Inspiration came to Louis H. Borders back in 1997. The co-founder of the Borders bookstore chain was reportedly opening a package of Japanese spices and specialty foods he had ordered from a catalog when he realized that Internet-based commerce would never take off until someone figured out a way to deliver products to people’s homes simply and inexpensively, as he told Business Week at the time. Determined to do just that, Borders came up with the concept for Webvan, an Internet venture whose ambitious goal was to revolutionize the low-margin, intensely competitive grocery business.

Armed with more than $122 million in initial funding from blue-chip companies such as CBS and Knight-Ridder and backing from top-notch Silicon Valley venture capital firms such as Benchmark Capital, Sequoia Capital and Softbank, Borders and his associates declared Webvan open for business in the San Francisco Bay area on June 2, 1999. “Webvan Group today set a new standard for Internet retailing,” the company declared.

As most people now know, for all its hubris Webvan has turned out to be one of the dot-com economy’s most spectacular failures. After burning its way through more than $1.2 billion in two years after its high-profile launch, the company declared bankruptcy in July this year. Most of its 2,000 employees were let go with minimal notice. Since then, the company has been liquidating its assets. Borders, through one of his companies, has petitioned the bankruptcy court to let him buy Webvan’s software technology platform for $2.5 million and the assumption of $500,000 in debt.

Does Webvan’s Icarian flameout mean that shoppers will never buy fruits and vegetables unless they can touch and smell them in a real-world store, and that the online grocery business has no future? For part of the answer, look across the Atlantic to a Britain-based supermarket chain called Tesco. The company’s online arm,, is on track to garner $420 million in revenues this year, and analysts reckon its profits from the online grocery business will be around $22 million. is said to have nearly 1 million registered users, 840,000 orders a year and is expanding into new catagories such as baby products and cases of wine. claims to be “the largest and most successful Internet based grocery home shopping service in the world.”

On the surface, Webvan and Tesco had the same goal: Both companies wanted to harness the power of the Internet to deliver groceries to shoppers. That, however, is where the similarity ended. Anyone who compares Webvan’s approach to selling groceries online with Tesco’s will see that each company pursued a strategy that was not just different from the other’s but poles apart. For example, while Webvan made huge bets on the Internet’s ability to change shoppers’behavior, Tesco made tiny ones. Webvan wanted to redesign the grocery industry’s infrastructure to make it more efficient; Tesco used the industry’s infrastructure to keep costs low. Webvan spent enormous sums trying to build a brand and a customer base; Tesco used its existing brand and customers to drive its online business.

Jerry Wind, a Wharton professor of marketing who explores the actions of both companies in a new book titled, Convergence Marketing (co-authored with Vijay Mahajan and Robert Gunther) notes that Webvan started with the notion that it would have to do everything from scratch, and that a new type of firm would be required to do it. “But the company did not take into account the logistics issues that were involved,” he says. “As such, Webvan had to create a whole logistics company. In contrast, Tesco followed a simple strategy. From the beginning it saw as one more channel through which to reach its existing customers as well as some new ones. It tried to provide a multi-channel experience to customers it had already attracted.” And that strategy allowed’s online grocery business to thrive.

It may be worthwhile contrasting the strategies of both Webvan and Tesco in greater detail to show how those differences led to different results.

Webvan: Speed Kills

From the beginning, an ambitious, winner-take-all attitude marked Webvan’s approach to selling groceries online. In the late spring of 1999, just as Webvan was getting ready to launch its website, Borders told the Wall Street Journal that Webvan planned to sell $300 million worth of groceries a year from a single warehouse in Oakland, Calif. “If it thrives, and even if it doesn’t, Mr. Borders plans to open another enormous grocery warehouse in Atlanta a few months later. Down the road are plans for at least 20 more such facilities throughout the U.S. in practically every city big enough to support a major-league sports team,” the Wall Street Journal wrote.

Borders raised an initial $120 million in venture capital and spent a significant part of it building the state-of-the-art warehouse, “a 330,000-square-foot behemoth adorned with five miles of conveyor belts and $3 million of electrical wiring,” according to the Journal. Although other online grocers such as Peapod were in trouble, Webvan had high hopes that it would be able to succeed where others had failed because it had invested heavily in high-tech infrastructure. Webvan executives believed that this investment would translate into higher productivity — and this would allow the company not only to beat out other online grocers but also traditional, bricks-and-mortar supermarkets.

