Delivery, Anyone?

An elevator fire forced Peapod Inc. employees to evacuate the company's headquarters outside Chicago twice last Thursday. The blaze was an apt metaphor for the online grocery industry, which last week saw two of Peapod's competitors-much-heralded Webvan Group Inc. and lesser-known Homeruns.com Inc.--go up in smoke.

"Online grocery is a murderously difficult market," says Miles Cook, a VP at consulting firm Bain & Co. "I don't expect to see a lot of online grocery stores in five years. It's a niche business."

Maybe, but don't tell that to some of the world's largest grocers. Dutch conglomerate Royal Ahold, Tesco in the United Kingdom, and Safeway all have anted up tens of millions of dollars to seek success where Webvan failed. So has Albertson's, which won't disclose the amount of its online investment.

What distinguishes the new players from Webvan is that they all own brick-and-mortar grocery stores and plan to use them to fill orders. Webvan's dream extended beyond delivering sacks of groceries: Webvan intended to break even on food sales, then use its automated network of distribution centers to sell more profitable items, from CDs to airline tickets. But after spending more than $800 million to set up its service, the 2-year-old company ran out of cash.

Established chains might be wise to rely less on automated warehouses and more on systems that make it easy for employees to locate, pack, and deliver ordered goods themselves. It's a balance Webvan didn't strike.

"You must have the right combination of technology and business processes," Peter Relan, Webvan's senior VP of technology, told InformationWeek a year ago. "The more you throw a lot of people and labor at it, the less profitable you become." But in the end, Webvan put technology ahead of the business.

"Webvan built a nice Web site, and people loved to use it, but [the company was] so enraptured with the technology, it forgot the need to make money," says Randy Covill, a senior retail analyst at AMR Research. "They were technologists first and grocers second, and that's backwards."

To recoup its $830 million investment, Webvan needed sales of $75 billion in a market with an average annual after-tax net profit of slightly more than 1%. Kroger, the industry leader with 3,500 stores nationwide, had $49 billion in sales last year. No wonder brick-and-mortar grocers say that only by serving both online and in-store customers can you make a go of the electronic grocery business. Patrick Steele, executive VP of IS and technology at Albertson's, says the challenge is figuring out how to leverage existing technology. "The costs we avoid are pricing, merchandising, and item selection," he says. "We already do that."

Another lesson some companies learned from Webvan's demise: High-tech distribution centers aren't always better than local stores for filling orders. At Webvan warehouses, the products came to the person collecting each order. Huge carousels of inventory deposited the appropriate item onto a conveyor belt and then into a bin that traveled on another belt through various product areas. The setup was visually impressive, but the company spent three times as much as it should have on the system, says Bain analyst Cook.

Safeway, which invested $30 million in GroceryWorks.com, learned a similar lesson. GroceryWorks spent more than $12 million in June of last year to build a fulfillment center in Houston and one near Dallas. Late last month, it closed them both and began filling orders from its stores. After Tesco invested $22 million in GroceryWorks last month, the Web site shut down. It will relaunch by year's end, using an in-store model based on Tesco's proprietary software for inventory management, replenishment, and truck routing.

Jeffrey Cushman, GroceryWorks' CFO, estimates that his company had to get 15,000 orders a week to break even using its fulfillment center; a store requires fewer than 100 to do the same. "The economics of a store are a lot better than a fulfillment center, but you need a partner with storefronts to pick from," he says.

Albertson's, the nation's second-largest grocer with 2,500 stores, agrees. The Boise, Idaho, company launched an online service in November 1999, sending employees to walk store aisles to fill orders. "If at some point it makes more sense to pick out of a centralized site, we have that capability," Steele says. "What we're really after is expanding our brand to consumers to allow them another way to shop with us."

The trouble with the in-store system is ensuring that customers get what they ordered. "Real-time store inventory is the challenge," says Meta Group analyst Gene Alvarez. Even with high-speed Internet connections linked to point-of-sale systems, "how do you know if someone just picked up the last can of peas somebody ordered online?"

