I once had to explain to a manager of hundreds of people why reproducibility in engineering (and science) is important. Not good.