Law #10: The Law of Division
14 Jun 2004

(This entry is part of a series I am writing on The 22 Immutable Laws of Marketing.)

The Law of Division observes that over time, a category tends to divide and become two or more categories.

A new market category starts out very broad.  For example, in the beginning of the automobile industry, the only category was "cars".  Over time, categories break up into smaller and more specialized subcategories.  Today, there are quite a few brands of car, each catering to a specialized niche.

This effect is an obvious and natural consequence of other laws.  Each company will try to setup a new category in which it can be #1.  Not all of these categories will end up becoming real, but some will.

This law is a good place to remind ourselves that Ries and Trout primarily consult for companies like Pepsi, McDonalds, and General Motors, not for small ISVs.  There is a bit of an impedance mismatch between their world and ours.  Those companies do business in mature industries selling mass market consumer products.  Those products are easily interchangeable.  I can switch from Pepsi to Coke with no major costs associated with the transition and deployment.  Categories split into subcategories over very slight differences in consumer preference.  Brand building is absolutely critical.  General Motors understands that some car buyers want to feel like they are buying something sporty, whereas others want to feel like they are buying something conservative.  So, they sell basically the same car under the Pontiac and the Buick name, managing each of these brands very carefully.  The underlying engineering is identical, but the message of these two brands is very different.

Software isn't usually like that.  When people are choosing between Dundas Chart and Chart FX, brand can be an issue, but it is certainly not the only issue.  Those two products are not identical under the hood.  They are not completely interchangeable.  There are significant differences in features.

So the Law of Division still matters, but it happens a lot less.  In software, categories don't divide quite as easily as in commodity markets.

But even though categories in software don't usually divide over matters of pure message, they do still divide, especially over matters of greater substance.  My own market is a decent example.  In the beginning, there was simply "source control".  Now, we've got several subcategories which came as the market divided over lines such as platform, process and integration.