One of the first commercial software packages I developed was a DSS to help schedule orders in a paper mill.

The system was designed to help the person who scheduled orders decide when to schedule particular orders to reduce waste in the mill. This was a very challenging problem—so challenging, in fact, that it usually took the scheduler a year or two to really learn how to do the job well.

The software was tested by a variety of paper mills over the years and always reduced the amount of waste, usually by about 25 percent but sometimes by 75 percent when a scheduler new to the job was doing the scheduling.

Although we ended up selling the package to most paper mills that tested it, we usually encountered significant resistance from the person doing the scheduling

(except when the scheduler was new to the job and the package clearly saved a significant amount). At the time, I assumed that the resistance to the system was related to the amount of waste reduced: the less waste reduced, the more resistance because the payback analysis showed it took longer to pay for the software.

Questions 1. What is another possible explanation for the different levels of resistance encountered at different mills?
2. How might this be addressed?