Some dotcom-era startups were just bad ideas, and others were bad timing. I think Webvan was in the bad timing category. Today there is nothing especially odd about the idea of ordering groceries over the Internet. Several successful companies use some kind of variant of that business model today with success. On April 26, 2001, Webvan had an especially bad day. It pulled out of the Atlanta market, laid off 885 employees, and planned a 1-for-25 reverse stock split in a desperate attempt to keep its stock listed. This was just 13 days after its CEO, George Shaheen, resigned, on April 13, 2001.
Webvan and the dotcom gold rush
Being first has its advantages if you survive, and selling groceries online isn’t a terrible idea. So that explains why Webvan’s November 4, 1999 IPO was successful, raising $375 million on top of the $396 million in venture capital it already had and valuing the company at $4.8 billion.
But it might have been a bad sign that the company had only made $395,000 in revenue at that point and lost more than $50 million to do it. But Amazon wasn’t profitable right away either. More worrying, though, was that Webvan peaked at $178.5 million in sales in 2000, and lost $525.4 million that year. And the IPO itself didn’t go as smoothly as it could have. It was originally planned for October 8, 1999, but Webvan ran afoul of SEC rules and the SEC intervened to mandate a cooling off period, an early suggestion that Webvan might have been more hype than substance.
So if Webvan wasn’t a terrible idea, what went wrong?
Problems specific to Webvan
Webvan made one major and puzzling mistake. None of its senior executives or major investors had any management experience in the grocery industry. Investors who have experience in your industry can provide valuable mentorship and advice. All of the startups I’ve worked for had investors with experience in their industry. If Webvan wasn’t able to attract investors with grocery experience, that might have been a good indicator that they had deeper problems.
Not hiring senior management with grocery experience seems like a major oversight. I know from a former coworker who worked in the grocery industry as a security analyst that brick-and-mortar grocery chains often expand slowly and into adjacent markets as a way to handle the logistical problems of selling perishable goods.
Hiring upper management who knew things like that might have influenced Webvan’s early decisions and helped them keep losses under control in those crucial early days. Its CEO, George Shaheen, gave up a $4 million annual salary at Andersen Consulting to head up Webvan. His specialty wasn’t groceries, but given his background in management consulting, it’s surprising he didn’t seek out lieutenants who did have grocery experience. I’m also surprised that founder Louis Borders, founder of the bookstore chain, didn’t think to seek out specific grocery experience.
Problems with being an online grocer in 1999
There were some problems that existed in 1999 that don’t necessarily exist today, and if they do, they are smaller issues now than they were at the turn of the century.
The Internet adoption problem
First, Internet adoption was still in the early majority phase. It wasn’t until August 17, 2000 that half the US population had Internet access. And people didn’t spend as much time online in 1999 as they do now. With smartphones, some people are online basically all the time today. In 1999, smartphones didn’t exist, broadband was scarce, and, believe it or not, some people still paid by the hour to go online.
Comfort level paying for things online
A second problem was a fairly large percentage of Internet users were still getting comfortable with the idea of paying for things online.
Inventory problems
Ebay had a significant advantage over Webvan in that they didn’t have to carry any inventory. They were just facilitating people selling their own stuff. Handling inventory and shipping was the sellers’ problem. Ebay just had to ensure sellers were meeting their commitments. Ebay’s overhead was much more similar to a software startup than a retail startup.
Webvan had to deal with other problems specific to the business model of selling perishable goods. The idea of buying perishable goods without being able to see it first was an uncomfortable leap of faith for many consumers in 1999. We got much more comfortable with that idea during the pandemic of 2020.
A company selling books or CDs like Amazon or CDnow also didn’t have to worry about the product expiring. Slow-moving inventory meant you had valuable capital tied up and it reduced your cashflow, but the item remained sellable. Groceries have a limited shelf life and a limited delivery area. This is one reason why, at the time of this writing, there’s still no general grocer who ships nationwide. Some will sell meal kits nationwide, but not a full line of groceries.
Webvan’s logistical problems
Just employee count gives a measure of the magnitude of Webvan’s logistical problem. Webvan had 4,500 employees. That is a huge number of employees for a startup that didn’t last 3 years. A software startup might only have 50 employees, even after four or five years in business.
