The Webvan distribution center in East Oakland is a technological wonderland, even if it is entirely devoted to the prosaic mechanics of grocery distribution. It is the back end of Webvan.com, an e-commerce site where shoppers can click through thumbnail images of thousands of individual items, put them into a virtual shopping cart and then arrange to have them delivered to their doorstep, all for not much more than ordinary grocery store prices. The huge warehouse is a sci-fi workplace where the employees wear bar-code scanners on their fingertips and miniature wrist computers that tell them what to do, where the software driving the machinery is so prescient that it makes sure the potato chips get packed on top of the six-pack of Coke, instead of underneath. It is also the nerve center of Webvan’s Bay Area hub-and-spoke routing system, designed to time deliveries so precisely that your order gets to your house with the ice cream still frozen and the tomatoes unblemished. The Webvan warehouse is a totally digitized, super-refrigerated behemoth, capable of replacing eighteen traditional grocery stores and packing goods to be shuttled to online shoppers all around the Bay Area. The future is so very now at the Webvan warehouse, which was designed to forever change the way we shop for groceries and the way retailers deliver them to us. It’s such hip, forward-thinking technology–too bad almost no one is using it.I’ve been itching to get inside the Oakland Webvan distribution center–or DC, in Webvanspeak–since the company opened it with great fanfare in June 1999. In those heady days of Internet business speculation, when bigger was better and much, much bigger was much, much better, Webvan had well-publicized plans to roll out identical facilities in 26 cities nationwide. There was talk of online grocers capturing a hefty slice of the $500 billion the nation’s shoppers spend annually on groceries. If Internet commerce caught on, pundits mused anxiously (or eagerly, depending), it might put traditional “bricks-and-mortar” grocery stores right out of business. Opening the Foster City-based company’s first distribution center in East Oakland–near a ready supply of blue-collar labor and an affluent, computer-friendly and food-obsessed consumer market–had seemed like a particularly ingenious move for the new company. I really wanted to check things out.
But perhaps because they were shielding the proprietary software that runs the conveyor belts, or perhaps because they were leery of the e-tailing competition that was out there at the time, the PR team Webvan was then employing guarded entry to the distribution center with a vigilance worthy of a facility that was warehousing alien spacecraft, rather than bulk foods. (Surely you remember the stiff competition from other early grocery e-tailers such as Kozmo, Streamline, PDQuick, or HomeGrocer? Or perhaps you don’t.) At any rate, very few lucky people got to tour the premises. For me, the promised trip to the new center never panned out, and neither did the ride in the refrigerated delivery van, nor the much less exciting substitute offer: a chance to watch a Webvan driver hand off groceries to a real live customer. Several minutes before deadline, I caved in and accepted Webvan’s final offer, which was a phone interview with a satisfied customer. The satisfied customer had many enthusiastic things to say about the new brush for his barbecue pit that he had purchased via the Web site, but it all seemed anticlimactic. As it turned out, e-commerce as a whole was about to become pretty darn underwhelming.
Credits: Chris Duffey Webvan employee Julie Knutson Al Valentine (left) and Donte Sawyer, former Webvan employees
What a difference two years made, to Webvan as it did to hundreds of other Internet businesses. By the end of this April, an estimated 17,554 workers nationwide had been handed pink slips due to the dot-com bomb; even the staid
Wall Street Journal’s Web site began snarkily keeping a tally of dot-com layoffs. Over the two years, Webvan lost $829.7 million in venture capital, and none of the facilities it currently runs operate at anything approaching capacity. It was only during the first quarter of this year that Webvan finally turned a profit at its Orange County facility; the others are still chalking up losses. Webvan stock, once trading at $34 per share, now hovers somewhere between nine and twenty cents. In April, after the stock’s value had remained below one dollar for several months, the powers that be at the Nasdaq warned Webvan that they would be delisted if they couldn’t bring up the share price. In a last-ditch move to drive up the stock’s value–a move regarded by financial analysts as analogous to the futile flailing of a drowning person coming up for air one last time–Webvan proposed a 25-for-one reverse stock split. They’ve also appealed the Nasdaq’s delisting decision; a hearing is set for June 6.
Worse for the company’s flagging morale, key personnel have started leaving. The first to go was Louis Borders, founder of the Borders bookstore chain, who originated the Webvan idea and helped bankroll the company’s startup. Borders stepped down as chairman last fall, then resigned from the board of Webvan Group, Inc. this February (although he continues to own a quarter of Webvan’s stock).
Next to go was CEO George Shaheen. Less than two years ago, Webvan had flamboyantly recruited him away from a $4 million a year gig at Anderson Consulting with nothing more than the promise of a measly $500,000 annual salary–and lots and lots of Webvan stock. At that time, it seemed like a fair trade. Shaheen received a $13.5 million signing bonus to buy 1.25 million shares of Webvan stock, and was issued options to buy another 15 million shares. When Shaheen left, the value of those options had shrunk to about $150,000; when needled by the New York Times about his choice, he replied simply, “I didn’t have any idea of the blood bath that would ensue.” (Webvan recently forgave a nearly $7 million loan to Shaheen which had been intended to help him pay taxes on Webvan shares he had purchased but not sold. The deal let him walk away from the wreckage nearly unscathed, unlike many dot-com CEOs who got hit by the little-known but disastrous alternative minimum tax, which required them to pay taxes on paper gains made from exercising stock options at the height of the Internet boom, even if those stocks are now worthless.) Webvan also cut him a sweet good-bye package–$375,000 a year for life.
