It's never too late to try
BACK in 1975, when Paul Allen and I were college kids, the two of us in my dorm room cooked up the first software program for a microcomputer.
Paul showed me a Popular Electronics story about the 'era of the computer in every home', and the two of us decided that software was the future.
This was the beginning of Microsoft.
Communication was simple: just Paul and me talking over Cokes and pizza.
Nobody else cared much about our opinions.
Things have changed in the past two decades.
I still enjoy junk food, but I also spend two hours a day reading and answering electronic mail from Microsoft's 15,000 employees.
Lots of E-mail comes from outside Microsoft, too.
Questions range from what it's like to be married (it's great), to what movies I like (Schindler's List and Shadowlands ) to complicated questions that would take a book to answer (and, as a matter of fact, I am writing a book).
The trouble is, I can spend my days answering outside E-mail and giving speeches, or I can run my company.
I try to do both, but I don't communicate enough to broad groups and a lot of my E-mail goes unanswered.
Think of this new monthly column as my E-mail to you.
I'm excited about writing it because it lets me communicate to a wide audience without being edited into a soundbite or filtered through somebody else's perceptions.
Not all questions come to me through E-mail, by the way. Sometimes people stop me at an airport, or a would-be entrepreneur corners me at a computer show or a college kid sends me a letter.
One student recently asked what to him was an important question.
It wasn't a big philosophical idea he wanted to discuss, although I suppose you could say he wanted to talk about the free-market economy. He wanted to know: 'Is it too late for me to get into the software industry and build a company and get rich?' I'm often asked that question, and my answer is always the same: this is a great time to be in software.
I won't say you can build another Microsoft, but you can shoot for US$2 million a year in sales by selling 10,000 copies of a $200 product.
That's pretty good and it happens all the time.
Because I remember how exciting it was to start a software company, I enjoy the success stories of others.
Small software companies are always cool.
They begin with a guy (or gal) who has an idea.
He or she gets together some friends who know how to program, and they build a product.
There's lots of craftsmanship in what they do because they care about it.
Usually they make the product for one customer and, because it's good, they find other buyers.
If you want to start a company, finding a niche is a strong strategy.
Forget about creating a word processor for writing, a spreadsheet for financial analysis or any other major product that has entrenched competition.
Instead, create a product that helps people do something specific or gives practical information in areas such as medicine, insurance, accounting, architecture or government processes.
Software like this will make many small fortunes.
If you're not satisfied with a small fortune, you'll have to get in on a generational change.
This is expensive and risky.
Every few years one generation of technology gives way to another.
Consider the arrival of the IBM PC in the early 1980s.
Microsoft bet that the IBM PC would be important and created the MS-DOS operating system for it.
The result was that Microsoft gained the leadership in operating-system software.
Nobody had heard of Lotus until it brilliantly capitalised on the generational change with Lotus 1-2-3, the first spreadsheet designed specifically and exclusively for the IBM PC.
Apple's Macintosh and Microsoft Windows were winners later, when the world embraced graphical computing and turned its back on clunky old programs that displayed only text.
To win big, you have to spot a generational change that big players are ignoring.
The bet can be expensive.
Recently some entrepreneurs gambled that software to let people scribble instead of type would lead to a new generation of word processors and spreadsheets.
They set out to create new products they thought would eclipse entrenched competition.
They were wrong.
That was an expensive bet.
What would I advise a college kid looking to become a software entrepreneur? Learn the ropes at an established software company.
Look for your niche.
Line up venture capital.
Find smart people.
And don't forget the Cokes and pizza.
Trust me, there will be plenty of late nights.
The Bill Gates Column will be published every two weeks. In every second column, Gates will answer questions from readers on any subject they wish to bring up. He calls it his 'E-mail to the World'. Readers can E-mail Gates their questions via the Internet to askbill microsoft.com, or by letter to: Gloria B. Anderson, Executive Editor, The New York Times Syndicate, 122 E. 42nd St, 14/F, New York, NY 10168, USA.