Can You Sell Groceries Like Books?

Louis Borders' online supermarket may be the most innovative E-commerce venture to date. If Webvan succeeds, the rewards will be huge

Two years ago, Louis H. Borders was opening a FedEx package that had arrived at his Silicon Valley home. It contained some hard-to-find Japanese spices and specialty foods that the founder of bookseller Borders Group Inc. had ordered from a catalog. That's when it hit him: Retail through the Internet would never become really big unless someone could discover a more efficient and cheaper way to get the products ordered to people's doorsteps. "I kept thinking I would need a loading dock outside my house to use the Internet to buy things," says the 50-year-old Borders, who left the bookstore chain in 1992 to start his own investment firm. "And on top of that, I'd have to pay $10 per package."

Since that prospect would be no more palatable to the average Jane than it was to Borders, he decided to devise something that would be. The result is Webvan Group Inc., a Foster City (Calif.) online grocery superstore that is arguably the most ambitious E-commerce initiative to date.

STODGY RIVALS. It's not just because Webvan is aiming to shake up the stodgy, highly fragmented grocery business. And it's not just that Webvan has raised $122 million so far from such bigwigs as CBS, Yahoo!, Softbank, and LVMH--an enormous sum given that the median amount of venture capital raised by upstarts is $20 million, according to VentureOne Corp. What makes Webvan's ambition such an attention-grabber is that the company intends to overcome one of the most expensive operating challenges imaginable for an Internet company: local delivery of high-quality dry and perishable goods to consumers' homes, at competitive prices, within a 30-minute delivery window chosen by the customer. "We are building the last mile to the consumer," says Borders. "It's a huge logistical problem."

No argument there. If Borders can pull it off, Webvan may well have nailed the as-yet-unrealized dream of the Internet, which is to streamline cumbersome, age-old business practices to bring consumers the ultimate in convenience. "If Webvan cracks this nut, it will reap huge rewards," says analyst Evelyn Black Dykema of Forrester Research Inc. "But can it do it?"

Clearly, scores of frustrated shoppers hope so. Since Webvan's June 2 launch in the San Francisco Bay area, more than 10,000 people have signed up for the service. That's not a bad start, considering that rival Peapod Inc. has taken 10 years to amass a customer base of just 100,000 households.

So far, no one has been able to make much of the online food business. In 1998, less than 1%, or $235 million, of the $720 billion grocery market went to Net companies, says Forrester. By 2003, online grocers are expected to make up little more than 2%. The problem: Consumers have been largely disappointed by the service, selection, and prices they have gotten from a slew of upstarts. Even Peapod, the best-known member of the online-grocer pack, has struggled to find an economic model that works: It is still losing money in five of eight markets.

WARM AND FUZZY. That's why Webvan says it's starting from scratch--almost. Borders has cobbled together the best people and operating practices from a myriad of cyber- and real-world businesses. Federal Express Corp. serves as the blueprint for Webvan's hub-and-spoke delivery system. Webvan also has dipped into FedEx' talent pool for some of its most senior executives. Traditional grocers have been the model for maintaining food quality in transit. Wal-Mart Stores Inc. has served as an example of breadth of product selection. And for his Web site, Borders has taken a page from Yahoo! for speed, for the shopping experience, and eBay for that warm-and-fuzzy feel.

Webvan has hired some 80 software programmers to create proprietary systems that automate and link every aspect of its business processes. "This is the first back-end reengineering of an entire industry," says Kevin Czinger, Webvan's chief financial officer, a former managing director at Goldman, Sachs & Co. Gone are the scores of stock personnel, warehouse employees, and cashiers that grocery stores typically employ.

At the heart of all this reengineering is the 330,000-square-foot distribution center in Oakland, Calif., which services an area of 40 square miles in any direction. This is the prototype for the other centers that Webvan plans to build. The $25 million facility, which includes 4 1/2 miles of conveyor belts and temperature-sensitive rooms to store items such as wine, cigars, and ice cream, can serve as many customers as 20 supermarkets, which typically measure about 40,000 square feet. The difference is that Webvan can do all this with less than half the labor and nearly double the selection of available products. Webvan eventually plans to sell about 50,000 items, including 300 varieties of fresh fruit, 750 kinds of cheese, 500 types of cereal, and 700 cuts of fresh meat and fish, including live lobsters. A typical grocery store carries an average of 30,000 items.