Industry analysts have speculated that the departure of two of the company’s most visible personalities might hobble its ability to raise more capital–and Webvan estimates it will need another $25 million to make it through the first quarter of 2002. In an effort to conserve capital, Webvan recently closed its facilities in Atlanta and Dallas as well as a substation in Sacramento. (Webvan still maintains service in Chicago, San Diego, Los Angeles, Orange County, and the Seattle/Portland area, as well as the Bay Area.) In February, Webvan laid off 273 workers. In April, cutbacks necessitated the layoff of 885 more–about half of them from the Foster City administrative office. The company said it was merely “right-sizing” the business.
The Webvan executives who remain insist their business formula will still work, and that the company’s recent travails have more to do with outside factors–the dip in the economy, the current skittishness of investors, and the sluggishness of consumers to warm up to online shopping–than with flaws in Webvan’s model. “We’ve always had a business plan and a path to profitability, whereas the majority of Internet companies just slapped [a plan] together, raised tons of money and said, ‘Don’t worry about it, that’ll come later,'” says spokesperson Amy Nobile, a member of Webvan’s new PR team. Webvan’s model can and will turn a profit, she says, but they have to start over on a smaller scale. “When the Internet started to boom, the priority [for venture capitalists] was, ‘Show us you have a national plan, show us you can scale this and be a national player.’ Having that national footprint and the plan to move into 26 cities was a big seller,” she says. “Over time, obviously, priorities kind of shifted and the new demand was that you have to prove the model works, first, and then expand later.”
Industry experts point to Webvan as well as fellow money-loser Amazon.com as the true litmus tests of whether or not e-commerce can succeed at all. (Amazon is the nation’s largest online retailer; Webvan is the second largest.) Both of these companies are online mega-marts that expect to finally achieve profitability due to the sheer volume and variety of goods they offer. Now that the more narrowly defined e-merchants–the Wine.coms and the Pets.coms of the world — have imploded, they say, there are few online competitors left to take on the likes of Webvan, which you can now use to buy anything from vegetables to books to beauty products to consumer electronics, not to mention wine and pet food.
Meanwhile, even if you were one of the many people who were secretly relieved that the Internet bubble burst, you should be pulling for Webvan. Unlike other dot-coms, bankruptcy for Webvan doesn’t just mean pink slips for techies and administrators–Webvan employs hundreds of truck drivers, warehouse workers, meat cutters, and other manual laborers, many of them recruited from the Oakland neighborhoods that surround the Webvan DC. These were the very people whom the political left fretted might be left out of the Bay Area’s tech boom–the people for whom the economic surge of the last two years mostly meant rising rents and gentrification. This is a place where the wealth is actually beginning to trickle down.
As for me, I just wanted to get into the DC, which in my imagination was not unlike the Willy Wonka Chocolate Factory. “Reporters often ask that,” muses Nobile when I try to delicately raise the Wonka question. She is nice about it, but she points out, very firmly, “There is no chocolate pond.” Nevertheless, it did not disappoint.
The Oakland distribution center is a squat, gray concrete building just a few blocks away from the Coliseum, and if you want to really see it in action, you have to get there well before 6:00 a.m., when most of the day’s orders are being packed for delivery and the joint is truly hopping. You should bring a parka, because the warehouse is cold. You should also be prepared to shout, because the interior reverberates with a constant rushing noise, the sound of massive refrigeration at work.
At first blush, the interior of the Webvan DC looks quite a bit like the inside of your neighborhood Costco, with its giant pallets of soft drinks and dog food bags stacked four tiers high near the entryway. What is more striking is the tangle of silver conveyor belt–over four miles of it–that runs around, over, and underneath everything and must be crossed using a series of yellow-painted stairs and catwalks. Along the way, the belt is equipped with photo eyes that read bar codes. Bins of groceries passing by the photo eye are scanned, a mainframe computer figures out which way each one should be going, and they are accordingly nudged this way or that on their journey. Webvan prides itself on being so full automated that humans rarely handle the produce. I am so busy gazing at the interplay of moving parts that general manager Matthew Mahood, who is doing duty as tour guide, has to gently nudge me out of the way of an approaching forklift.
The roller-coaster conveyor belt inside the DC is designed to sort Webvan’s merchandise according to its weight, humidity needs, and–most obviously–the temperature at which it must be kept. Ambient, or room-temperature items, are packed into yellow bins and head stage right on the conveyor system; chilled items are packed into green bins and exit stage left towards the refrigerated portion of the building. Freezer items (blue bins) are maintained at a temperature of minus 20 degrees. Workers in this part of the DC wear full-body “freezer suits”–olive-colored padded jumpsuits that reach all the way over the tops of their heads–and work in fifty-minute shifts with warm-up breaks in between. Everyone else gets a Webvan fleece jacket; here and there you can see people with woolen gloves stuffed underneath their sanitary latex ones.
Here, slightly simplified, is how the order-fulfillment process works: After customers have placed their Internet order, Webvan’s software crunches through a massively complicated program and combines it with all the other orders into a schedule for the day. Meanwhile, in the wee hours of the morning, vendors’ trucks are pulling up outside the DC to drop off their shipments. Since my tour takes place mid-morning, I get to see only the tail end of the unloading process in action, but it’s enough to get the idea. Near a loading bay, two people are cutting open cardboard boxes and dumping their contents into yellow bins. A third man with a bar-code scanner zaps the code on the plastic bin, then the UPC on one of the items inside the bin–in this case, jars of raspberry jelly–and enters the total number of jelly jars on a computer keyboard. He then hefts the bin onto the conveyor belt and it sedately rolls away. The system now not only knows that it has a dozen jars of raspberry jelly in a certain plastic bin, but it also has collected 150 different bits of information about that jelly, including its weight, volume, and perishability. This process is repeated thousands of times–Webvan estimates that they carry about 20,000 separate products.
