Regression methods, logistic or normal distribution based methods are meat and potato of analytical methods.
Every one is taught and software are awash with procedures and some times even extending them to some additional functionality. In the aisles of SAS and SPSS developers and users heard the story of how the whole company was built based on these methods in the early stages of development of the company.
However, statisticians, economists, social and behavioral scientists, and recently computer scientists have introduced a host of other methods, which are not pervasive yet among analysts, though significant amount of solution methods are available.
One such methods is Latent Variable Models.
There is one abstraction of successful analytics practices that points to latent variable models as the ultimate creator of knowledge nuggets in social sciences and I would add to that that it is true in business too, especially if you are looking for an advanced level of CRM where the intelligence in consumer inputs can be coming in so many signal variations that the intelligence can be systematically extracted using latent variable analysis. This is a common opportunity, in general, in sample surveys for attitudinal and behavioral aspects of consumers.
A simple factor analysis is the beginning of this area.