CNET también está disponible en español.

Ir a español

Don't show this again

Tech Industry

What an e-tailer learned from Dell

FreshDirect has a made-to-order philosophy that eliminates the middleman and creates a more efficient supply chain. Heard that line before? This time it may actually work.

Despite the carnage among the same-day home-delivery e-tailers, a new online grocer recently began delivery operations in New York City. Like its deceased predecessors Webvan, Kozmo and Urbanfetch, is a pure-play start-up pursuing a dream of rapid home delivery of food products. But unlike its predecessors, FreshDirect has a well-thought-out operating strategy--and it's a model that might just work.

FreshDirect doesn't look to other retailers, and certainly not to other e-tailers, for inspiration. As FreshDirect's co-founder Jason Ackerman recently told Fortune magazine: "The only reason we chose the Internet was that it helped us reach people at a lower transaction cost. It allows us to do for food what (Dell Computer CEO) Michael Dell did for computers." Like the Dell, FreshDirect employs a make-to-order philosophy to eliminate the middleman and create a more efficient supply chain. The Internet offers a critical tool to achieve the operating efficiencies that FreshDirect needs to profitably serve its targeted customers.

FreshDirect doesn't look to other retailers, and certainly not to other e-tailers, for inspiration.
FreshDirect's operating strategy also turns the traditional grocery business model, which offers wide variety to a diverse consumer population linked largely by geographic proximity, on its head. FreshDirect offers a selection of basic but still high-quality staples and sells specialty items only when they will clearly pay off. The question is whether the consumer appetite will match the volume and offerings of food scaled to deliver.

The inspiration behind FreshDirect dates to 1998, during the heyday of dot-com mania when erstwhile e-tailer Webvan was also making the venture-capital rounds for its June 1999 launch. Unlike Webvan, however, the founders of FreshDirect never sought to become one of the few companies to "earn the right to cross into a person's home," the goal espoused by Webvan's former chief executive, George Shaheen.

In fact, Louis Border, the creator and founder of Webvan, never wanted to be in the grocery business, but saw it as a pathway to achieve dominance in the "last mile" to the consumer's home. Groceries, as an ongoing "replenishment" purchase, would provide the base load for delivering a plethora of products bought online. In early business plans, Border described a vision of carrying a million stock-keeping units.

FreshDirect's focus, by contrast, is not the service, but the product. "It's not about convenience. Online shopping and home delivery can be inconvenient," Ackerman told strategy+business. "We focus on fresh products and offer higher quality at a lower cost by eliminating waste throughout the grocery supply chain in each individual processing step."

A different model
Rather than reproduce a traditional supermarket online, the FreshDirect management team has set out to create a totally new model. About 50 percent of sales in a typical grocery store comes from such packaged goods as cereal, soda, and laundry detergent; the other 50 percent comes from fresh products--meat, seafood, produce, deli products and baked goods.

The typical grocery store carries about 25,000 different packaged goods items and roughly 2,200 perishable products. In contrast, FreshDirect offers 5,000 perishable products but only 3,000 choices in packaged goods. As a result, it generates a revenue pattern dramatically different from that of a traditional grocer: an expected 75 percent in perishables and only 25 percent packaged goods. Given the higher margins for perishable goods, this change in the mix of goods alone probably generates an 8 percentage point improvement in gross margins--even without considering the efficiency gains that FreshDirect hopes to achieve.

To make this model work, FreshDirect built a 300,000-square-foot facility in Long Island City, which it calls a processing center rather than a distribution center. This distinction underscores the company's emphasis on good-quality fresh foods rather than the delivery process. FreshDirect offers the New Economy equivalent of the local fresh market, but it can bring the fresh product to your door because central processing lets it create scale efficiencies unachievable by the local market.

Rather than reproduce a traditional supermarket online, the FreshDirect management team has set out to create a totally new model.
Thanks to its made-to-order approach and scaled, vertical integration, FreshDirect shortens the supply chain to produce a fresher product at a lower cost.

Consider the seafood operation. FreshDirect's representatives place initial orders at the docks in lower Manhattan as the catch arrives during the day and into the evening. At midnight, FreshDirect stops taking consumer orders for the following day and provides an exact order quantity to the seafood buyers.