Once the bins hit the belt, they glide off to be sorted onto shelves and racks. We head over to view what is possibly the world’s most aggressive-looking and complicated shelving unit, a giant revolving rack capable of holding 5,000 different kinds of products. Imagine a long, thin, five-tiered grocery shelf, capable of spinning like a dry cleaning carousel on amphetamines. Or rather, five dry cleaning carousels on amphetamines: one worker loads five racks at a time. Employee Julie Knutson, wearing a Webvan zip-up fleece jacket and a long blond ponytail, is replenishing inventory for the coming night’s picking and is totally focused on a silent communication between herself and the machine.
She stands with the five racks on one side of her; on the other, there’s a two-tiered conveyor belt called the “dashboard,” where she receives a steady stream of products to be shelved. She hits a button and immediately a dozen full yellow bins roll onto the dashboard. A green LED display lights up under one of the bins: it’s number 3874–rice cakes, as it turns out. Knutson pulls the tray from the dashboard and whirls around to face the carousel behind her. Now a green LED light with an arrow is pointing to the empty tray on the carousel where she should dump the rice cakes, into slot 1B. When she’s done that, she smacks a yellow “task complete” button with the palm of her hand, and the computer gives her her next assignment. While she’s pulling the next item–a bin full of Reynolds Wrap–off the dashboard, the carousel racks frenetically revolve behind her so that by the time she turns back around, the right slot is in front of her and the green arrow is guiding her to it. Knutson can do this remarkably quickly and, better yet, does not seem to have been rendered insane by the task. “Ever see that game at carnivals, the Whack-A-Mole?” asks Mahood. “This is like that. Whenever you finish one task, there’s another. It will drive you nuts.”
Since it’s after 9:00 a.m., most of the staff are busy preparing the racks and bins for the coming day’s orders; in the early morning the process goes in reverse, as items are plucked from the carousels and shelves and put into plastic totes that will be shipped out to the customers. Only the more lightweight and nonperishable items get picked from carousels like the one Knutson is using–everything else is kept on more standard shelves, and pickers select the items from them by hand. For this process, pickers’ marching orders are delivered to them via an RF (radio frequency) gun, a gray plastic device the size of a pocket novel that they strap onto their wrist with the aid of some seriously heavy-duty Velcro. The face of the device has a readout screen as well as a whole series of alphanumeric buttons. A thin cord leads to a tiny bar-code scanner that they attach to their pointer finger with tiny elasticized straps. You set off the scanner by pressing a little trigger with your thumb.
Mahood takes us to the pod where the heaviest items–the ones that must be packed into the delivery tray first–are loaded. At first they appear to have been arrayed on the shelf without rhyme or reason, with bins of beer, cat litter, bottled water, and boxes of crackers all jumbled next to one another. In fact, items are arranged on the Webvan shelf by the “velocity,” or frequency, with which they are ordered, with the most popular scattered throughout. This is to prevent pickers from bunching up in front of one particularly high-velocity item. The picker’s movements are so minutely controlled by the software that the program even calculates the minimum number of steps a worker must take to put items in the totes.
To begin filling an order, the picker first loads up a rolling cart with perhaps a dozen yellow plastic totes; each one will hold a different customer’s order. By pointing and shooting the RF gun, the worker scans the bar code on each bin. The computer now knows which orders to fill, and quickly compiles a series of tasks for the picker to perform. As the picker pushes the cart along the shelf, the display screen on his wrist tells him which grocery items to put into which tote. Each time, the picker scans the product’s UPC code as well as the tote’s bar code to make sure the right things go in each bin. A friendly employee lets me try his RF gun on; it’s like wearing a brick, and my arm lolls about lopsidedly until I get used to the extra weight. Zap. Six-pack of soda. Zap. Tub of cat litter. After a few minutes, my wrist aches, but I feel extremely cool. I am the comic-book hero of grocery checkout.
The tour continues; we stop in various brilliantly lit and chilly rooms where the mysteries of grocery assemblage unfold. In one room, two women in white aprons are bagging bananas in clumps of three, four, five, or six. (Bananas are the most popular item ordered on the Webvan site; the others in the top five are milk, oranges, strawberries, and avocados.) In another part of the shop, we find a glass case enclosing several dozen brands of cigars. In one blindingly white room, two women slice meats and arrange them on deli platters. Webvan used to run a full butcher shop, but now most meat-cutting work is outsourced; what’s left is mostly the task of assembling party platters of meats, cheeses, and fruits. We duck into the freezer, but only for a second, as the eyeglass-wearing and camera-toting members of our party complain heartily about fogged-up lenses.
We jog (briskly, with hands tingling) through the produce section, just behind a team of quality-management inspectors who are looking over the shelves. Produce is the top seller for the entire grocery business, and here we find plenty of it: individually wrapped lemons and Asian pears, plastic bags of jicama sticks and arugula, tiny bowls of fruit salad and shrimp louie, all stamped with the Webvan logo. Webvan used to have an entire section of the company devoted to cooking heat-and-serve meals, but it wasn’t as popular as they expected. The vestiges of it remain with elaborate-sounding ready-to-eat items like the “grilled chicken focaccia sandwich with roast peppers and roast garlic mayo” which also line the produce section shelves. Pickers in this section, I imagine, occasionally have to wipe the drool from their chins.
Finally, all the picking is done. A single customer ordering frozen, chilled, and ambient-temperature items would have blue, green, and yellow totes going through the conveyor system all at once–magically, they all find each other at the end of the process, where they are stacked, sorted by destination, and loaded onto the refrigerated vans. The Webvan driver then cranks up the radio, rockets to your house, arrives during a sixty-minute time window chosen by you, and helps you unpack the goodies in your very own kitchen. There. Wasn’t that easier than going to the store?