Unlike shoppers in traditional grocery stores, who moved around aisles with carts, Webvan workers would stand at automated carousels equipped with nearly 9,000 products. Thanks to its unique technology, Webvan executives predicted, its workers would be 10 times as productive as traditional shoppers — and this would translate into faster profitability. Borders claimed the Oakland warehouse would be profitable in six to 12 months, while other warehouses might break even in as little as 60 days. “I don’t see any reason why an Internet company should take five to 10 years to be profitable,” Borders argued.

If higher worker productivity was one key element of Webvan’s strategy, another was its assumption that time-starved shoppers would respond overwhelmingly to the convenience of being able to order products on Webvan’s website 24 hours a day. These would be home-delivered within a 30-minute window of their choosing. This, the company said, would be accomplished by having a fleet of customized delivery vans to handle distribution. So efficient would this process be, Webvan executives believed, that customers would be able to shop at Webvan at the same or lower prices as they did at traditional grocey stores. “Prices are up to 5% less on average than typical supermarkets, and delivery is free for orders of $50 or more,” the company said.

Based on these twin assumptions of super-efficient worker productivity and customer-friendly delivery, Webvan embarked upon aggressive growth immediately after its website was launched. By July 1999 the company announced that it had hired the Bechtel Group, an engineering firm in San Francisco, to build 26 highly automoted warehouses for $1 billion. Each warehouse was to be modeled on the facility in Oakland. Webvan clearly wanted to grow — and fast.

Two factors underlay Webvan’s aggressive drive for growth. The first was the threat of emerging competition. Peapod had a head-start over Webvan in the online grocery market, though it was bleeding cash. A greater challenge stemmed from HomeGrocer, a Seattle-based online grocery firm. At around the same time that Webvan launched its operations, announced it had bought a stake in HomeGrocer. The Amazon-HomeGrocer combination could have affected Webvan’s prospects significantly. For Webvan, the only way to head off that threat seemed to be to make a run for dominance.

Webvan executives believed the threat of competition made the company’s drive for market dominance necessary. The second factor — easy availability of capital — made that drive possible.

In 1999 capital was flowing in tidal waves towards technology and Internet companies, especially those backed by leading Silicon Valley venture-capitalists such as Benchmark Capital and Sequoia Capital — both of which were solidly in Webvan’s corner. That year venture-capital investments reached an all-time high of $48.3 billion, an increase of more than 150% over 1998’s total, according to the NVCA and Venture Economics. More than 90% of that capital went to high-tech and web-based companies. Before a company could qualify to grab a piece of that action, however, it had to convince potential investors that it was willing to live by the Internet economy’s unwritten rule of growing at breakneck speed.

Even if someone at Webvan had wanted to try its online grocery model in one city, improve upon it, and then expand to other cities, the financial climate of those times would have had little patience with that approach. Many people involved with Internet start-ups believed that they had a narrow window of opportunity, and that they had to act fast before it slammed shut. In an interview with The New York Times, David Beirne, a venture capitalist with Benchmark Partners and an early backer of Webvan, described the situation as a Catch-22. “We had a unique opportunity to raise a lot of capital and build a business faster than Sam Walton rolled out Wal-Mart,” he said. “But in order to raise the money, we had to promise investors rapid growth.”

If rapid growth was what Webvan’s investors wanted, that is what they got. The company began rolling out massive warehouses — at a cost of more than $30 million per warehouse — in areas such as Suwanee, Georgia (serving the Atlanta market) and Carol Stream, Illinois (serving the Chicago area). Smaller distribution centers were set up in areas such as Los Angeles and San Diego, among others. On November 5, with hardly a few months of online product sales under its belt, Webvan went public in a stock offering co-underwritten by some of Wall Street’s most blue-blooded investment banks: Goldman Sachs, Merrill Lynch, BancBoston Robertson Stephens, Bear Stearns & Co. and Salomon Smith Barney. Webvan sold 25 million shares priced at $15 each, but so heady was the buzz surrounding its IPO that the stock soared to a short-lived high of $34 on its first day of trading, giving Webvan a market capitalization of $7.6 billion.

Over the next year and a half, Borders and other Webvan executives strove mightily to remain true to their vision for the company. Among their most ambitious moves was to recruit George Shaheen, the CEO of Andersen Consulting, as Webvan’s CEO, with Borders taking the chairman’s post. As the months passed, however, it became clear that Webvan was unable to get away from one simple fact: the company was spending more money on acquiring products than it could make by selling them. Some analysts reckon that Webvan lost more than $130 per order, including depreciation, marketing and other overhead.

