The online grocery business just keeps growing. So why can't anybody make any money at it?

Here's a conundrum: On one hand, the market for online grocery retailing has been thriving. Only last spring research firm Jupiter Media Metrix pegged it at $1 billion, up from $600 million last year and $200 million the year before. More than a million people buy at least some of their food on the Web.

On the other hand, this apparently booming market is littered with casualties. Streamline, HomeRuns, and other early entrants are dead. Peapod had to be rescued by deep-pocketed acquirers and still isn't close to overall profitability. Webvan, which had 46% of the industry's sales, went under in July -- an event that led Jupiter to cut its 2001 projection by about $200 million.

So it is with new markets: they don't exist in the abstract. On the contrary, a market appears only when a company's business model allows it to create and maintain one. Dissect the key elements of online food and you understand why this particular market is so tenuous.

In the grocery business, a 2% to 3% return on sales is considered healthy. Supermarkets earn that thin margin only by rigorously controlling three kinds of costs. They keep cost of goods low by buying in huge volume -- "full truckloads of the same product straight from the factory," as Tim Laseter, vice-president in Booz-Allen & Hamilton's operations practice, puts it. They control marketing expenses by attracting regular (repeat) customers and by enticing those customers to spend nearly all of their food budget at the store. They minimize labor costs by getting shoppers to assemble their own orders and transport them home.

Can the three elements of this business model be adapted to online selling? The answers are yes, maybe, and probably not.

Buying power? Yes (with help). Webvan, flush with cash from its initial public offering, expanded rapidly into several metropolitan areas. That would have created large-scale buying volume, except that none of the markets grew as fast as Webvan had hoped. Smaller Peapod, running into trouble, stumbled on a more fruitful approach: it sold itself to giant Royal Ahold, the Netherlands-based owner of Stop & Shop and other U.S. chains. Suddenly, Peapod could take advantage of Royal Ahold's $30-billion-a-year buying clout. Its cost of goods fell 5% to 7%.

Marketing? Maybe. Any online seller must persuade customers to give it a try. But low-margin grocers must persuade shoppers to use the service week in and week out. That's a tough order: any glitch in Web-site operations or order fulfillment discourages repeat business. Worse, customers may rely on Web ordering for only a portion of their groceries -- packaged goods, say. "If a store has only a quarter of my grocery budget, they have to acquire three other customers just like me" to make it up, says Jupiter senior analyst Ken Cassar, who follows the industry. "That's expensive." The jury is still out on whether the customer base will stabilize. A survey last spring by Gomez Inc. found that 11% of Internet users had ordered groceries online during the previous three months. About the same percentage of users had once ordered groceries online but not during that three-month period.

Labor? Oops. Self-service supermarkets drove neighborhood stores out of business partly by economizing on labor costs. Online stores reverse the trend: they pay employees to assemble and deliver the order yet still hope to offer competitive prices. The secret to this economic sleight of hand was supposed to be technology: state-of-the-art warehouses, wearable scanners, and elaborate software to ensure fast, error-free picking and packing; and sophisticated routing software to minimize drivers' delivery time. Does it work? Webvan's high-tech, $30-million-a-pop distribution centers were designed to cut head count 40%, says Booz-Allen's Laseter, and might actually have done so if they had ever operated at capacity. None came close. Peapod last spring claimed to be turning an operating profit in one market, Chicago. But look at the numbers. Gross margin per order, $47. Cost of assembly and delivery, $31. Allocation of fixed expenses and partial corporate overhead, $13. That leaves a "profit" of $3, except that the calculation still excludes all marketing costs and the remaining overhead. Still, what's a grocer to do? Raise delivery charges too much -- or make customers pick up orders at a store -- and you may shrink the market beyond recognition.

There are lessons to be drawn from the story so far. The online grocery industry may be booming, but it has been propped up by all the free money that investors once poured into dot-com dreams such as Webvan's. If the industry lasts -- and it may not because of those intractable labor costs -- it will be a tough, low-margin business of uncertain size. Why did anybody ever think it would be otherwise?

John Case is a contributing editor at Inc.

Cream of the Crop

One Web grocer is making money: Tesco, the big British chain, which expects online sales this year of about $420 million. The secret? In-store picking and packing, says the company, which saves the cost of building warehouses. A harder-to-copy advantage: location. "Margins for grocers in the U.K. are 6% to 8%, versus 2% in the United States," says Miles Cook, vice-president and director of Bain & Co.'s supply-chain practice. "At those margins you can afford to offer delivery."

Still, the Tesco model presents problems of its own. A store's online service may cannibalize its existing customer base and thus produce correspondingly lower margins. Get enough customers, points out analyst Ken Cassar of Jupiter Media Metrix, and suddenly your in-store pickers are fighting with moms or dads for that last package of strawberry Pop-Tarts. "The bigger it gets, the more poorly it works," he says.



Food for Thought

60-Second Business Plan
Hell-Bent for Lather

Business for Sale
Hey, Sailor, Wanna Get Lucky?

Please e-mail your comments to