In 1996, the idea for Webvan sprang from the fertile mind of Louis Borders as the result of an unfortunate spice-ordering incident. Borders, an MIT-trained engineer, had a history of problem-solving and was no stranger to start-ups; in addition to founding the Borders Books and Music chain, he had also cofounded a software consulting company and an investment firm. But this flummoxed him: he had ordered the spices over the Internet; the delivery person had arrived. Borders was not home; the delivery person left a sticker on his doorjamb; a new delivery time was scheduled; Borders was not home; another sticker. And so on.
According to Webvan legend, Borders promptly whipped out a pencil and paper and got to work designing what the company later dubbed “the complete shopping experience”–an e-commerce venture that would take charge of all aspects of completing a customer’s order, from the Web site to warehouse to refrigerated van. Unlike companies such as Amazon, which use third-party shippers like UPS to deliver packages, Borders’ grocery store would take the product right to your door, which in e-commerce vernacular is termed “the last mile.” It would be efficient, environmentally friendly (reducing car trips to the store as well as paper and plastic grocery bags) and most importantly, a time-saving boon for busy people. Webvan’s summary annual report for 1999 compared e-commerce to the advent of automated banking–once people got used to the idea of making transactions through a machine, shopping would never be the same again. “Make a trip to the store, parade a shopping cart up and down the aisles, scour the shelves for the best quality products, wait in a checkout line, lug the bags home and unload them,” it read. “We accept this mundane and time-consuming task because we have no reliable alternative. Or do we?”
Venture capitalists liked Borders’ idea. Financiers like Softbank, Knight-Ridder, CBS Corp., Sequoia Capital, and Benchmark Capital all put up money. Combined with the funds raised by former competitor HomeGrocer, which Webvan acquired in September 2000, Webvan has raked in a total of $1.2 billion in venture capital–the second-biggest funding pool after that raised by Amazon. After Webvan’s initial public offering in November 1999, the company was hailed by the press and Wall Street alike as innovative and trend-setting. Webvan sold its first shares at $15 a pop; by the end of the day they had soared in value to $24.88. Webvan very quickly became a $8.45 billion company, at least on paper. (The IPO was marred only by a two-week delay ordered by the Security Exchange Commission–it seems someone had been blabbing business details not included in the company’s prospectus during what is supposed to be a pre-IPO “quiet period.”) And Borders wasn’t the only one pitching “last mile” delivery services; the Webvan model was replicated to a lesser extent by companies like Kozmo, Urbanfetch, SameDay, and Pink Dot, which would deliver snacks, movies, or gift items to your house via delivery vans or couriers mounted on scooters and bikes.
Today, of the “last mile” online merchants, only the grocery operations like Webvan are left standing. Kozmo–the brand perhaps most familiar to Bay Area residents–folded this April; all of them disappeared without ever turning a profit. Even the grocery ventures are having a rough time, and the ones that remain have done so by dint of partnering with bricks-and-mortar retailers. Webvan’s nearest remaining online competitor is Peapod.com, which was recently acquired in an eleventh-hour deal by Dutch grocer Royal Ahold. A few traditional grocers are gingerly testing the workability of online arms, although the industry as a whole has been slow to embrace e-commerce perhaps out of fear that they would put their own stores out of business. Albertsons.com now does business in Seattle, and Safeway has teamed up with GroceryWorks.com to try out the online delivery business in Texas. Otherwise, Webvan, which had the splashiest debut, is the only one left in the pool.
As it turns out, going the last mile is extremely expensive, especially for companies dealing in low-margin items like groceries. For companies like Kozmo, which would blithely deliver a single movie or candy bar to a customer, the cost of paying the delivery person far outweighed the value of the sale. Not long after applauding Webvan’s launch, analysts went from calling Webvan’s approach “innovative” to “profit-proof.” Critics now argue that due to flaws inherent to the home-delivery model, Webvan won’t be around for long unless it finds a way to boost order size enough to make up for delivery costs. In a report scathingly titled “The Last Mile to Nowhere,” Tim Laseter, a vice president at management consultants Booz-Allen & Hamilton, writes that home delivery could be profitable only in extremely densely-populated, affluent areas–and not even there until more people warm up to online shopping. Laseter warns that early estimates of the potential for online sales–for example, Forrester Research, Inc. had predicted $184 billion by 2004–were grossly exaggerated. “We believe the last mile may lead to the gallows rather than to the promised land,” the report gloomily concludes.
City dwellers, who tend to be tech-savvy, are likely to be receptive to e-commerce ventures, agrees strategy consultant Jasjit Mangat, of consumer packaged goods consulting firm Swander Pace & Co. But, he says, even if they like Webvan’s service, their grocery orders are typically too small to be profitable. “Your drops in these areas tend to be smaller than you would like them to be, because they are inhabited by singles or couples with no kids, the DINKs [double-income, no kids],” he says. “The families of three, four, five, or larger are out in suburbia, but once you get out there and have multiple-hundred-dollar drops, the houses are five to a block instead of fifty apartments to a block. It decreases your ability to make multiple drops per day, you have to drive a longer distance, and it takes longer to unload and carry in groceries.”
Webvan has indeed felt financial strain as the result of orders being both smaller and less frequent than the company had hoped. By their own estimate, they need about $125 million in sales a year to get their finances in the black, and they’ve been running at about two-thirds of that. They’ve fought a mostly losing battle to meet their break-even point, which they had estimated at 8,000 orders per day at about $103 per order. That translates to between 2,800 and 3,000 orders daily at the Oakland facility–although those numbers have steadily been revised downwards as expectations have diminished. Donte Sawyer, a former meat cutter at the Oakland DC who was laid off this spring, recalls the sinking morale among Webvan employees. “They used to write down how many orders we had a day on a little board, but when the numbers started going down, they stopped writing them,” he says. “After a while, the whole board was sitting off the wall on the floor.”