In an attempt to gain economies of scale, which might have led to profitability, Webvan in September 2000 merged with its erstwhile rival HomeGrocer, but that too could not postpone the decline. In documents filed with the Security and Exchange Commission, Webvan reported that in the fiscal year ended December 31, 2000, the company had lost $453 million on sales of $178 million. By April 2001 Shaheen was out, and the company was scaling back dramatically. This included dropping plans for construction of new warehouses as well as slashing marketing expenses. These actions added to the perception that Webvan was in trouble and unable to stanch its financial hemorrhage.

Goldman Sachs, meanwhile, was making intense efforts to find a buyer or new investors for Webvan. When these efforts failed, Webvan had little choice but to announce on July 9 that it was closing its operations and would declare bankruptcy.

Flawed Assumptions

In retrospect, what did Webvan do wrong? Robert E. Mittelstaedt, Jr., vice dean and director of the Wharton School’s Aresty Institute of Executive Education, believes that the company’s baseless assumptions led to its blunders. To recount, Webvan assumed:

  1. That a very large number of people would prefer to buy groceries online and have them delivered at home rather than buying them at a physical supermarket. This belief led them to reckon that Webvan’s sales would explode, and that people would place a high value on not having to go to a physical supermarket.
  2. That so much inefficiency existed in the grocery industry’s infrastructure that Webvan would garner a bigger margin if it rebuilt the whole infrastructure — by doing all its own warehousing and logistics and tried to move further up the value chain by cutting out the wholesalers.
  3. That if a website gave shoppers more choice and a wider selection of products, that people would be willing to pay at least the same price if not a premium for the privilege of shopping online as they did in a physical store.

As time was to show, each of these assumptions was wrong. According to Mittelstaedt, Webvan’s biggest mistake was assuming that people did not want to shop in a supermarket. “It turns out that a large number of shoppers haven’t made their purchase decisions before going to the store,” Mittelstaedt says. “This is where Webvan ignored the basic laws of economics: You can’t get people to buy something they don’t need. When it comes to groceries, you can’t get a person to buy a delivery service that is convenient for them if they have not decided what to order.”

Had Webvan made its groceries dramatically cheaper — selling them, say, at half-price — then conceivably some people would have thought more about their needs and organized their shopping behavior to make the process work. But if the groceries “are the same price online as they are in the stores, it doesn’t have the same incentives except for a very small percentage of the population that finds buying online more convenient,” Mittestaedt notes.

Webvan’s second mistake was to try and reinvent the whole infrastructure that the grocery industry has evolved over the last 100 years. It turns out that the infrastructure may be more efficient than people realize. “Webvan spent huge amounts trying to integrate this infrastructure, and it also didn’t do it as efficiently as it thought it would,” Mittelstaedt says. “It would have taken 10 years for Webvan to learn how to integrate its infrastructure, and it ran out of capital long before that.”

Webvan’s third mistake was to choose San Francisco as its starting point. “The company assumed that that market had people with high incomes, higher interest in quality of food, and they thought that this would be a good place to start, but that market is very difficult from a traffic standpoint,” Mittelstaedt says. “It’s got hills, houses that are hard to reach, and it’s a mess from the standpoint of having to deliver things. You have to drive a lot to get to where you want to go, and it’s a traffic nightmare. That added to Webvan’s implementation problems.”

Mittelstaedt points out that if all these errors are added together, the result is that Webvan invested a billion dollars based on very shaky assumptions that don’t hold up under economic scrutiny. An electric company that wanted to invest $1 billion in building a new power station would have to look long and hard at the demand for electricity before making such a decision. A chemical company would have to look thoroughly at the demand for plastics before deciding to build a billion-dollar plastics plant.

Webvan, however, did not go through that exercise. So rosy was the view inside the dot-com bubble that it did not need to — and the company and its investors eventually paid the price for that mindset.

Tesco: Slow and Steady

With $32 billion in annual sales, Tesco bills itself as the “number one food retailer in the U.K. and the largest e-grocer in the world.” When it wanted to enter the world of e-business, however, its approach was dramatically different that Webvan’s. Tesco executives recently told Business Week that back in 1996, the company tested whether shoppers were willing to buy groceries online by introducing a single website at one store in Osterley, England. In fact, as Business Week notes, “Tesco’s big bet was to bet small.”

