Computer History Museum Software Industry Special Interest Group Preserving the History of the Software Industry
Home History Other Resources Preservation
Developments
overview  
the 50s  
the 60s  
the 70s  
the 80s  
 
Companies   
overview  
the 50s  
the 60s  
the 70s  
the 80s  
 
People  
Anecdotes  


How the ICP Directory Began


by Larry Welke (as told to Luanne Johnson)

Copyright ©1998, Luanne Johnson, All Rights Reserved

I started publishing the International Computer Programs Directory, a catalog of computer programs that were for sale, in January of 1967, two and a half years before IBM announced unbundling, the event that supposedly created a market for software products. I’ve been told that I was really prescient to anticipate that there would be customers willing to buy software products at a time when most users expected to get software for free from the computer manufacturers, but it seemed like an obvious idea to me at the time.

I was in charge of customer data services at the Merchants National Bank in Indianapolis and kept selling services that our customers needed but we didn’t have. So I’d go back to the programming staff and say I need this by the end of the month. And they'd say two years. And nothing I was selling was new or different or unusual. Banking in those days was heavily regulated and basic banking functions were done literally by the book so every banking application was done exactly the same way by every bank. And I knew that other banks had computer programs that would do the things that our customers needed.

I started to compile a list of the programs that other banks had available and then in June of 1966, the American Banking Association Automation Conference had a swap room. If you had something to sell, you put it on one wall. And if you had something you wanted to buy, you put it on another wall with your requirements. It was the most popular room of the conference. I mean, there were more bankers there than in the bar.

The ABA said that they were going to come out with a catalog as a consequence of the popularity of the room. But we figured anything that was a volunteer effort, within an association, had no chance of ever happening, so two of the fellows at the bank and I formed a general partnership to do it on a private basis. We each kicked in $250. And after 30 days, literally 30 days, the other two figured it wasn't going anyplace and they pulled out.

So I started publishing the catalog on my own and ran it part-time for two years, doing the work at night on my dining room table. It turned out that, not only were customers willing to pay to get their hands on computer programs that they needed, they were willing to pay for a catalog that would tell them what was available.

Most of the computer programs that were listed for the first couple of years hadn’t been written as software products. They were in-house programs written by companies that were trying to recoup some of their development costs by selling copies of the programs to other users. I started with the list I had of programs that were available from banks and began writing letters to any company that I ran across in the trade press that looked like it might have a computer program of some kind that could be listed in the catalog. A lot of the companies that listed in the early issues of the catalog were service bureaus that had developed generalized versions of programs to run basic business applications like payroll or customer billing so that they could sell their computer processing services to a lot of different customers without having to do a lot of expensive new program development.

But people very quickly learned you can't take somebody else's program and make it into a product. You can take a hotel room and make it into your living room, but you can't take your living room and make it into a hotel room. It just doesn't work. There were many horror stories of people who tried to take somebody's program and turn it into a product. The architecture was wrong. The programming was unstructured.

So it wasn’t long before companies began to show up that developed software specifically for the purpose of selling it as a generalized product to lots of different users. We estimated that in 1969 about 9% of the programs we listed were developed specifically as software products. By 1973, that number was up to 49%. And the idea of selling software as a product was no longer considered to be some kind of crazy, hair-brained scheme.

top top of page
back return to the list of anecdotes

© 2007 Computer History Museum. All rights reserved.
1401 N. Shoreline Blvd., Mountain View CA 94043    Ph 650-810-1010