The prescribed quantities arrive at the Long Island City processing facility around 3 a.m., to be cut according to customer orders by early to mid-morning. Customer deliveries begin at 4 p.m. the same day, resulting in a "dock-to-door" time that is often less than 24 hours. On average, FreshDirect's seafood department has about one day's worth of inventory, compared with the seafood counter at a well-run grocery store, which has seven to nine days' worth of inventory. Michael Dell might even be proud.

Not only does vertical integration shorten the supply chain to improve product quality, it also cuts out the middleman, allowing FreshDirect to garner higher margins while still offering a broad selection of fresh food, greater customization, and low prices.

Dell also eliminated the middlemen of the computer industry supply chain--the value-added resellers and retailers--by offering its own brand assembled to order from the same quality of components used by the established brands. FreshDirect applies this model to coffee beans: The grocer offers 55 standard varieties of coffee from 22 different coffee beans, which are roasted and ground to order. But if those options don't fit the consumer's needs, FreshDirect will prepare a customized blend by combining the 22 beans in any proportion desired. Although roasting and blending add further costs, the raw beans can be procured for as little as 70 cents per pound, leaving room for both low consumer prices and attractive margins.

Limiting the brands
Despite its clear focus on perishables, FreshDirect does not completely ignore the standard grocery items. FreshDirect is a small operation, especially compared with Kroger, the leading pure grocery store chain in the United States at $50 billion in annual revenue, or even Albertsons or Safeway, the No. 2 and 3 chains, each with revenues well over $30 billion. So the start-up faces a substantial disadvantage when it negotiates pricing with giant consumer goods companies such as Unilever, Coca-Cola and Procter & Gamble.

FreshDirect compensates for its lack of conventional negotiating leverage by letting major consumer goods companies compete for exclusive positions on its Web site--a marketing advantage these companies can rarely achieve in traditional grocery stores. For example, FreshDirect offers six Coca-Cola flavors in three sizes for a total of 18 different SKUs. But not a single bottle or can of Pepsi is for sale. Similarly, a search for Keebler on the FreshDirect Web site turns up 28 snack products, whereas a search for Nabisco produces no hits.

In general, FreshDirect seeks to maximize its volume per item by limiting the number of SKUs it carries. If you are loyal to a specific brand of powdered detergent, then you might be disappointed in what FreshDirect can deliver: It offers only two brands, Tide and Ivory Snow. Tide does come in three different types but only in the 78-ounce size. The Ivory Snow comes only in one type and a 26-ounce box.

Webvan, Kozmo, and Urbanfetch offered free delivery; FreshDirect charges $3.95 for delivery and requires a $40 minimum order.
FreshDirect offers a wider selection of cereal--a total of 73 choices, according to our search late last year. But to keep the SKU count down, only five of the 73 come in more than one size. The high-volume brands--Cheerios, Kellogg's Corn Flakes, Kellogg's Raisin Bran, Special K and Total--each come in two sizes. Compare that with the 125-plus brands in a typical suburban grocery store, 20 or more of which usually come in two or three sizes. Adding in private-label store brands, a typical store carries more than 200 different cereal SKUs.

FreshDirect does offer a large selection of what it calls specialty items--such as condiments, cookies, crackers, spices, oils and vinegars--from which it can capture higher margins. A "specialty" tab on the home page seeks to attract its most discerning customers to those products, instead of similar, but less expensive, items sold under the more mundane "grocery" tab.

For example, FreshDirect carries 26 different SKUs of mustard. For those consumers seeking a value brand, French's Classic Yellow in the squeeze bottle is sold at a price below that at a traditional grocery store. But the company also offers 25 other mustards, including some exotic imports at prices as much as 10 times higher than the value brand's.

Will customers accept FreshDirect's less extensive range of packaged goods to enjoy the fresh product advantages that it offers? Ackerman believes so.

"A typical wholesale grocery store like Costco offers a mere 4,000 SKUs of the most basic products like soda, diapers, and toilet paper in bulk quantities at unbeatable prices," he says. "The dramatic growth of the warehouse format offers ample evidence that customers will accept a limited product offering of basic commodities to get lower prices. In fact, our ideal customer would be someone who buys bulk staples once a month from Costco and buys everything else from us once a week."

No promised land
FreshDirect's management holds no illusions about a promised land in the "last mile," but sees the last mile as a necessary evil that supports the make-to-order model.