Early on in its e-business experiment, Tesco realized that it would have to address one key question: Should it supply shoppers with groceries taken off the shelves of its existing stores, or would demand be so high as to require the construction of dedicated warehouses? Tesco decided not to invest in the construction of special warehouses until it had a better sense of online consumer demand. The company kept testing and readjusting its online sales process, letting customers order groceries on the Internet and supplying them from its existing stores, for nearly two years — which was not only an extremely long period in “Internet time,” but also coincided with the height of the dot-com boom.

At the time, Tesco was often criticized as a company that did not “get it,” and which stood timidly by letting other, so-called “purer” web-based retailers forge ahead. By plodding along at its tortoise-like pace, however, Tesco learned a lesson that its hare-like rivals did not — that for the time being, online grocery shopping represented a niche trend rather than a full-blown mass market. By 2000, for example, though’s annualized online sales were running at a rate of $420 million a year, this was less than 2% of the company’s total revenues of $32 billion.

Taking the gradual approach helped learn at least two significant lessons. First, rather than promising ambitious home deliveries, the company experimented with having customers order their groceries online, but pick them up at a store. Customers saved on the time and effort it took to pick products off the shelf, and they found a bag of groceries waiting for them when they arrived. At the same time, however, if they wanted to add one or two items to their order, they had the option to do so.

In addition to pre-packaging orders for shoppers, also began to deliver groceries to customers’homes near each store. In an important departure from Webvan’s strategy, however, the company imposed a delivery charge right from the beginning. Not only did this approach help recover part of its delivery costs, it had another positive result: The company saw the shoppers’order sizes increase as households strove to get maximum mileage for the delivery charge.

This approach kept growing. On September 18, 2001, Terry Leahy, chief executive of Tesco plc announced that this year’s sales were “up 77% on last year, a period when we were still rolling out the service. Grocery homeshopping made good profits, however overall made a small loss of £3 million ( in the first half, reflecting the launch cost of new sites such as our wine warehouse.” He added that “made excellent progress and we now reach 94% of the UK population. In the first half our grocery homeshopping operation achieved like for like sales of nearly 40% and created 600 new jobs.”

In an effort to extend its model to the U.S., last June announced a partnership with Safeway, one of the largest food and drug retailers in the U.S. The company, which is slightly bigger than Tesco — its annual revenues are $32 billion — operates more than 1,700 supermarkets in the U.S. and Canada. Since January 2000 Safeway has been providing online grocery shopping through a Texas-based unit called GroceryWorks. As part of the deal, has bought a 35% stake in GroceryWorks for an investment of $22 million in cash as well as intellectual property and technical resources, while Safeway holds 50% of GroceryWorks. According to Leahy, the objective was to introduce the model to American grocery shoppers in collaboration with Safeway. “With Tesco’s know how and the Safeway brand we have the perfect combination to bring grocery home shopping to the world’s largest market,” Leahy said.

Will’s approach work in a market where Webvan failed? According to Mittelstaedt, it well could. “The reason why this works is that Tesco used the technology to make the existing shopping process — that people were used to — more efficient rather than trying to totally reinvent a process that people were not used to.”

Wind, too, sees considerable potential in Tesco’s approach. Like Mittelstaedt, he says that Tesco “basically has to worry just about distribution from the store to the home. This is a far more economical model — and it offers the opportunity for considerable cross-selling. Tesco has found out that by adopting this model, its online customers have increased their purchases from the stores, and people who used to shop in the stores have increased their shopping on the web. So there is a crossover effect between the two channels.”

Wind explains that this model is also starting to catch on in other parts of the world. Caprabo, a European supermarket chain that is “replicating the Tesco model in Spain, has announced that it expects to break even this year — in less than a year,” he says. “The reason is simple: The entire cost of launching was as much as opening a 20,000 sq. ft. store. And the volume of sales has been amazing. Within the first three months, even though it was just one store out a network of 52 stores that introduced online sales, that one store already accounts for 6% of Caprabo’s total sales volume.”

Ask Wind why he thinks Webvan flopped so mightily, and he scoffs: “Webvan had the arrogance to think it could conquer the world. Money was cheap, and venture capitalists were willing to fund any dumb idea. But primarily Webvan forgot to look at consumer behavior. And it forgot basic economics.”