The cost is a consequence of a complex system out of hand, not a target metric. I would argue that paying $30k a year would be beneficial, _if_ the resulting system incentivized cost reduction. So the if is not in the cost but in the belief in the consequence of the change