[sgift]

Not GP, but usually things are missing or not up to date. Also, people are really bad at defining what exactly is a "good" result. When we introduce scheduling at customers it can take twenty to thirty iterations until we have a good scoring function with all constraints. And then the data needs to be kept up to date or it gets slowly worse over time (requirements change, data rot). This often leads to people saying "the system is bad, I'll just do it by hand" instead of investing the time to find the underlying issue. Especially if it happens gradually. First you rework one schedule, then two and so on. Or if they are not able to enact a change of the system. In many companies the people who do the schedule don't know who to tell that the system is getting worse or get ignored. And they have a job to do after all ..

Stereotypical discussion with customers after a "bad" scheduling result: "Hey, why did your system not schedule Dave to do any of these jobs? All others are at 100% already, but he has nothing to do." "The job all need a Fubar cert, Dave doesn't have one." "Of course he has a Fubar cert." "Not according to the system." "Oh, we probably forgot it, because all schedulers know it anyway."