One of Webvan’s first moves to drive up order size was to diversify its inventory. You can now use Webvan to order a hodgepodge of nongrocery items, including books, music, movies, consumer electronics, children’s toys, beauty supplies, postal stamps, transit passes, and baby clothing: almost all of these items are more expensive–and have a higher profit margin–than basic groceries. The idea is to convince customers that the site can replace not just trips to the grocery store, but a whole host of other errands as well. Last year, Webvan overhauled the design of both its site and its logo; the logo, formerly an image of a paper grocery bag overflowing with foodstuffs, was now a more ambiguous green and blue “W.” Nobile says that the plan is working, that loyal customers are getting used to thinking, as they say at Webvan, “outside of the bag.” “You can always tell a first-time customer because they’ve got one apple and one peach and one piece of broccoli, and then in the coming weeks they start adding shampoo and razors and cleaning supplies, and then all of a sudden, a book,” she says. “It keeps expanding–wine, beer, maybe a Palm Pilot.”
Industry analysts hailed the company’s diversification as a step in the right direction. “Webvan has always been more about the van than the Web, and as many ways as you can use the van, the better off you are,” says Mangat. “Groceries are and have always been a low-margin business, but if they drive traffic and you can add on high-margin items, that’s an incentive to profitability. We think of that as a necessary expansion, something solid and good. They used groceries as a foot in the door, and then brought in other high-end items.”
Nobile puts it more bluntly: “We weren’t named ‘Groceryvan’; we were named ‘Webvan’ for a reason,” she says. “People have to eat groceries on a regular basis. It was a way to get into their home.”
And not just any home. “We’re really targeting households with kids,” says Nobile. “We’ve learned over time that they need us the most and we need them the most. They order a lot–more than the average Joe–and they are frequent orderers. It’s a love affair and a win-win situation. That doesn’t mean we don’t benefit the elderly who can’t seem to make it out there, or the disabled, or working couples, but if you really want the sweet spot, it’s households with kids.”
In order to attract mom and pop, Webvan has recently ramped up every promotional trick in the book–they’ve started using coupons, sponsored a “tell a friend” program that gives you discounts if you refer a new customer to Webvan, and pumped out a slew of direct-mail advertisements as well as a whole new radio and TV ad campaign. (Perhaps you have seen the one where the harried woman, her eyes glazed over, says “I want to spend time with my kids, but I need olive oil.”) Webvan has even started throwing parties so that groups of moms can get together and go shopping online.
It’s still a work in progress, but the new marketing campaign definitely adheres to Webvan’s original theme of saving time for busy people. On our way through the Oakland DC, we asked Mahood who Webvan’s customer base is. He held up a giant package of Huggies. “Our target demographic is households with kids,” he said. “How many people with kids want to drag them to the grocery store?”
Here are the e-commerce buzzwords from two summers ago: “footprint,” “scalability,” and “eyeballs.” “Footprint” referred to how much geographic space a company could serve. “Scalability” measured how well you could take a program that worked on a few computers and make it work on many. “Eyeballs” was the term people used to indicate not just customers, but the number of people who merely looked at your site and, conceivably, remembered your brand name. They are the reason Webvan is in this mess. Booz-Allen & Hamilton’s Tim Laseter says that during the initial “Internet land grab”–when hundreds of companies rushed to stake their claim on the Internet, businesses erroneously assumed zero-to-sixty was the best way to go. “If you had the choice to be the most famous or most profitable online grocer today, I think most people would take profitable,” he says, but “During the heyday of the dot-com revolution, people were arguing that the old economy numbers don’t matter anymore–only the number of eyeballs and the percentage growth rate. That drives you to replicate and expand your footprint, and it drove companies to expand too fast.”
Webvan, he says, went national before its feasibility had been tested. “If you look back at Wal-Mart or Circuit City, they ran for years and years fine-tuning their sales model before they replicated it. Once they had something that was working well they built more all over the place. Webvan made a big gamble on it in Oakland. The level of automation, the offering of prepared meals, the size of the facility–there were a lot of questions they could have fine-tuned over time before they went and replicated the model.”
Growing fast was expensive; it cost $35 million to build each distribution center. In 1999, Webvan contracted with Bechtel, the San Francisco-based engineering giant, to build 26 identical facilities across the US. Only four of them were built–in Seattle, Chicago, Baltimore, and North Bergen, New Jersey–and of those, only the first two are actually being used. (The Oakland DC is considered a hybrid facility–the actual building was inherited from HomeGrocer, but all the internal workings are Webvan’s own design.) Webvan officials say the company’s operations will eventually swell back out to the original 26-city plan, but, agrees Nobile, “If we had to do it again, we would have built it a little bit smaller and added on.”
For the construction of all 26 sites, Bechtel was to be paid $1 billion, plus they would get 1.8 million shares of stock as an incentive for meeting work standards and deadlines. “We did meet those incentives and earn stock,” says Bechtel spokesman Alexander Winslow. “So now we have this stock that, well, you know how much it’s worth.” For those of you not doing the math, at an average price of twelve cents a share, Bechtel’s shares amount to about $216,000. Optimism had been infectious in the late ’90s, and not just among the dot-coms.
