Webvan

» see webvan.com as it lived, on the wayback machine

~ the obit ~

webvan was founded in december 1996 by louis borders, who had earlier co-founded the borders bookstore chain. the idea was an online grocery store that would deliver fresh food to your house inside a 30-minute scheduled window. you placed your order through a website, and a webvan truck would show up at your door the next morning out of a giant automated warehouse somewhere in the metro area.

the company spent before it earned. it placed a $1 billion order with bechtel to build out a network of huge automated distribution centers in cities it had not yet entered. it acquired a competitor called homegrocer in 2000 for $1.2 billion in stock. it went public in november 1999, raised $375 million on the IPO, and reached a peak market valuation north of $4.8 billion. for a window of about a year webvan was held up as one of the businesses dot-com era investors cited when they argued that the new economy would absorb every category, including the supermarket aisle.

i never used webvan. it was a bay area product first, then expanded to a handful of US cities, but it never operated outside the united states. by 1999 i was 15, online from turkey on a 33.6k modem, and the idea of getting groceries delivered through a web browser would have sounded like science fiction in a country where most people still walked to the corner bakkal for bread. i learned about webvan years later from articles about the dot-com crash where it appeared as one of the names that went up the fastest and came down the hardest.

~ the rap sheet ~

~ the shutdown ~

the unit economics never worked. delivering a basket of groceries to a house in a one-truck-per-route window cost more than the markup on the basket itself. webvan tried to solve this with scale. the assumption was that once warehouse throughput got large enough the per-order cost would drop below the margin. the throughput never got there. through 2000 the company posted $178.5 million in sales against expenses of $525.4 million. for every dollar of grocery revenue it spent close to three dollars to operate.

the leadership lost confidence in the model through early 2001. ceo george shaheen had joined from andersen consulting in 1999 with a lifetime payment contract worth $375,000 a year. he resigned in april 2001 and layoffs started immediately. the company shut down customer service in early july and filed for chapter 11 bankruptcy on july 9, 2001. roughly 2,000 employees lost their jobs that week. the warehouses were sold off in pieces over the following months.

the cause of death was the bet that capital could substitute for unit economics. webvan raised more money in 1999 and 2000 than nearly any other dot-com retailer, and it spent the money fastest of all of them. the model needed customer density that home grocery delivery in 1999 could not produce. amazon would eventually solve a version of the same problem twenty years later by buying whole foods, but the answer turned out to be physical store networks rather than greenfield warehouses, which is exactly what webvan refused to do.

~ what survived ~

the case study. business school programs have used webvan as a teaching example of dot-com era overinvestment for the last two decades. the specific mistake was building peak-capacity infrastructure before peak-demand existed. that pattern shows up over and over in later venture cycles, but the webvan version is the one with the cleanest paper trail because the numbers were public from the IPO onward.

the warehouse blueprint. webvan's automated distribution centers were genuinely advanced for 1999. parts of that engineering work were later cited by amazon and ocado as influences when they built their own grocery fulfillment networks in the 2010s. the technology was not wrong. the timing was. webvan tried to run the playbook a decade before the customer behavior had caught up to it.

the language. for a few years after the bankruptcy "another webvan" became shorthand among silicon valley investors for any startup that was building physical capacity ahead of provable demand. the phrase faded by the late 2000s when a new generation of founders showed up who had not personally watched the trucks stop running, but for that window webvan was the warning every series-a deck got measured against.

@}-->--

              )
             (_)
            .-'-.
            |   |
            |   |
            |   |
            |   |
          __|   |__   .-.
       .-'  |   |  `-:   :
      :     `---'     :-'
       `-._       _.-'
           '""""""

« previous  ·  random  ·  next »