Tesco Bets Small--and Wins Big

Britain's top supermarket chain was slammed for its go-slow approach to selling goods over the Net. Now it's the world's largest online grocer

"But oh, ho, ho, who's got the last laugh now?" --IRA GERSHWIN, 1936

John Browett, CEO of British online grocer Tesco.com, isn't smug enough to amble around the company's headquarters singing the old Gershwin tune. But few would begrudge him if he did. Assailed by analysts during the peak of the dot-com boom for its go-slow approach to selling groceries over the Internet, Britain's No. 1 supermarket chain has watched one rival after another put up the white flag. Now, Tesco.com has assumed the mantle of the world's largest and most successful online grocer. "We've been a bit lucky," says Browett, "but we've also been right."

Tesco's big bet was to bet small. In 1996, when the Web was exploding and online groceries seemed like a brilliant idea, Tesco PLC (TESOF ) dipped its toe ever-so-gently into the water, outfitting a single store in Osterley, England, to accept orders by phone, fax, and a crude Web site. The idea was to test whether customers would buy groceries without shopping in conventional supermarkets. Equally important, Tesco had to figure out whether it made more sense to pick those groceries off the shelves of its stores or build separate warehouses to fill online orders.

By March, 1998, the company had proved there was sufficient demand and that picking from stores worked. But it had to keep tweaking the process to get the economics right. It wasn't until September, 1999, that it rolled service out to 100 stores.


  Now, Tesco.com is firing on all cylinders. It has expanded to 250 outlets--more than a third of the chain's 690 British stores--enabling it to deliver to 91% of Britain's population. The business is on track to turn in revenues this year of more than $450 million and boasts a respectable net operating margin from groceries of around 5%, or more than $22 million, analysts estimate. Last year, the dot-com unit lost $13 million due to the cost of expanding into new businesses such as CDs and videos, but it was profitable on groceries. "They were the only company in the world to really get it," says retail analyst David V. McCarthy of Schroder Salomon Smith Barney.

What Tesco got was that selling groceries over the Net was going to be small potatoes for the foreseeable future. After all, the chain is expected to book sales this year of $30 billion, making its online operation a mere 1.5% of revenues. So instead of spending a fortune to build distribution warehouses outfitted with fancy technology, Tesco chose a decidedly low-tech approach. Fewer than two dozen employees are needed to pull products off the shelves in each store and schlep them in vans to customers in the neighborhood. It's kind of like an electronic version of the 1950s delivery boy.

Today, Tesco.com handles more than 3.7 million orders per year--and half of its online customers weren't previous Tesco patrons. Now, the company is building on that foundation to expand into other businesses, such as baby products and wine by the case. "We've got a chance to become the leading `last mile' delivery service in Britain, because we're taking it very incrementally," Browett says.


  That's a lesson that failed dot-coms likely wish they had learned. Many Net startups were undone by focusing so much energy on growth that they never knew whether their business models worked until they hit the wall. Says Browett, who at 37 could easily be mistaken for the head of a Web upstart: "You can't make a run for revenues and then work out the cost structure later." Despite that timeless logic, e-commerce gurus from McKinsey & Co. and Andersen Consulting (now Accenture Ltd.) questioned Tesco in 1999 for not building warehouses, prompting the company to recheck its math to make sure it wasn't heading down the wrong path. "We were clearly out on a limb when the hype was at its peak," Browett says.

Staying the course proved even sweeter for Tesco after the ignominious failure of Webvan Group Inc. The Foster City (Calif.) startup, one of the most richly funded in history, went bankrupt in July. It burned through $1.2 billion in two years trying to establish a purely Web-based grocer in the U.S. Webvan's strategy was vintage dot-com: It shot for the moon, aiming to build two dozen automated warehouses around the country, costing up to $35 million apiece, that were supposed to cut 40% off the labor expense of handling groceries. Each was meant to serve a 60-mile radius encompassing millions of potential customers. But after building only three warehouses--in Oakland, Calif., Atlanta, and Chicago--the numbers got worse and worse.

Webvan's Waterloo: Customer demand wasn't high enough to operate the facilities at anywhere near their capacity, so fixed costs swamped revenues. "The first 10,000 customers were easy to find--San Franciscans sitting on the edge of their couches waiting for Web groceries," says analyst Ken Cassar of Internet researcher Jupiter Media Metrix Inc. "But the rest of the country isn't the same." He and other analysts figure Webvan lost $5 to $30 on an operating basis for every order it delivered. When you crank in depreciation, marketing, and other overhead, the loss jumps to a staggering $132 per order.


  By contrast, Tesco's decision to pick groceries out of existing supermarkets kept startup costs low. The company spent just $58 million over four years to launch its online grocery operation, and has since laid out $29 million more to expand into nonfood items. The wisdom of that tortoise-vs.-hare approach has been validated by a host of other companies. "Now that the high-cost wacko Internet business model turns out to have been a disaster, Tesco's store-based picking is the direction everybody's going," says analyst Andrew P. Wolf of BB&T Capital Markets in Richmond, Va.

