Two years later, Jenkins no longer has to worry about Webvan -- it went bankrupt, along with most of its rivals. Meanwhile, however, Jenkins has moved ahead with his own plans for Publix in the online world.
PublixDirect, the chain's new internet shopping and home-delivery service, launched in late September. By January, PublixDirect's 38 trucks averaged 2,500 deliveries a week throughout Broward County and in south Palm Beach County. Atlanta should get the service by the end of this year. Orlando will be PublixDirect's third location.
Why spend an estimated $33 million in startup costs and $18 million a year to operate PublixDirect when the competition has pretty much disappeared?
Publix lifer Tom O'Connor, the facilities vice president Jenkins tapped to lead the chain's e-commerce efforts, explains: Over the last decade, discount warehouse stores like Wal-Mart and specialty shops such as PetsMart have stolen market share from traditional supermarkets. When Webvan finally opened in Atlanta in May 2000, Publix saw its own sales drop. "We just couldn't lose too many more pieces," O'Connor says. "If we don't do it and someone else does ..."
Despite the failure of so many e-grocers, analysts believe there's still potential in online shopping and delivery as high-speed internet connections become more prevalent and consumers feel more pressed for time. A September report from Forrester Research predicts that by 2006, 14.2 million households will spend $23.5 billion on e-groceries. The likely benefactors are traditional chains like Publix with established supply chains and loyal customers -- "online grocery's second wind," the report calls it.
Still, it's not clear that online grocers can make money. Building high-tech distribution centers is expensive, and attracting enough customers to cover those costs requires changing entrenched habits. Many shoppers want to see and touch the food they buy. What's more, they believe the prices are cheaper in stores.
The answer to how Publix thinks it can beat these challenges rests in a quiet industrial district in Pompano Beach, where O'Connor built PublixDirect's first distribution center. At 145,000 square feet, this warehouse is more than double the size of Publix's biggest store, and O'Connor says it can handle four to five times the volume.
Walking through the bustling aisles, it becomes clear why Jenkins chose an expert in building design and construction like O'Connor to head PublixDirect instead of some techie. Designing the warehouse for maximum efficiency is critical to keeping down costs, and technology is just one component.
The action starts at midnight, when techs download the previous day's orders, which powerful software crunches into routes for the drivers. At 4 a.m., the "pickers" arrive, strap small computers onto their wrists and hit the aisles. The computers tell them where to go for the items; pickers can pack anywhere from four to 10 orders at a time.
Workers then place the crates -- color-coded for dry, refrigerated and frozen -- onto the conveyor belt, where a laser beam scans their bar codes and routes them into the proper truck for delivery. The first loads start arriving in homes at 8 a.m.
Moving goods quickly is also essential to keeping customers happy. "We know that this has got to be convenient, and to be convenient, orders have to be 100% accurate and on time," O'Connor says. Seventy percent of his customers are repeat users, he says. And although customers have to be home to receive delivery, the service offers one-hour delivery windows.
So far, after a few glitches during the launch, PublixDirect seems to have it down. I've used the service five times, the orders were accurate and on time, except for once when the delivery arrived nearly an hour late.
The website allows customers to create shopping lists by clicking through a 9,000-strong menu of products; after the time-consuming setup, filling the virtual shopping cart is a breeze. The selection isn't as comprehensive as a typical store, but the major items are all here, and O'Connor says that the store has started selling bakery-made cakes and will soon sell sushi. He's even working on creating a system that allows coupons.
O'Connor knows he needs to please customers to reach profitability. Forrester Research reports that a distribution center has to process 4,000 orders of at least $120 each per week to break even. Having already reached 2,500 orders a week, after expanding to 10 additional ZIP codes on Jan. 2, PublixDirect is more than halfway there. O'Connor is coy about divulging numbers but says he agrees with the Forrester analysis.
He ticks off the savings he has over a typical Publix store: He's paying warehouse rent instead of suburban shopping center prices; he doesn't have to worry about decor or cashiers or parking lots. The center employs 150; a typical store, he says, 200 to 300.
As part of a huge supermarket chain, PublixDirect gets the same bulk discounts on merchandise as other stores. He also has access to the same administrative support, meaning he doesn't have to create human resources, payroll and marketing departments from scratch the way startup e-grocers did. In fact, he says, his marketing budget is almost nil; he's done posters at the stores and mailers in new delivery areas but mostly relies on word of mouth.
Drivers and delivery trucks add back in hefty expenses, however; plus, PublixDirect maintains a headquarters in Atlanta that's staffed with 60, including 25 information technology workers and five industrial engineers. (O'Connor says Atlanta was chosen because of its deep pool of IT workers; the chain's Lakeland headquarters, he says, already struggles to keep those workers.) The $7.95 service fee per delivery helps cover those extra costs.
"At the end of the day," O'Connor says, "we think we can be equal to or better than the established supermarket profit. But that's yet to be proven."