Despite Webvan's collapse, Peapod, which went public in 1997, still believes it can fill orders efficiently by skipping the stores altogether. It operates nine distribution facilities. But Peapod does leverage its grocery-store ties with Ahold, which invested $73 million in the online grocer last year and now owns 58% of it. Peapod doesn't pick groceries off the shelves at Ahold-owned stores, which include Stop & Shop and Giant Foods, but it gets the same prices as Ahold does-better deals than Webvan ever got, because Webvan had no relationship with an established food retailer.

Although Webvan failed, Peapod executives admire its IT prowess. "Webvan put together a good technology platform," says Thomas Parkinson, Peapod's chief technology officer. The problem: "They overscaled for the business."

Parkinson should know. He's been doing this for more than a decade, making Peapod perhaps the best example of the struggle to turn a profit online. The Skokie, Ill., company, founded with $25,000 in 1989 by Parkinson and his brother, Andrew-who worked at Domino's Pizza for a year to learn the home-delivery business-didn't even use delivery trucks until two years ago. In the early days, Peapod gave software to customers and sold them the modems they needed to dial in directly to the company. The Parkinson brothers would pick and pack three orders at a time, then deliver them in their own cars.

In the first quarter of this year, Peapod finally made it into the black in the Chicago area, which accounts for a third of its 150,000 customers. Companywide, Peapod posted a $15.5 million net loss on $25 million in sales. It expects to turn a profit in Long Island, N.Y., and Fairfield County, Conn., by the end of the second quarter. And it should see increased business in its two other markets, Boston and the District of Columbia, since it competed there with now-defunct Homeruns.com.

Peapod relies heavily on IT to keep things running smoothly. The easy part was developing a Web site, Parkinson says. Peapod's biggest investments are in warehouse management, material handling, and pick-and-scan technologies, for which the company took a year to write the requirements.

Four months ago, Peapod began installing a $2 million warehouse-management system from M-Group Systems Inc. The company customized the software so that a person filling an order can select a single item, such as one banana, from a crate. Usually, supermarkets don't need to manage stock that way.

However, the real challenge is delivery: covering the last mile from warehouse to doorstep and making transportation as efficient as possible. Eighteen months ago, Peapod installed SmartFlow, software developed in-house that allocates orders to trucks and schedules routes to make sure each trip is a profitable one. The total cost was $1 million. "It's a yield-management system, just like the airline industry," Parkinson says. "In our business, everybody wants delivery Saturday morning from 8 to 10." When it can't accommodate all its customers, Peapod offers discounts to coax shoppers into accepting another delivery time.

Delivery also is important to Albertson's. To help determine the most-efficient routes, it uses RoadNet software from UPS Logistics Group. It has also added new features to wireless devices from Symbol Technologies Inc. that let items be pulled from shelves more efficiently. Albertson's already uses the Symbol devices for price checking, receiving, and inventory checks.

One online grocery company, NetGrocer.com Inc., uses a different model for deliveries. It outsources the task to Federal Express, which has trailers at the company's only distribution center in New Jersey and picks up nonperishable items packed in boxes twice a day. FedEx handles all the shipping and tracking. Shipping is based on order size--a $100 order costs $9.99--and NetGrocer subsidizes the cost. Still, the business isn't profitable.

One online grocer that's profitable is Tesco. The company is serving almost 1 million registered customers in the United Kingdom, handling 70,000 orders per week, and seeing annual sales of $420 million. But Cook points out that Tesco has an advantage over its U.S. counterparts: Margins on grocery sales in the United Kingdom are around 8%. "They can lose close to 6% [with online operations] and still be making a profit," he says. With the revamped GroceryWorks, the company won't have that luxury due to the tight margins in the United States.

One big question remains: Will enough customers be willing to pay for delivery to keep these companies in business? In Seattle, where Albertson's operates the five stores that fill online orders, the answer seems to be yes. Albertson's Web business has jumped 300% since Webvan shut down.

As a result, Albertson's last week added two stores to the ones that service online delivery there and will add another two by July 23. It hopes to capitalize on the new demand by picking inventory off store shelves, focusing on high-density neighborhoods, and keeping a close watch on transportation costs. Now, all eyes are on the bottom line.