Some of Webvan’s earliest fans were its own employees. At first, Webvan recruited heavily from traditional grocery stores or shipping companies like FedEx. Many of Webvan’s first employees were experienced grocery workers who were leaving union jobs, but they agreed to join a non-union shop because Webvan offered them an enticing package: comparable pay, promises of advancement, and offers of between 200 and 600 shares of stock. Plus, Webvan was an experimental company, an exciting place to be back in the summer of 1999. “I thought it would be an opportunity to be part of the cutting edge, a futuristic outlook on the way people shop for groceries,” says Al Valentine, one of the company’s first employees who worked in Webvan’s meat, seafood, and deli department until he was laid off this spring.
“The reason I decided to work for Webvan was that the recruiter I spoke to was the happiest person on earth,” says Devon Lewis (not his real name), a former customer-service supervisor who was laid off last month. “Everyone I spoke to was unbelievably happy. It was unreal.”
But after the first year of operations, the optimism began to evaporate. “These guys reel you in real good,” agrees former deli worker Sawyer. “They give you a lot of false promises, and they matched the pay I was making and assured me there was so much opportunity there for African Americans, and it was a growing company. My benefits started the day I was hired, they gave me 600 shares of stock, the breakrooms were catered with breakfast and lunch. I didn’t need to bring anything but a few quarters for soda or water. Then they slowly but surely began to take it all away from us.”
Valentine, Sawyer, and a host of other former Oakland DC employees–most of them laid off early this year when Webvan engineered a cost-cutting switch from freshly cut to prepackaged meats–tell of a slow decline in working conditions. Some of the more common complaints include being required to work through lunch and without breaks, frequent changes in procedures and scheduling, stressful overtime shifts especially around holidays, the lack of a redress system for grievances, the company’s switch to a new medical plan with higher premiums, concerns about top-heavy management, and the fact that their stock options were now nearly worthless. While some of these are gripes you would expect to hear from any recently laid-off employees, in the case of Webvan a few startling claims rise to the top.
Many say, for example, that the state-of-the-art DC was designed for the needs of groceries, not people–the shop is kept cold to minimize bacteria, but some felt it was too cold for workers. “In the deli it was sometimes twenty, thirty degrees and sometimes your hands and feet get numb–it’s like pins sticking you in the hands,” says former deli worker Anthony Stripling. “But a lot of us needed the money so we just sucked it up.” The work could also be physically demanding; Valentine developed both tendinitis and carpal tunnel syndrome after working long overtime shifts during the last Christmas season–one day at work his hands swelled up so dramatically that his wife had to come in and cut his wedding band off his finger. He has since had surgery on both hands.
Because many employees worked odd hours and the DC was so far from restaurants, Webvan had initially fed its workers at mealtime. “They had food galore in the beginning; that was a total perk,” says former customer-service representative Terry Machado, who was terminated when Webvan closed its Oakland call center in April. “They had catered lunches from the DC, any junk food, Rice Krispies treats, cookies, soda, coffee, and you could eat all day long. It was hard to go to lunch anyplace else and it was good that people could keep drinking caffeine. They cut out the food a couple of months before they closed the call center, so we knew they were cutting things.”
The cessation of free meals went over like a ton of bricks with employees. Worse, some of them say that the cost-cutting extended to the products that were being sent out to customers. Sawyer says that once Webvan started using prepackaged meats instead of cutting meat themselves, the quality went down. “People would open [prepackaged seafood] up and it smelled like all hell, like someone had died in the place,” he says. “They were sending out bad product–blue, green meat, dark meat with blood in the trays.” And, he adds, “When they ran out of product, they were substituting stuff people didn’t want. If you wanted hot dogs, they would give you New York sausage, and they were doing this with other meats just to make the sale. Everything to keep from shorting the product.” That included substituting chemically processed meats for organic, he says.
If you ask these former employees why Webvan is in its current dire straits, they’re unanimous on one point: too much waste. One of Webvan’s most frequent customer complaints revolves around missed delivery times, and because of that, many totes returned to the warehouse every day untouched. Then the food seemed to simply disappear. “Even though it stayed under refrigeration, the stuff that went out and came back, we never saw it again in my department. We never used anything twice,” muses Sawyer. “Other grocery stores do that–if you put a steak out and it doesn’t sell, you rewrap it and sell it the next day, but these guys never did that. Where did all those thousands of dollars go? We just kept opening new boxes.”
Valentine, who was in charge of tracking the out-of-code products that went either to food banks or the trash, says he felt that Webvan was overordering then throwing out or giving away perfectly good food that they couldn’t sell. “It almost seemed as if Webvan was purposely self-destructing,” he says. “If they’re ordering a hundred bananas and we’re only selling fifty, if we’re not selling game hens but they continue to order them, and we’re throwing away half a ton [of merchandise] every week, and they continue those type of actions, we felt as a group of individuals that the powers that be were deliberately creating their own demise.”
Granted, Valentine has a bone to pick with Webvan. Not only was he terminated after he tried to claim worker’s comp for his injuries, but he had previously applied for management positions three times, only to be turned down. Several other people in his department supported him in his repeated bids for management; they feel he was turned down because he is African American, and point out that Webvan’s upper management is mostly white, and the rank-and-file mostly black. “Webvan was making them promises of advancement, training, higher pay, but instead of that we started getting subtractions, while the upper management itself was getting raises and promotions,” he says. “It diminished morale; our sick leave started going up drastically. People quit, but Webvan was unable to replace these professionals because it was nonunion in the first place, and after word spread, nobody was going to touch Webvan with a ten-foot pole.”
Once Webvan hit rough economic times, the relationship between management and workers only got more strained. “When I first got there, it was a fun place to work, but after a while they wouldn’t tell you anything,” says Stripling. “We would go ask managers different things, and they wouldn’t look you in the eye. They would be looking down at your feet.”
