I generally agree with Al Ries. His insights into positioning have been remarkable over the years. He is a fanatic about NOT allowing brand or line extensions. His position is they usually (always) result in poorer performance than if you created a new brand.
In a recent article he took the position that Amazon.com would have been better off creating multiple brands for their various offerings. He even argues that if Amazon had started an online shoe store they could have not spent $1.3B buying Zappos. His argument is that while Amazon has been successful, their profits to date are poor and that is at least partly (largely?) due to line/brand extension. In the article he goes on to discuss other companies with similar failings (as he sees it) including Xerox and IBM.
His arguments about Xerox are correct in my view. His IBM argument, especially when comparing them to HP (which is also a big brand/line extender) does not hold water. IBM’s errors were strategic in nature not brand extension related. And his Amazon argument is flawed because Amazon is not a product company it is a retailer. Wal-Mart and others are quite successful selling a WIDE range of “stuff.” And Home Depot and Lowes are successful with a more focused strategy.
If you view Amazon as the easiest online store to buy from (which is my view), then who cares what they sell as long as they maintain the expected Amazon experience. That is brand integrity. He furthers my argument by citing Kindle as a new brand Amazon created that is doing better than Sony’s e-reader. And the Kindle is a product and Amazon is a store. QED.