Law #2: The Law of the Category
The Law of the Category says that if you cannot be first in your category, setup a new category. This is really just another way of explaining a concept called "differentiation".
New entrepreneurs tend to think purely in terms of finding a product which is better than the competition. But so very often, it is more important to be different than to be better. Every difference defines a category. And for each category, somebody is the leader. In other words, a large market is really just a cluster of small markets. Tackle the large market, and you will probably lose. Tackle a small market, and you might just win.
Don't think of MacOS as the number two desktop computing platform. Instead, think of them as the number one desktop computer in the graphic design category. The difference highlights the category in which Apple is number one.
Note that I am fully aware of the relative size of these two categories. Microsoft is number one in a category which is many times the size of the category in which Apple is number one. Finding a category in which Apple is number one is not an effort to claim equality. Rather, it simply explains who buys Macintosh and what differentiator is important to them.
The point of creating a category is to make sure you and your customers understand what your key differentiation is. What makes you different? To whom does that difference matter? In the minds of those people, you are number one.
But do make sure there are enough of those people. Every difference defines a category, but not every category matters. For example, suppose you want to create the number one IDE for programmers who develop enterprise accounting software in Forth. I daresay you won't encounter much opposition in your effort to win this category. However, you won't encounter much revenue there either.
Understanding categories has been critical for SourceGear as we do business in the source control market. People sometimes ask us how we can possibly sell source control tools when CVS is free. The reason is that "open source" is merely one category in the source control space. This market has quite a few categories, and the competition across those categories is minimal. If the customer wants a workflow-based tool, they only look at the tools in that category. If cross-platform support is critical, the customer isn't paying much attention to the Windows-centric products.
Perforce has played this game very well by setting up their own category called "fast source control tools". Speed is their differentiator, and they seem to be having a nice measure of success with this approach.
SourceGear Vault plays in the Windows-centric category of source control tools. I don't have data to prove it, but I'm pretty sure that right now we are number one with the collection of people who are using SourceSafe and want a painless transition to something better. We chose our category very specifically.
Ries and Trout are right -- setup a category in which you can be first. But size does matter. Make sure the category you choose is not too large and not too small.