Commerce One

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

~ the goodbye ~

i was twenty when commerce one filed for bankruptcy in october 2004, in the middle of engineering school in ankara, and the b2b exchange story had finished playing out before i had any reason to care about it. nobody in turkey was buying enterprise b2b procurement software, the whole category was an american and european corporate-buyer phenomenon that lived in trade-press articles i would not have read. the company shows up in the dot-com retrospectives as the cleanest example of an enterprise-side bubble loss, the kind of stock that consumer-internet narratives leave out because it never had a consumer interface.

the company started in 1994 as distrivision, became commerce one in 1997 when mark hoffman took over as ceo, ipo'd in july 1999 and saw its stock rise 190 percent on the first trading day. peak market cap hit $21.5 billion in 2000. the pitch was b2b digital marketplaces that would let large enterprises run their procurement, supplier discovery and auction-based purchasing on a shared electronic exchange. ariba was the bigger rival. sap and oracle were eventually the bigger threat.

in december 2000 commerce one and oracle co-founded covisint, the joint procurement platform for ford, general motors, daimler, renault and nissan. for a moment the trajectory looked like every major industrial vertical would have its own commerce-one-built exchange. that did not happen at the scale the stock price assumed. by 2002 customers had pulled back. by 2004 the company was insolvent. it filed for chapter 11 on october 6 2004. patent assets went to novell for $15.5 million. operating assets went to comvest partners for $4.1 million. perfect commerce picked up the brand in february 2006. a 99.9 percent decline from the 2000 peak.

~ the chronicle ~

~ the breakout ~

the b2b exchange thesis in 1999 went like this. every large corporation buys things from a long tail of suppliers using slow and paper-heavy procurement processes. an internet-based exchange could compress that into electronic auctions and real-time inventory matching. industry analysts at the time projected the b2b e-commerce market would hit several trillion dollars by 2003. commerce one's stock was priced against that projection. ariba's was priced even higher.

the actual product was a procurement software stack plus a hosted marketplace called marketsite. it scaled into a real business. by 2000 the company had over 1000 employees and quarterly revenue in the hundreds of millions. that revenue was nowhere near enough to justify a $21.5 billion market cap. the gap closed the hard way.

~ the decline ~

covisint was supposed to be the proof point. the automotive joint venture launched in december 2000, both ford and general motors taking 14.4 million shares each. for a stretch the idea that every industry vertical would build a covisint-style exchange held up. when the broader procurement-software market stopped buying through marketplaces in the 2002 to 2003 window, the exchange thesis fell apart. sap and oracle had built their own procurement modules and were selling them inside their existing erp footprint. customers were not going to pay commerce one for a separate platform on top of an erp seat they already had.

the financials followed. commerce one had been burning cash on the assumption of growth into the procurement thesis. when the thesis broke, layoffs followed in 2002 and 2003. by mid-2004 the company was looking for a buyer.

~ how it ended ~

the october 6 2004 chapter 11 was a clean wind-down. the patent portfolio (including the xml-based schema for object-oriented xml, or sox, which had influenced w3c standards) sold to novell for $15.5 million. the rest of the operating business went to comvest partners for $4.1 million. neither buyer kept the brand active in a recognizable form. perfect commerce acquired the leftover brand in february 2006. the peak-to-trough number, from $21.5 billion to a roughly $20 million asset sale, is the cleanest single data point in the dot-com bubble bust on the enterprise side.

the legacy is the sox xml work and the covisint joint venture, which itself eventually wound down in 2017 after compuware bought it from the automakers. commerce one is the grave that explains why the enterprise side of the bubble is mostly forgotten. it was bigger than pets.com but had no consumer brand to make it memorable.

@}-->--

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

« previous  ·  random  ·  next »