The prime example: In June, Safeway Inc. (SWY ), the No. 3 supermarket chain in the U.S., said that it would partner with Tesco to deliver groceries to online customers. The British chain is investing $22 million for a 35% stake in GroceryWorks, a money-losing Net startup half owned by Safeway. The companies already have shuttered three GroceryWorks warehouses in Texas. They plan to begin rolling out the Tesco.com system this fall or early next year in a handful of Safeway's 1,500 U.S. stores. "We liked Tesco's track record," says Safeway spokeswoman Debra Lambert. "They understand how to combine technology with bricks and mortar."

That's not to say there aren't drawbacks to Tesco's approach. Orders are automatically routed from a data-processing facility in Dundee, Scotland, to the nearest store, so customers are limited to buying only what's available there. If it happens to be one of Tesco's smaller outlets, they may have only 20,000 items to choose from, vs. 40,000 at larger stores. Analysts worry, too, that the Tesco.com model won't "scale up" when the business gets bigger. Although its fixed costs are low, Tesco.com has relatively high variable costs because more orders require more labor for picking and delivery. As a result, Tesco.com won't likely be able to reap the economies of scale that Webvan expected from its warehouses--meaning it may never become much more profitable than it is today.

On top of that, Tesco will face growing competition. Despite Browett's claim that the secret of Tesco's success lies in painstakingly tested and refined processes, some analysts think competitors can fairly easily copy or improve upon them. For instance, archrival J Sainsbury PLC, the grande dame of British supermarkets, is belatedly rolling out delivery service at 36 of its stores but will also operate warehouses near London and Manchester. "We think picking centers are the way to go in the long run," says Sainsbury spokesman Matt Samuel, "but we're using stores to get to market faster."


  None of this worries Browett. He thinks Tesco.com can hit $4.35 billion in sales--about 10 times today's levels--before running out of headroom in its stores. The biggest concern: Aisles could get clogged with pickers, interfering with everyday shoppers. But long before reaching that point, Browett says, Tesco.com likely will embrace a hybrid approach, using warehouses in dense regions and store-picking for rural customers. As for selection, he says, even Tesco's smallest markets can offer more items than Webvan's warehouses did because human pickers can grab a single jar of capers or sate sauce from a shelf, while Webvan stocked goods by the palette, limiting each warehouse to about 15,000 products.

What really riles Browett, though, is the notion that rivals will have an easy time catching up. "This looks simple on the surface, but the detail underlying it has turned out to be very, very hard," he says. That's where Tesco's by-the-numbers management has proven to be an invaluable asset. Founded in the 1920s by immigrant businessman Jack Cohen, for decades the company played scrappy second fiddle to Sainsbury's. The turning point came seven years ago, when Tesco marketing execs won an internal power struggle with the traditionally dominant purchasing managers. That resulted in the elevation of then-marketing head Terry Leahy to his current role as CEO of Tesco PLC.

From that point, says Richard Hyman, a principal at British retail consultancy Verdict Research Ltd., Tesco has obsessively focused on satisfying customer demand. By the late 1990s, it surged past Sainsbury's in sales and market share, despite its less upscale image. "We think Tesco is the best retailer in Britain, period," Hyman says. In recent years, it has expanded into Central Europe and Asia, where it now gets 12.5% of sales, and has opened catalog and financial-services units.


  But the dot-com operation may be Tesco's most unlikely triumph. Launched as a skunkworks project with six mid-level managers reporting directly to Leahy, Tesco.com got off to an inauspicious start. For two-and-a-half years--an eternity in Internet time--the unit's managers tinkered with the formula. "We went down some blind alleys and back," admits Tesco.com Chief Operating Officer Carolyn Bradley.

After rejecting phone and fax orders as too expensive and error-prone, Tesco settled three years ago on a system that lets customers place orders only over the Web. But the process of picking products off the shelf was punishingly inefficient. In the supermarket business, where margins are thinner than a slice of prosciutto, a few pennies per item can make the difference between profit and loss. And the one thing Tesco wasn't willing to do, Bradley says, was to lose money on its dot-com operation.

When the company retrenched in 1998, it came up with a nifty solution. Rather than having pickers traverse the entire store filling orders for individual customers, each supermarket is divided into six zones--groceries, produce, bakery, chilled foods, frozen foods, and "secure" products such as liquor and cigarettes. Each picker, outfitted with a rolling cart, scours a single zone retrieving products for six customers at a time. To save valuable seconds and improve accuracy, each item is scanned at the moment it's picked. Then, customer shipments are assembled in the back room and stacked in vans for delivery. Tesco.com typically fills two to three waves of orders per day, which allows customers to buy as late as noon and receive a delivery by 10 that night.