Of course, replicating this system worldwide will be no easy feat, which is why Webvan announced on July 9 that it has enlisted the services of Bechtel Group. The construction giant will be responsible for building Webvan's distribution centers and delivery infrastructure in 26 new markets over the next two years. The cost: more than $1 billion.

Despite the hefty investment, Webvan estimates that it will spend less than 1% of its revenue on bricks and mortar when all is said and done, while conventional stores must spend an average of 6%. The result, says Webvan, is at least a 10% increase in the industry's traditionally low operating margins. And that's why the company, unlike its online-grocer rivals, plans to pass on the savings to consumers and deliver any order above $50 for free.

Skeptics say Webvan's promise is too good to be true. With so many niggling details to nail down, they say the odds that something will go wrong are impossibly high. Moreover, the massive capital investment required to roll out Webvan in big areas--from Washington, D.C., to Chicago to Los Angeles--makes the prospect of imminent profitability dim. "We want to keep our investment down to get into the 40 markets with 400,000-plus households," says rival Peapod's CEO Andrew B. Parkinson, who spends about $2 million per warehouse vs. Webvan's $25 million. "If you don't keep your investment costs down, you don't have a model that can be expanded."

Borders is undaunted. He claims that each of Webvan's facilities, because of their efficiencies, will start making money within nine months of launch. Behind Borders' boast is his company's ability to assemble an average 25-item order in less than one hour. At peak performance, the company expects to bring in $300 million in annual revenue per facility by handling more than 8,000 orders a day, involving a total of 225,000 items. The average conventional supermarket brings in about $12 million in annual revenue.

How can Webvan do it? For starters, high-tech tracking systems monitor orders from the moment they are placed on the company's Web site. Workers known as pickers are stationed throughout the distribution center to assemble orders in plastic boxes known as totes, which are color-coded depending on whether items are refrigerated, frozen, or dry. But instead of traipsing down endless aisles, the pickers travel no more than 19 1/2 feet in any one direction to reach 8,000 bins of goods that are brought to the picker on rotating carousels. "When 85% to 90% of a person's time [at a traditional store] is spent traveling to locate and assemble orders, you realize why this makes sense," says Gary Dahl, formerly a senior executive at American Stores Co. and now Webvan's vice-president for wholesale. "You want to keep people stationary so you can keep them busy."

Once a picker has finished with the task, the tote is then transported via conveyor belt to other areas of the facility housing different items. After an order has made the rounds, it is loaded onto trucks refrigerated at 35F that take it to one of 12 docking stations throughout the Bay Area. From there, the totes are loaded onto one of the company's more than 60 vans so that drivers can take the orders directly to the customers' homes. None of some 70 so-called couriers travels more than 10 miles away in any direction.

Borders, who pursued graduate studies in mathematics at the Massachusetts Institute of Technology, has spent many a night with his old math books calibrating such things as the 19 1/2-foot distance between products and workers to achieve maximum productivity. "We mocked the whole thing up in a small warehouse," says Borders. "With 10 feet, we couldn't get enough inventory to the person. Thirty feet was too far to walk." He also calculated how many products each tote should hold and how far trucks should drive from docking stations to customers' homes.

Of course, all this logistical wizardry can't prevent the inevitable snafus. And there have been plenty. Some examples: Heartier containers had to be bought to keep soft cheeses from getting crushed. Celery didn't fit into the produce bags, so Webvan had to buy bigger ones. And some totes were tipping over while traveling around the distribution center, so weights were placed inside them to keep them upright.

AMBITIOUS PLAN. Webvan is sure to experience more such mishaps--particularly given its ultra-aggressive expansion schedule. Construction on facility No. 2 is already under way in Atlanta, and the center is scheduled to open early next year. Deals are pending in other locales, both in the U.S. and abroad.

Webvan says it has to act swiftly to build its brand and head off competitors, including Peapod and, a venture backed by Webvan says that once it has achieved enough volume in each market, it can easily offer other services and products at incremental costs. Borders muses about selling books, videos, office equipment, and consumer electronics, as well as offering dry cleaning and film processing. "We see this as a cornucopia of opportunity," he says. Either that, or Webvan is going to turn out to be the biggest disappointment the Internet has seen yet.

— With assistance by David Leonhardt

    Before it's here, it's on the Bloomberg Terminal. LEARN MORE