Computer Usage Company

~ the last words ~

What this company invented had stopped having a name long before I wrote a line of code anyone paid for. Software was a product. You bought it, or you hired someone to write it, and nobody thought the arrangement needed explaining. In March 1955 it needed explaining badly, because it did not exist.

Elmer C. Kubie and John W. Sheldon had worked together at IBM, where computers came with programmers attached and the programming was something the vendor or the customer did as part of the deal. They left to sell the programming by itself. Computer Usage Company, Inc. was incorporated in Delaware on 3 October 1955, on an initial investment of $40,000 that had to support the two founders and a staff of five.

The first project was a program written for California Research Corporation to simulate the flow of oil. That is a good first customer for a company with no precedent to point at, because reservoir simulation is a job where the arithmetic is obviously beyond a room full of people with desk calculators, and the client can see the value without a sales argument about what software is.

It grew for twelve years and went public in April 1960. The high-water mark was 1967, with over 700 people in 12 offices and revenues over $13 million. Then it declared its first loss as a public company in 1969 and spent the next seventeen years sliding. In 1986 it declared bankruptcy and was liquidated under Chapter 7.

~ the record ~

~ five people and forty thousand dollars ~

The staffing records are the most interesting thing in the company's paperwork, because they show what the first software firm thought a software firm was. By 1960 it had grown to three managers, 37 mathematicians, six physicists and three engineers. Not a programmer in the list, by title. The job was understood as applied mathematics that happened to be done on a machine.

The offices followed the customers rather than the talent, inverting what this industry later did. Washington, D.C. opened in 1959, for the government work, and Los Angeles in the spring of 1961.

~ selling a thing nobody sold ~

Both the Wikipedia article and the IT History Society describe it as the first independent company to market custom software programming services, with the careful hedge that it is "generally considered" the first to develop software independently. That hedge is doing real work. RAND's System Development Division was spun off as the System Development Corporation in 1957, which is a competing claim with a serious case behind it, and the honest version is that CUC was first by a couple of years in a field so new that being first in it was not yet an advantage.

The company also spread the idea deliberately. A Computer Usage Education subsidiary was formed in 1965 under Ascher Opler, publishing software books and running courses, and one of its best sellers was Programming the IBM System/360, which Opler edited for Wiley in 1966. The subsidiary sold the same knowledge the parent sold, in books and classrooms rather than under contract.

~ the slow squeeze ~

Sheldon left in 1964 and Kubie in July 1968, and the losses started the year after that. The record shows a long run of losing years and no single catastrophe anywhere in it.

The last numbers are stark enough to need no commentary. In 1985 CUC lost $2.4 million on revenues of only $1.5 million. A company whose revenue in 1967 was over $13 million was, eighteen years later, losing more than it billed.

Kubie died in 2004. His obituary says that in 1955 he co-founded the first independent software company, which is the correct claim stated plainly, and is more than the company got while it was trading.

@}-->--

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

« previous  ·  random  ·  next »