Taguchi proposed an alternative approach to traditional quality loss functions by arguing that not all values within specification limits are equal, and that quality loss includes costs beyond manufacturing like financial losses to society. Specifically, he proposed a quadratic loss function that calculates loss based on how close a quality characteristic is to its target value, rather than just whether it falls within limits. This recognizes that processes with less variation have lower quality costs. Taguchi's approach considered that quality loss extends beyond production to the customer, and argued for minimizing variability to improve quality and business results.