Webvan also couldn’t sell nationwide. They could only afford so many distribution centers, so they had to pick populous areas with high Internet adoption to plant distribution centers. I can’t argue with any of their 10 choices in isolation. But they had too many and they were definitely too far apart geographically. Opening in San Francisco and Sacramento to start with, then adding Orange County and San Diego would have been a better strategy. Then they could have considered further expansion. Portland and Seattle are great tech hubs, and they are on the west coast, but maybe they’re too far from San Francisco. Maybe Phoenix and Las Vegas would have worked better since they’re geographically closer to the other markets and have very good connectivity to Orange County.
How Webvan met its end
Because what ended up happening was Webvan ran out of money and had to start pulling out of markets less than 2 years after its IPO, at a time when a company that age needed to be expanding. Indeed, it had planned to be in 26 markets by the end of 2001, a goal that proved far too ambitious.
And then they had to take desperate measures just to keep its stock listed on NASDAQ and to stay in business one more quarter. In addition to pulling out of Atlanta in April 2001, Webvan pulled out of Sacramento and Dallas earlier in 2001. But the bleeding continued.
Webvan ran out of money, investors ran out of patience, and the company ran out of options before spring had fully turned over to summer. That’s how it went from a market capitalization of $4.8 billion with 4,500 employees to liquidation in less than two years. It stopped taking orders in June 2001 and was out of business by July. The dotcom era was wild. Random thought: I suddenly don’t feel as bad about my Jose Canseco baseball cards.
Webvan was a solid idea. The timing and the execution definitely could have been better.
David Farquhar is a computer security professional, entrepreneur, and author. He has written professionally about computers since 1991, so he was writing about retro computers when they were still new. He has been working in IT professionally since 1994 and has specialized in vulnerability management since 2013. He holds Security+ and CISSP certifications. Today he blogs five times a week, mostly about retro computers and retro gaming covering the time period from 1975 to 2000.
As has been said in several documentaries, all you had to do was add the .com suffix to a company’s name for it to be worth more than gold.
It was an important time, failures were important for learning, and the initial euphoria ended. Many companies had great ideas but no experience. Webvan was simply unviable in 1999, but in theory it seemed revolutionary.
We have the case of Google, which waited for the euphoria, had the patience to analyze the market and is now one of the largest in the world. Other companies like Amazon felt the blow but learned to reinvent themselves.
The history of the dot-com bubble is one of the most interesting in the world.
This is a terrific analysis of the challenges surrounding the early online grocery space. I did some post mortem strategic analysis of Webvan, Peapod, et al in the wreckage of the dot com crash and you’ve identified all the key themes and for the right reasons. This kind of analysis is both fascinating and hard to do well. I always find it super interesting to understand how obviously smart people, who’d clearly learned their era’s ‘lessons of the past’, still managed to miss crucial things and, if possible, identify the one or two which primarily doomed them. The tricky part for us armchair analysts is to avoid putting on a cape and becoming Captain Hindsight, the annoying super hero who only shows up after the fact and points what you should have done. So I appreciate that you’ve did the homework on the historical context they were facing and been charitable in your analysis. For anyone interested in more, there have been some good academic papers and HBR-style case studies done on the space.
I’d add one ambient meta-theme of that time which provides context for Webvan and many other similar dot-com era failures. From today’s perspective, it’s puzzling how the business plans of so many of the new and large dot coms gave themselves so little financial runway. The fundamental mistake was that they came into existence in an expanded bubble but assumed that was the “new normal”. The bizarrely short runways and terrifying burn rates didn’t seem as crazy to them as they do today because for ten consecutive years more investment capital had always been available for even cheaper next year than last year. I can recall startups seriously debating not accepting investment being thrown at them solely because “it’ll be even cheaper next year.”
Pair that with then-prevalent idea that emerging consumer internet startups were competing in a drag race where the only thing that mattered was getting more users than your competitors. Like an F1 race on a too-narrow track, it was all about getting pole position – because passing after the first lap was considered almost impossible.
Finally, I’ll add that you identified one of my favorite ‘mistakes very smart people keep making’. Undervaluing the size of the submerged iceberg of relevant domain experience. This classic hubris-in-action constantly repeats, even today. And its especially common in technologists (in which I include myself) and successful, highly-paid consultants (particularly from the Big Five consulting firms). Sam Walton was only able to scale Walmart to the stars because he was both *very* smart and had spent years stocking shelves and driving inventory to stores. Smart people venture into new domains at their peril whenever they do so without a deeply experienced sherpa guide on the senior team.
Do you remember iwon? it was a website like yahoo but when you click you get points up to 1000 and everyday someone got $10k and in the end of the year someone wins $10 million.
i assume they went out of business. did the people really get their money?