I think R's users are mostly mathematicians in probability and statistics.
This is also the main advantage of R, that is, R has a library with rich mathematical functions, while other free scripts can't still compete it till now;only charged SPSS and MATLAB include such content; Python’s panda package still has a gap to reach it.
It could be said this is a niche market, after all, most users are unable to understand and do not need those profound knowledge about statistics and mathematics.
Day-to-day data analytics is based on structured data, its difficulty lies in the complexity of process steps, rather than esoteric operations, such as, compute rising stocks within 5 days, find users whose arrears have exceed three months, and so on.
Besides statistical function package, R is lackluster, cryptic in syntax, and so bad in performance; it is more likely that R will be replaced by Python.
For structured data computing, R has slightly superior set-style syntax and data frame object to Python, allows you to write less loops and shorter code (but obscure). But more advanced script is none but esProc. The data object esProc provides is far more powerful than data frame, and a set-style syntax is more nature, easy to understand; the performance to traverse file data is ten times higher than R.