actually developed one of the most advanced sets of rules that I have seen for omnichannel retailing.
Most omnichannel retailers favor using either DC- or store-based inventory and using distributed order management software to manage the exceptions to those rules. For example, Cabela's first seeks to use distribution center inventory and only goes to the stores as a last resort if the DC is out of a product. Stage Stores currently chooses to use store inventory whenever possible and only fulfills an e-commerce order from the DC if there is no other choice. But, as Gough Grubbs noted earlier, this strategy is shifting more toward the DC. Unlike these retailers, Sears uses a more nuanced approach and calculates the lowest-cost fulfillment path on an order-by-order basis to determine how to best source that particular order. There is no preference for either the DC or the stores—simply a preference for the most efficient fulfillment method.
For the time being, Manhattan Associates seems to be the dominant player in DOM technology—the only major player that is actively promoting a (relatively) mature DOM platform. However, Manhattan can expect to have some company soon. JDA (formerly RedPrairie) reports that it will roll out an integrated DOM system with interfaces to WMS in the fall. It wouldn't be surprising to see some mid-tier software providers roll out DOM platforms soon as well, providing a less costly alternative aimed at mid-market customers.
Once a decision has been made to fill an order from a retail store, it becomes critical for retail workers to be able to efficiently and accurately pick, pack, and ship it. Many retailers are deploying modified versions of warehouse management software in their stores to make this possible.
Sears is an excellent example. The company has been quietly building out an omnichannel network leveraging store inventory and can now serve 81 percent of the U.S. population via one-day ground delivery service. As part of this strategy, Sears recently began a pilot program using Highjump's WMS as the execution engine for picking these orders. The company realized that the average retail store worker represents a different demographic than the average warehouse worker: probably younger, less experienced with the concept of picking orders, and more familiar with a different generation of technology. Sears recognized these differences and chose to deploy the in-store Highjump WMS on iPads, using ring-style bar-code scanners. "The technology, which uses touchscreen user interfaces, is familiar to the average store worker, which reduces training time and improves pick speed," Jeff Starecheski, vice president of logistics services at Sears Holdings, told me.
The store planogram is loaded into the WMS, and workers are directed to pick orders by department, using a "cluster pick" methodology more often seen in a warehouse than in a retail store. During the most recent holiday season, Sears was able to process hundreds of orders per day from the store network and filled greater than 99 percent of those using two-day ground service.
Other software vendors have reported successfully adapting their WMS for retail store use. Manhattan Associates released an in-store inventory and fulfillment system module two years ago, drawing heavily on its WMS heritage. The company now reports more than 4,000 store locations currently using the software, which features touchscreen interfaces tailored to younger workers.
Retailers that are developing an omnichannel fulfillment strategy have no shortage of technology solutions to provide decision-support and execution capabilities. The coming year will likely see additional software vendors enter the market with offerings, adapting their WMS systems to support store fulfillment as well as developing distributed order management capabilities, thus allowing retailers to take flexibility and service to a new level for consumers.