ONDON, July 19 — John Browett, the chief executive of Tesco.com, the online arm of Britain's largest supermarket chain, is no stranger to the intricacies of home delivery. So when he visited the warehouse of an Internet grocery company in the United States last year, he noticed that it was using a fancy conveyor belt to ferry boxes around a corner, when another gadget would have done the job for half the price.

Company executives defended their choice, Mr. Browett recalled in a recent interview, as worth the expense because it enabled them to start up sooner.

"People were making some very strange decisions," he said. "They were saying things like, `I'm going to get the revenue first and work out the cost structure later.' "

Tesco was not in such a hurry; what was the rush? Instead of trying to revolutionize the mundane chore of grocery shopping with a high-tech approach relying on expensive centralized warehouses, Tesco simply added a modest online delivery service to its traditional supermarket operation.

And now that Webvan, the most prominent of the various stand-alone Internet food shopping companies has failed after pouring $1.2 billion into the attempt, Tesco's tortoiselike approach to the business is clearly more likely to succeed.

Tesco neither designed a glitzy Web site nor spent millions on the latest distribution facilities. It has invested just $56 million in its online endeavor, a business that now has $422 million in annual sales, making it the largest in the world. Last year, Tesco earned a pretax profit of $1.6 billion on total sales of $30 billion.

Moreover, Tesco says it makes an undisclosed annual profit on its online service, which analysts estimate to be in the range of $7 million. And the company is exporting its expertise to the United States in a deal with GroceryWorks, an Internet venture now half-owned by Safeway (news/quote).

Still, while Tesco's store-oriented approach appears to be winning out, the debate over which method works best is not quite over. Publix Super Markets (news/quote) of Lakeland, Fla., by contrast, is building big fulfillment centers for its online operations. Still others, like Ahold, a Dutch company that recently bought the rest of Peapod, a grocery shopping start-up serving Chicago and some cities in the Northeast, is experimenting with a hybrid bricks-and-clicks method.

Jan Hol, an Ahold spokesman, said Peapod is making a profit in the Chicago market, although he declined to provide details.

With such deep-pocketed supermarket chains taking the place of idealistic but cash-short entrepreneurs, this will be a battle drawn out over decades rather than years. Even the most optimistic forecasts now predict that food deliveries by online ordering will remain a relatively small share of the overall grocery business for the foreseeable future. Jupiter Media Metrix (news/quote) predicts that by 2006, online food shopping in the United States will reach $11.3 billion, or 2 percent of all grocery sales, up from $800 million, or 0.2 percent of total sales this year.

For the better part of the last decade, Tesco has been concerned with improving operations at its stores rather than experimenting in cyberspace. That penchant for efficiency paid off by helping Tesco increase its share of the retail food market in Britain from 10 percent in 1991 to 16 percent today, putting it ahead of Sainsbury as the country's leading supermarket chain.

Tesco took its first online order in 1996, inspired by futuristic technologies that Terry Leahy, its chief executive, had seen demonstrated at an Andersen Consulting Smart Store a year earlier. But Tesco was still unsure whether to pursue the warehouse model or to have its 690 British stores double as distribution centers.

Both strategies had risks. Building huge warehouses would cost millions of pounds that Tesco was hesitant to spend on an unproven service. But packing and picking groceries from stores could clog the aisles and frustrate customers. So Tesco ran a series of calculations that turned up surprising results.

To get enough volume to justify the cost of building a warehouse, Tesco determined it would have to serve too wide a swath of the population from one location. In London, for instance, a single warehouse would have had to deliver to an area stretching from the northwest to the center of town, a distance that would have taken hours to cover even if London's traffic were not as notoriously jammed as it is.

"The vans would have left on Friday," Mr. Browett said, "and not returned until Saturday."

But by delivering from local stores, Tesco found, no route would take more than 25 minutes.

Once it settled on its approach, Tesco set about creating a system that would allow it to pick and pack groceries efficiently from stores. Rather than walking up and down the aisles pulling a product from the shelves individually, employees use carts equipped with a computer that determines the most effective path, allowing six orders to be filled simultaneously. The products are automatically scanned into a virtual cash register, all of which has helped to reduce picking time by one-third, Mr. Browett said.

Last, Tesco determined there was no way to make a profit unless it charged a delivery fee per order of £5, or about $7. Many competitors in the United States shied away from imposing such a hefty upfront fee for fear of turning off potential customers.

Shoppers like Susan Goulding, a 34-year-old resident of Middlesex, England, are willing to pay.

"It's much more convenient to order the stuff over the site and have them deliver it rather than spending a few hours in the supermarket," Mrs. Gould said, adding that since she gave birth to a son 10 months ago she does her grocery shopping online every week.