Pandesic

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

~ the eulogy ~

In the summer of 2000 I was sixteen and the internet at home came through a modem that made a noise before it worked. Pandesic died that July inside a layer of the industry that was completely invisible from where I sat. Nobody on a dial-up line in Turkey had any reason to know that the software behind a shopping page even had a vendor.

It started in August 1997 as a fifty-fifty joint venture between Intel and SAP, which on paper was the strongest pairing anyone could assemble for the problem. SAP owned the enterprise software that ran actual companies. Intel owned the hardware everything ran on. The plan was to take a cut-down version of SAP's system, put it on Intel machines, host it, and sell the whole arrangement to mid-sized companies that wanted to sell things online without building a technology department first.

The target customer was a business doing between ten and a hundred million dollars a year. By the middle of 1999 there were about forty of them. OshKosh B'Gosh and The Children's Place were the names anyone recognized, and Citibank was in there too. Around four hundred people worked on it, out of offices spread across three continents.

The pricing was the interesting part and probably the fatal one. Pandesic took a share of what customers sold rather than charging a license fee up front, which meant the venture carried the customer's risk as well as its own. If a client's online store sold nothing, Pandesic had installed a full system for free and then waited.

~ the dossier ~

~ two parents, one child ~

A joint venture inherits its parents' arguments. Intel thinks in silicon and volume. SAP thinks in enterprise software and multi-year implementations. A startup that has to change direction in six weeks cannot do that when both owners have to agree first, and the people sent over to run it were experienced executives from large companies rather than anyone who had launched something from nothing.

That gap is what the Harvard case study is actually about. Not one of the people running Pandesic had built a new venture before, so when the first strategy did not work there was nobody in the room whose instinct was to throw it away. The product kept getting bigger and more complete while the customers it was aimed at kept wanting something smaller and faster to install.

~ the july announcement ~

The end came in the ordinary corporate order. Chief executive Harold Hughes and president Pete Wolcott both resigned in the last week of July 2000, and Catherine Yetts was named interim chief executive to run the closing. On 31 July the venture said it was winding down operations because it did not see a timely road to profitability.

It stopped signing new contracts immediately and kept a retention-paid crew to move existing customers onto something else. Over $100 million had gone in. Intel and SAP each absorbed their half and moved on, and both companies are still here, which is the difference between a joint venture failing and a startup failing. Eight months later the whole sector followed Pandesic down and nobody remembered it had gone first.

@}-->--

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

« previous  ·  random  ·  next »