I use SAS, a statistical processing software, for most of my work. While the software is very easy to learn, the way it is internally coded is ridiculously bad. It is another good case where the business team is good and knows what most of the users want. But the application is developed by software engineers who have absolutely no idea about those mathematical concepts. They also don’t seem to appreciate good software engineering practices. Extremely unhelpful error logs, bad interface and I can keep the crib going on.
SAS had nearly monopolized the statistical software market. Until R came. Initially R was used only by folks from academia. It had negligible market value once you exited academia. But things have changed a lot in recent years. Open source is a marvel. Lot of packages are available in R. SAS is a black-box. R has the distinct advantage of the analyst knowing what is going on in that package. It is a good competition.
I hope SAS improves their error logs and interface. Otherwise, it is going to lose the market leader position quite soon.