When I start a consulting engagement, I ask my new clients “what’s a Company X?”  This is shorthand for “what do you stand for, how do you do it, and why should anyone care?”

When I hear “Uber for pet food” or “Airbnb for cupcakes” or “YouTube killer for soccer moms” my conclusion is that they’ve decided, perhaps unintentionally, to compete in a category that already exists. The strategy is to disrupt the existing category with a novel twist.

The problem is it doesn’t usually work. Seth Godin’s piece today talks about being lost on the shelf in a digital world where the number of SKUs is theoretically infinite and the ability of the customer to put options side-by-side for comparison is instant and free. Lots of choice, lots of pretty similar products, many with novel twists.

Categories with lots of players tend to become a market of a dominator, a strong second, a distant third, and everybody else. Everybody else competes on the price set by the market dominator, except without the same economies of scale as the top two or three players.

Al Ries has been writing about categories for years and suggests at least two options. You can either become a category killer or create a new category. In the former, you create something that’s so obviously better that it kills the existing category and attracts the customers to the new category. To demonstrate this, he uses the classic example of P.R. Mallory creating Duracell and the alkaline battery market. The result was that the carbon battery market was killed and drained of growth and profits. (How many non alkaline batteries do you have in your house?)

Category creation is a variant of category killing. Like Uber and Airbnb, operating in this fashion results in creation of an entirely new category–a category of one–where you have the opportunity to become the dominant player. Creating new categories is at the core of building new, strong brands. Al Ries provides a number of classic examples in his AdAge post earlier this year.

Takeaway: Worried about being lost on the shelf? Don’t build “Uber for…” Build a category of one, where your brand is the category.