Some Earlier Ventures

My first startup had a similar founding strategy to what I am proposing here. I developed the first product, a system for automatically rewriting existing business software to make it more understandable (and thus more maintainable) by programmers. It was a missionary market at the time because software companies rarely targeted the maintenance of exiting code (about a trillion lines of legacy COBOL). Customers knew the pain point, but I had to educate investors and the media about this invisible market segment. The first round was funded by CitiCorp Venture Capital who provided not just the capital, but helped recruit the management team to surround me.

The early market had almost no credible competitors, which allowed us to establish the leading position, but we bore the entire cost of educating the market. Following our success, IBM entered the market with a similar product and technology. This validated the market, and made us the market leader in a much larger segment. IBM never understood why we were so regularly able to beat them, given that we had practically identical technologies. Their disadvantage was that they had no product marketing function in the company, so their developers rarely worked directly with, or even listened to, customers. So we knew the pragmatic aspects of programmer preferences that they did not.

Mark Elliott was then Vice President, Software Marketing for IBM Software Group

When we sold Language Technology to a larger company, IBM hired me to help make them more externally, market focused in their software business. I was part of the first wave of outside hires that came in with Lou Gerstner.

After two years there, I left to found another startup — this one to fix the looming Y2K problem. The necessary technology was similar to what I developed in my first company, but I had to start from scratch. This time I raised seed funding to build the products first. Having learned from my first startup, I raised considerably less capital because I did not intend to spend it all on the very expensive direct sales force needed for the mainframe business market. Once the product was finished, I found an established channel partner, Computer Associates (then second in size to IBM), to be the distributer. This was not a missionary market. Everyone knew what was at stake. The market was already very crowded with vendors offering partially automated services. It did not need an evangelist. It needed a big, established sales force. So I was more than happy to private label the product to Computer Associates, letting them take the credit for the only fully automated solution. No pictures from this one. I was anonymous by design.

The product was released in the last year and a half before the two-digit year ticked over to ‘00’. It earned approximately $34 million.