By the middle of last summer, the grocery industry professionals Webvan had originally hired started leaving, and the company replaced them with temps, or with less experienced workers. It got so bad that Webvan started offering a $2,000 bonus for bringing in a new recruit. In the customer-service department, where there were increasing complaints about incorrect grocery orders, the sentiment grew that new workers weren’t quite as proficient as the old ones. “People in the DC were largely temps and always being cycled through, so there was no real continuity of someone having their job for a long time,” says Lewis. “There were a lot of errors, missing items, and we’d end up apologizing profusely. A customer may or may not keep the incorrect item, and then we’d have to schedule another delivery which of course would cost more money.”
In fact, customer-service reps would do more than apologize–they could issue coupons and refunds. By Webvan’s own estimate, in 1999 they lost $1 million on make-goods to customers. There are tales of customer-service reps who, appalled at the egregiousness of certain errors, ran out to grocery stores and bought the missing items with their own money. And then there was that horrible Thanksgiving when a slew of customers just didn’t get their turkeys at all.
Not all customers could be quelled with coupons. “Sometimes people would get e-mail letters telling them fifteen to twenty items were missing,” says Machado. “Some people were like, ‘If we can’t get the whole delivery, we don’t want it,’ and that’s a really big way to lose money.” She feels that in recent months, service quality was truly beginning to slip. “Some people would say that they’d been having really good deliveries since the beginning, but lately the drivers weren’t as clean, courteous, pleasant, or willing to help people unload, which I think was because they were getting a large delivery load and had less time to do stuff,” says Machado. “Customers would say, ‘I really love your service, but the last four orders had items missing, so the next time, that’s it.'”
Even though Webvan is highly automated, there is still room for human error. After all, people working late at night can make mistakes, and the system was so complex that if someone accidentally dumped a product in the wrong bin at the beginning of the process–say, putting cans of Spaghetti-O’s in the bin that should have held peanut butter–the error would permeate the whole system. Likewise, if a customer ordered an item that the Web site listed as in stock, but the next morning the vendor failed to deliver it to the DC on time, it just wouldn’t get packed.
And then sometimes the high-tech system just wouldn’t work as well as it was supposed to; a glitch in one part of the conveyor belt system, for instance, could stop operations throughout the DC. Rich Hedges, spokesperson for the United Food and Commercial Workers (UFCW) and a man who is used to the pitfalls of the grocery business, can cite a whole list of technical problems that Webvan encountered. “Their conveyor was supposed to keep things frozen till they were ready to go, but it didn’t work right because the cold jammed it up,” he says. “Their freezers have like forty-foot ceilings–the cubic feet that you’re cooling compared to a grocery store is just tremendous. Every truck has to be a reefer, but they have no cool curtains, so every time they open the door they have forty feet of cool going out. It’s the most energy-inefficient company. If all the greens would realize what an energy waste this company is…” he says, trailing off.
Early on, Hedges and the UFCW, as well as the Teamsters, had approached the DC employees about unionizing, but it wasn’t until the summer of 2000 that some Webvan employees got serious about organizing. Hedges estimates that about 300 employees signaled their interest, and the unions started coming onto the premises to meet with employees and circulate authorization cards for workers to sign.
Tensions between the unions and Webvan management quickly got ugly. Hedges blasts Webvan as “a company with 21st-century technology and 19th-century labor relations.” In February, the Teamsters Local 70 and the UFCW Locals 120 and 870 joined to file a complaint with the National Labor Relations Board regarding what they claim are union-busting tactics used by Webvan management. Among other complaints, the unions charged that Webvan was illegally prohibiting its employees from approaching unions or from gathering on Webvan property to discuss workplace issues during their free time. The unions also say Webvan prohibited its workers from using company e-mail to send information regarding the unionizing effort.
“They did what most employers typically do–they brought in some consultants who advised them on anti-union tactics, they held meetings and coerced employees not to sign [authorization] cards, they tried to convince them that they didn’t need a union to represent them, and that was somewhat effective,” says Odus Hall, spokesperson for the Teamsters. “Typically, they argue that ‘we’re a team and we don’t want to interject a third party between the rank-and-file and the management,’ but the reality was far from the nirvana idea that they put out there. The reality is like being in the same family, but you’re the part of the family that lives on the wrong side of town.” Valentine, Sawyer, and others say that Webvan openly discouraged the organizing effort by attaching anti-union notes to workers’ paychecks, corralling groups of workers for lectures, and telling employees that they would have to pay exorbitant union dues.
Webvan stoutly denies the unions’ claims; Nobile calls their complaints “completely unfounded.” “We’ve encouraged our associates to do whatever they want to do on their own time,” she says, although she says that Webvan did tell the unions they weren’t allowed on company property due to the hazards of having strange people wander into the highly mechanized DC. “There’s some real dangers with people barging in and coming onto the property; that’s not acceptable.”
Unionizing dot-com workers is a relatively new idea, although there have been semisuccessful attempts at Amazon and E-Town. But Webvan certainly seems to have anticipated the idea–in the company’s SEC filing for the year 2000, Webvan listed labor organizing as a possible reason why the company’s business plan may not perform as expected. The company’s employee handbook explicitly spells out some rules that the unions insist are patently illegal: forbidding the solicitation or distribution of literature by labor or fraternal organizations, either in person or by e-mail, for instance. Currently, the union’s charges are under review by the National Labor Relations Board in Washington, DC; the board is expected to issue a complaint within the next few weeks.