FreshDirect has learned from its predecessors' mistakes. Webvan, Kozmo, and Urbanfetch offered free delivery; FreshDirect charges $3.95 for delivery and also requires a $40 minimum order. This delivery charge is competitive with the $2.50 to $5 charge of a typical Manhattan grocery store. Such pricing generally covers the delivery cost and meets consumer expectations, according to the startup's market research. And, as was the case with Webvan and Urbanfetch, FreshDirect's drivers don't accept tips.

To control the cost of the "necessary evil" of home delivery, FreshDirect doesn't offer same-day delivery--only next-day delivery. In addition, it delivers only between 4 p.m. and midnight and promises delivery within a two-hour window. In contrast, to emphasize the convenience of delivery, Webvan offered slots from 6 a.m. to 10 p.m. and initially promised delivery within a 30-minute window. Such precision demanded sophisticated routing software, but still yielded regular redeliveries. Although it is still early in its launch of services, FreshDirect claims redeliveries of less than one-fourth of 1 percent.

The location of its processing facility is the most critical factor in FreshDirect's cost-efficient delivery model.
The location of its processing facility is the most critical factor in FreshDirect's cost-efficient delivery model. Long Island City, in the New York City borough of Queens, is less than a quarter-mile from the Midtown tunnel connecting Long Island and Manhattan. This gives FreshDirect easy access to the greatest population density in the United States: Roughly 4 million people can be found within a 10-mile radius of its processing center. Thanks to this population density, FreshDirect averages nine to 10 deliveries per hour with a two-person crew--more than triple the three deliveries per hour that Webvan struggled to achieve with its single-driver model.

Like Dell at its inception, FreshDirect offers a fundamentally different consumer offering. Dell, as an unknown with limited funding for marketing, focused on execution to build its reputation, delivering a high-quality product through a make-to-order process using the same standard components its competitors used, but without the middleman. Dell's edge in speed created a cost advantage that ultimately redefined the PC industry. Its competitors, saddled with longstanding channel relationships and made-to-stock manufacturing models, have been struggling for years to make the transition to the Dell model, but they still remain short of the system Dell created from scratch.

FreshDirect hopes to follow a similar path by focusing on execution in its target New York City market. Its system should offer an immediate competitive advantage given the city's unique delivery density and especially fragmented competition. Beyond New York City, however, FreshDirect would face different operating economics and formidable competitors, such as Safeway, which has an Internet delivery partnership with the U.K.'s, and Peapod, which is backed by the Netherlands' Royal Ahold chain. These global supermarket industry giants may prove more adept at transforming their business models than the leaders in the PC industry were during Dell's early days.

Ackerman says the most poignant lesson to remember from Webvan's demise is to avoid excessive expansion: "This is a very complex business, and the customer demands perfection every time we fill an order. Webvan's rapid expansion was matter how good the executive team."

FreshDirect's goal of $500 million in annual revenue for the Long Island facility seems plausible. Its current financial plans forecast positive cash flow from operations by the middle of 2003. Only when this goal is met will management contemplate expansion, the company says. FreshDirect's initial priority is to open additional processing facilities (up to three) to serve the same New York City market, with a goal of $2 billion in sales and a 5 percent market share within five years.

Although FreshDirect is seeking a small percentage of a large market, it aspires to volume that is nevertheless well beyond its predecessor's. Webvan's original $30 million distribution center in Oakland, Calif., was designed to support 8,000 orders per day and generate $300 million in revenue. Instead, it peaked at a mere 3,200 orders per day and never broke even.

Fortunately, FreshDirect has avoided the ills of unbridled spending that plagued most start-ups during the height of dot-com mania. Its offices are austere and its staff overhead minimal. Because of the high costs of a prime New York City location coupled with greater vertical integration, FreshDirect has consumed $100 million so far, a large sum, but far below the $1 billion that Webvan blew through during its short existence.

In contrast to Webvan's high-cost marketing effort, FreshDirect is leveraging billboards, public relations, and word-of-mouth--a so-called guerilla marketing effort. Financed by private investors, including significant funding from the founders themselves, FreshDirect will certainly not spend money like Webvan did. Patient capital and tight purse strings may offer a better formula for survival than a heady capital market and get-quick-rich entrepreneurs.

To read more articles like this one, visit

Copyright © 2003 Booz Allen Hamilton Inc.

Reprinted with permission from strategy+business, a quarterly management magazine published by Booz Allen Hamilton.