Value of Information

The idea that measuring something is only worth doing if reducing that uncertainty would actually change a decision, and by enough to justify the cost. Most variables aren't worth measuring; a few are worth a great deal.