I also use https://shifton.com/shift-scheduling , and, frankly, I have encountered the same problems. The system is certainly convenient, but automatic scheduling, especially if you have a large team, can sometimes go wrong. Why does this happen? Firstly, the shifton algorithm relies on pre-set parameters - minimum and maximum shifts, as well as employees' working hours. But here lies the main point: the system does not always perfectly understand individual preferences, and does not take into account the nuances that may arise in practice, for example, the physical activity of employees or their desire to work on certain days. When I first started using this system, I thought that automation would solve all the problems, but then I realized that additional settings need to be made to correctly distribute shifts. Shifton allows you to set parameters such as the minimum and maximum number of shifts for each employee, as well as their availability (for example, someone cannot work night shifts, while someone, on the contrary, prefers them). However, even when setting these parameters, the system does not always distribute shifts correctly. Here's what I did to improve the situation. First, I started using a combined approach - automatic scheduling, but with subsequent manual adjustments. At the beginning of the week, Shifton generates a suggested schedule, but I always check it and make changes if I see that someone has too many shifts or they are uneven. For example, if I have an employee who works a lot of night shifts, I try not to give him an extra shift at night, even if the system rules allow it.

I also use Shifton, and I agree that relying entirely on automatic shift distribution is not always a good idea. However, if you set up the system correctly and take into account all the characteristics of the employees, you can significantly simplify the scheduling process. I also add additional parameters to the settings, such as the number of days off, rest time after night shifts, etc.