Unionizing efforts at the Oakland DC seem to be in something of a hold pattern, given Webvan’s uncertain economic fate. Nevertheless, the union reps say they still want to achieve representation for Webvan’s current staff so that someone will advocate for them during bankruptcy and closure proceedings, should Webvan go under. “Generally, unrepresented employees are somewhere behind the garbage company and the water company in their claims during a bankruptcy,” says Hall dryly. Still, he says, “It’s never in our interest for an employer to go out of business. You can’t get a collective bargaining agreement with an employer that doesn’t exist. So we hope they survive, but it doesn’t look good.”
Has Webvan reached the end of the road? The company’s own independent auditors, Deloitte & Touche, fanned the flames of investor doubt earlier this year by attaching a statement to Webvan’s recent SEC 10-K filing saying that the company’s recurring losses “raise substantial doubt about its ability to continue as a going concern.” April’s layoffs lowered morale and occasioned many dire headlines. (The best tend to appear on the
Industry Standard’s Web site, which runs zingers like “Knock, knock. Who’s there? If you’re in Dallas, not Webvan.”) “The press are just ruthless right now,” sighs Nobile. “The media is really not seeing the big picture. They just want to see who the next E-Toys is. The minute they see a layoff it’s like ‘Oops, they’re dead.’ We feel like we really had to make this decision to move forward. We know this works. It’s not like, ‘Let’s just take this downward spiral.’ It’s, ‘Let’s take this step back to move two steps forward.'” Of course, Webvan management has had to radically adjust its goals for the next year, going from plans for pie-in-the-sky expansion to making dramatic cuts to reduce its expenses. “A mere ninety days ago, we stated that we would need $40-60 million in the first quarter of 2002,” says Bud Grebey, another Webvan spokesperson. “We believe with the changes we made and are making we now we will only need $25 million, so we’ve cut our capital needs by at least half.” Webvan is in discussion with existing investors for that money, and Grebey says they anticipate being self-sustaining and profitable by the second half of 2002. “Clearly,” he says, “there’s difficult news with regard to colleagues being laid off, but there’s also–at least from the perspective of the business and customers and shareholders–a real silver lining to that news, and that is that we’ve eliminated the risk of failure.”
Well, even if they haven’t quite eliminated it, they may have staved it off. Webvan still has $210 million left to burn through this year, as well as two intangible assets–brand-name recognition and customer loyalty. They take heart in the fact that over 84 percent of their orders are from repeat customers, and that during the first quarter of 2001, the average order size had finally jutted past the break-even point (at $114 an order). “If the customers we’ve got now ordered just once or twice more during a quarter then we’ll wildly exceed our expectations,” says Nobile. “We’ve got the folks–it’s not that we need millions and millions of customers we don’t have. It’s not that hard a goal to reach, and we’re working on it slowly and surely.”
At the end of April, new CEO Robert Swan even took the unusual step of sending a personalized e-mail to Webvan customers asking them to bear with the company through what he called “the current climate of uncertainty in e-commerce.” The letter asks customers to continue to refer their friends to the site, and encourages them to order nongrocery items to drive their order sizes up over $100. The company is also pulling out all the stops with new initiatives, like the recently piloted Webvan@Work program, which targets business managers who need to buy office basics like toner, copy paper, and coffee–and maybe a sushi platter or two. Other last-ditch moves include hiking delivery fees to encourage orders over $100, a flirtation with the idea of same-day delivery service, and the advent of the “Webvan Rewards” program, which gives shoppers bonus points redeemable for reserved delivery times and customer-service privileges. The company has also started sending out surprise gifts to its best customers.
But those who see Webvan as nearing the end of the road are already predicting what will happen next. On one hand, there are those who foresee a merger with Amazon, which already owns six percent of the combined Webvan/HomeGrocer operation. After all, if you could mix Webvan’s affluent, urban customer base with Amazon’s huge product selection, you might finally see shoppers placing orders large enough to be truly profitable.
Then there are those who advocate a “clicks-and-bricks” approach: partnering with a traditional grocer looking to ease into online sales. “Webvan’s model is one the traditional grocers would have never taken; they would have obsoleted eighteen grocery stores. It took a pure-play company like Webvan to go do this, but because they overextended themselves, now they need to be rescued,” says Laseter. “They never would have done it on their own, but if someone can buy Webvan’s assets for pennies on the dollar–which is not inconceivable–then the economics actually get pretty good. They don’t have all this overhead cost for building those distribution centers, their break-even point is lower, and that gives them time to let customers grow into this model.”
Despite the economic slump of the past few months, Laseter still sees online grocery shopping as a rising phenomenon. “It wasn’t as dramatic and fast a shift as what everyone was predicting, but there really is a good value proposition for a subset of consumers–a growing subset of customers,” he says. “As bad as the story is about e-tailers, online retailing is still a growth market. It reached $30 billion last year and that still is only one percent of the total [US retail], so there’s room for growth.” So while electronic shopping right now serves mainly a niche market–early adopters who have mostly stayed loyal to the concept–it’s a niche that may well become increasingly profitable as more homes get wired and more people get used to making purchases on the Internet. As Nobile puts it, “This is a sea change. The adult generation right now is on the cusp–we’re learning it, and some people feel comfortable doing it. The next generation coming after ours is not only going to want it, but demand it.”
The bottom line is: can that demand accelerate fast enough to save Webvan? The company has enough cash to get it through the year; whether or not that is enough time for them to finally turn a profit–or convince deep-pocketed investors to feed them yet more cash until they do turn a profit–is still almost impossible to speculate. For now, even the company’s harshest critics agree that even if Webvan doesn’t survive, online grocery shopping most likely will, albeit probably at a more sedate pace and in a less flashy manner. “Webvan has been a pioneer in many respects,” says consultant Jasjit Mangat. “It was quite a large idea they stepped out to tackle, and who’s to say that if the capital market was in a different state right now they wouldn’t still have the wind in their sails?”
