Which can be done, but it is a long process. Most attempts only get a fraction of the big brand, and the exceptions generally have more to do with the failings of the big brand than the competition. Once people find something they like, as long as that thing doesn't do something stupid they won't be in a hurry to look at the competition in general. If there is any switching cost they are even less interested in trying something else. Which means that the not number one competitors need to be perfect in everything - which is hard when they are not getting as much revenue to begin with, and thus cannot afford to try out any seemingly good ideas that turn out bad only after you try them...