The main problem with SPSS is that it is an extremely poorly written piece of software from a technical perspective.
One symptom of that is that it is very bloated, slow, and takes an order of magnitude longer to start than any other application we have in use. The much more serious problem is, though, that you regularly experience issues that simply would not occur with other applications that are not alpha-level pre-releases, and that the app constantly feels like it is barely held together by virtual duck tape.
We have about 25 SPSS users in our department (about half of them on macOS) and the number of SPSS-related support requests dwarfs every other type of support request by far. Often, they are just really bizarre, random problems that typically only affect some, but not other users. Just to give you an idea of the type of problems, here are a few examples of issues we encountered over the course of the past couple years:
- The (local single user) license stops working after users connect to a new Wi-Fi network (e.g., at a conference, at a hotel, etc.).
- The submenu "File->New" only contains "Script", but no longer "Data", "Syntax", and "Output".
- SPSS crashes reliably every time you open a second data file.
- You can no longer run commands from the "Analyze" menu unless you keep an (empty) syntax window in the background.
- SPSS crashes when you try to open large output files that you created with a previous version of SPSS.
- Opening a syntax file (without disabling syntax coloring before in the preferences) takes a few minutes and editing is extremely sluggish. Opening the same file with the previous version of SPSS with syntax coloring turned on does not cause any performance issues.
- Overall a high number of crashes and hangs that you can't reliably reproduce.
Most depressing of all, even though IBM sells a new version of SPSS every year, there does not seem to be any improvement in the quality or reliability of the application. If anything, the current versions seem to actually be less stable than those released 5 years ago.
Fortunately, because of the poor reliability of SPSS, the exorbitant prices when you no longer qualify for academic discount, and the availability of much better (and often free/open source) alternatives, many in the social sciences are moving away from SPSS so that it is no longer the de-facto standard statistics application that most people use and that every undergrad student is taught. Learning R, for instance, may appear a bit more involved at first, but it will save you so many headaches over the long term.