  It's not just the process but also judicious use of technology that lets Tesco.com keep expenses to a minimum. Each picking cart, for instance, is topped with a wireless touch-pad computer that plans the optimal route through the store and tells pickers what to grab, one item at a time. That lets them average just 30 seconds per item, so a typical order of 64 items can be filled in 32 minutes, at a cost of about $8.50, including labor and depreciation, say analysts. Despite Tesco's efficiencies, that's still pretty steep--some 7% of the average $123 order.

The company makes up the difference in several ways. First, it saves about 3% of the order value by not using checkout clerks, Jupiter figures. The real saving grace is that online orders tend to have higher gross margins--more than 30%, vs. Tesco's typical 25%, Schroder Salomon Smith Barney analysis shows. That's because online shoppers are more affluent and buy more profitable products, such as organic vegetables, quality meats, and private-label packaged goods. "Our success is dependent on the fact that Tesco.com's margins tend to be higher," concedes Marketing Vice-President Tim Mason, though the company wouldn't provide figures or comment on analyst estimates.

There's another surprising factor that has spelled the difference between success and failure for Tesco. When the company rolled out Web shopping, it bucked conventional wisdom and imposed a 5 ($7.25) delivery fee per order, an amount it figured the market could bear. By contrast, Webvan offered free delivery for orders over $50, which ended up costing it millions in unrecovered expenses. Tesco.com insists customers are willing to pay for service--and the quadrupling of its orders over the past year seems to bear that out.

Customer Krista Levey, 27, an executive assistant at Arthur Andersen who buys from Tesco.com about once a month, says the fee "is nothing" compared to the convenience of not having to lug home cases of bottled water on the London Underground.


  Charging for delivery proved to be a masterstroke. First, it largely covers the cost of the vans and drivers who blanket the country. Tesco.com takes in about $27 million per year from the fees, close to the estimated $34 million cost of deliveries, figures Booz, Allen & Hamilton Inc. analyst Timothy Laseter. Imposing a fee also boosts the likelihood that customers will be at home during the two-hour window for their deliveries, since they have to pay again for redelivery. That's a big win for Tesco, given that returning merchandise to the store and restocking it could savage margins.

Even more important, the delivery fee has helped raise the typical order size because customers want to get their money's worth. So the average purchase from Tesco.com is three times a typical $35 supermarket transaction, a vital contributor to the online operation's solid gross margins. Eliminating the fee in an effort to stoke demand "would take away the incentive to spend up," says Schroder's McCarthy. And indeed, Tesco.com has no interest in boosting sales if the result is a loss on operations.


  Truth be told, Tesco.com also enjoys advantages Webvan never could have recreated even with another $1 billion in funding. By being a part of the Tesco empire, it can piggyback on the parent company's advertising, branding, and customer database. As one of the best-known and most trusted names in Britain, Tesco confers instant legitimacy on its dot-com unit. Plus, the online operation gets free ads in Tesco's quarterly mailing to its 10 million affinity-card holders and has linked its Web site to store databases so customers can easily reorder products they've previously purchased online or in a supermarket. On top of that, having pickers in Tesco stores provides constant publicity for the Web service--a benefit Webvan could never enjoy because it had no retail presence.

Tangible and intangible advantages such as these have prompted some analysts to question whether Tesco.com would turn a profit if it were a stand-alone business. ABN Amro's Mark Wasilewski, for one, thinks the parent may not be charging its dot-com unit enough in-store costs--depreciation, utilities, marketing, and so on--as a way of making Tesco.com's books look better. Browett dismisses the charge. "There's no point in fooling yourself," he says. Every unit, whether Tesco's financial-services arm or its dot-com operation, has to carry its own weight, he insists.

Besides, he adds, the criticism entirely misses the point: Tesco isn't trying to create a stand-alone business. Tesco.com is merely an additional sales channel that lets the company boost revenues and push more products through its system. As long as it's not leaking red ink, it's a net gain.


  So what's next for Tesco.com? Analysts are confident it will keep expanding its business in Britain. By 2004, predicts Wasilewski, it will hit $2.2 billion in revenues--still just 7.5% of Tesco's total--and generate net profits of $181 million. Tesco also has announced the first international expansion of its dot-com business, other than the Safeway deal: By the first quarter of 2002, it aims to launch online shopping in South Korea, where it operates seven supermarkets, with 11 more on the way. Tesco chose South Korea because it has the highest residential penetration of broadband Net connections in the world, offering fertile opportunity for online shopping.

Meanwhile, at home, Tesco.com is rolling out new services such as its baby center and wine club. These offer not just commerce but also chat areas and simple content--information about, say, infant development or top French vintages. With innovations such as these, Browett may soon start singing the old Gershwin song out loud.

By Andy Reinhardt

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