I vaguely recall research about process changes reliably boosting productivity temporarily but eventually reverting to the mean. Does anyone recall if that effect has a name or a canonical study?