HomeGrocer.com

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

~ the obit ~

homegrocer was an american suburban grocery product, and not even a broadly american one. its delivery footprint was the seattle metro area, portland oregon, several southern california zipcodes (irvine, fullerton, carson, azusa, san diego) and dallas. that was it. the trucks had a peach painted on the side and the drivers wore peach uniforms. you ordered from the website, gave them a one-hour delivery window, and a refrigerated truck pulled up to your driveway in a single-family-home neighborhood within that hour. the model assumed two-car suburbia, big freezers, large average basket size, and the kind of zipcode density where one truck could service eight households on a single afternoon loop.

i was 13 in 1997, on a 56k modem in turkey, and not part of any of those zipcodes. there was no analogous service in ankara or istanbul. grocery delivery in turkey at the time meant the corner bakkal sending the apprentice up the stairs with a bag, or migros and gima delivering large orders by phone-and-handcart. the suburban-house-driveway-american-suv operating model that homegrocer required did not really exist anywhere in turkey, and would not exist as a real e-grocery layer until getir and migros sanal market scaled it twenty years later under completely different urban-density assumptions.

terry drayton ran the company tight in its first two years. the founders had grocery-industry experience between them, the unit economics were defensible, and the press coverage through 1999 emphasized that homegrocer was the boring grocery operation in a category increasingly defined by webvan's louis-borders moonshot. amazon led a 42-million-dollar round in 1999. the march 2000 ipo raised 288 million dollars at the top of the dot-com bubble. by november 2000 the operation was running about 1.5 million dollars in revenue per day with 2,400 employees, which made it the largest online grocer that has ever existed at that point in time.

then the acquisition happened. webvan paid 1.2 billion in stock in september 2000 to buy the smaller, less-funded, operationally-better competitor. the seattle times timeline lays it out: webvan switched homegrocer's warehouse stack to webvan's own technology platform, and orders in san diego dropped from 1,000 to 500 a day, then to 300 inside a week. by 2001 both companies were running out of cash. on june 26 2001 the entire combined operation filed chapter 7 and the trucks stopped. amazon bought the homegrocer brand and customer list out of bankruptcy and used neither.

~ the rap sheet ~

~ how it landed ~

the homegrocer thesis was an unsexy one for 1999. drayton's pitch was that grocery is a 600-billion-dollar us category, that online penetration could plausibly reach 5 percent over a decade, and that the company that won would be the one that figured out cold-chain warehouse logistics first. the press wanted to write about webvan, which was raising 800 million and building 26-acre automated warehouses in every metro. homegrocer was getting written about as the small-ball version: smaller warehouses, denser delivery routes, lower capex per market.

that smaller-ball version was the one that actually worked at the operations layer. homegrocer's pacific-northwest warehouses were profitable per-order through most of 2000. the san diego and dallas warehouses were ramping. the founders later told reporters they thought they would have hit positive overall ebitda inside another 18 months without the acquisition.

~ the deal that killed it ~

webvan offered 1.2 billion in stock in june 2000. the deal closed in september. on paper this was a defensive merger that consolidated the category against amazon's eventual entry. in operations it was a disaster from week one. webvan insisted on migrating homegrocer's warehouse management software to its own internally-built platform, which was designed for the huge oakland-style automated facility model rather than the smaller pacific-northwest warehouses homegrocer actually had. the customer experience cratered. san diego daily orders fell from 1,000 to 300 inside the migration. routing windows missed. cold items showed up warm.

by the time webvan reversed course in early 2001 the customer base was gone. the combined company was losing 25 million dollars a month. its stock fell below the threshold for nasdaq listing. terry drayton had left the company shortly after the merger over disagreements about operational autonomy.

~ what got buried ~

the peach trucks. they were one of the cleanest pieces of consumer-grade brand identity in the entire dot-com bust era: refrigerated trucks, big peach decal, drivers in matching uniforms ringing the doorbell. customer-loyalty surveys at peak gave homegrocer an 87-percent repeat order rate, higher than any other dot-com retailer of its era. when the trucks stopped in june 2001 most of that loyalty had nowhere to go. amazon picked up the assets in 2001 and shelved the brand. amazon fresh launched five years later under a different name and a different operating model. the homegrocer thesis (smaller warehouses, denser routes, two-hour windows in actual suburban neighborhoods) is what amazon fresh and instacart eventually proved correct, just at a different scale and through a different cap table.

@}-->--

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

« previous  ·  random  ·  next »