In graduate school I discovered that, while I liked to learn mathematics, what I like to do was analyze data, and develop new ways of doing that.
While consulting I enjoy helping scientists and engineers design and analyze experiments, especially when a new approach is needed. I have been particularly interested in the statistical problems encountered by researchers in biology.
I believe that the importance of teaching statistics to undergraduates is underrated. We live in a statistically illiterate world. In our service courses we usually have just one chance to demonstrate to students the value of the accelerated learning which a statistical approach can give.
This applies even more strongly in the area of quality improvement, in which the great need is to teach a very large number of people how to use a few simple, highly effective techniques for applying, essentially, the scientific method.