You can look up Philippe Kahn in Wikipedia if you want. He started Borland International on his own and took it from zero to $60+ million per year and an IPO in less than four years. Borland has been bought and sold several times over since then. Philippe has built some other companies, he’s become famous and wealthy and he’s earned it.
I was a co-founder of Borland International, one of four members of the original board when it was founded. I had been recommended to Philippe as a business plan consultant and he had needed a business plan. We met, we worked together, and things clicked. When he offered to give me stock and asked me to join the board as the company started, I agreed.
That was in 1983. I was 35 years old but I was also a recent MBA, only 2 years out of Stanford. Philippe had far more to teach me about business than I realized. Not that he wasn’t schooled — he had a good degree in math from France — but he wasn’t MBA-schooled. And I, on the other hand, trusted analysis first and intuition later.
So as Philippe guided Borland from start-up to success, we disagreed repeatedly as he chose business strategies that defied schooling and analysis, and, over and over, he was right, and the MBA analysis was wrong. Never have I made so much money while being so often wrong.
Take pricing as an example. Turbo Pascal, which was line for line, pound for pound, one of the best software products ever made, fell into our lap in October of 1983. [Side note: that’s a good story, you can read it in Fire and the Valley, and I intend to tell it in this blog, but not now]. That was just a few months after the JRT scandal, in which somebody brought out a $30 Pascal package to compete against the $450 mainstream offering, only to go broke after charging a lot of credit cards that weren’t refunded. Furthermore, my MBA analysis pointed out, with the leader at $450 per unit there was no reason to go cheap. Too cheap would hurt credibility, I said. And we were brand new, we didn’t have working capital to handle volume.
Philippe, however, politely ignored my logic and set Turbo Pascal at $49.95 per unit. And he was so right, I was so wrong, if I hadn’t had equity to console me it would have hurt a lot. The pricing move was brilliant, that plus some very gutsy marketing got Turbo Pascal’s wings up and soaring very fast, and Borland International never looked back.
I have a second example: Quattro Pro. Philippe aimed his competing product squarely at the industry leader in 1985 and published the first "Lotus 1-2-3 compatible" PC spreadsheet. I said it was crazy to take on Lotus at that point in our history, Philippe did anyhow, and, once again, he was right and I was wrong. And again, because I was a shareholder, I benefited.
I keep this story not because I like to chronicle mistakes (although I don’t mind doing that, it doesn’t hurt and it seems useful to others) but because I think this illustrates something that happens all too often. A good educated guess often trumps classic analysis.