Drill down into this research paper and you will find an interesting table of 27 problems that might be present in a well-organized software development company. Here are the top concerns: bottlenecks, estimation, late detection of defects, redundant steps, and automatically tailoring the process to the project. The items are not unexpected, but may help other managers, researchers, and toolsmiths to have a checklist.
The organization is a large mature software development company. It follows the capability maturity model integration (CMMI) program from Carnegie Mellon University. This defines maturity levels for processes: initial, managed, defined, quantitatively managed, and optimizing. The company is working on improving its well-defined process by using quantitive data gathered by software development tools.
I wonder about the next steps in this meta-project. How will the language in the table be translated into commands for process mining tools? Further, I once had the painful experience of mining industrial logs and producing ridiculous results. Will this happen here? Third, this evolutionary procedure of small changes in the process is a kind of hill climbing that can lead a system to a merely local optimum. Perhaps a leaner and more agile organization might be better?