Optimization Impossible?


Lowe, A.J.; Ridgway, K.   (2000, ASQ)  

Quality Progress    Vol. 33    No. 7
QICID: 13937    July 2000    pp. 59-64
List $10.00
Member $5.00

You will need to be signed in.
New to ASQ? Register here.

Article Abstract

A basic element in the implementation of a quality function deployment (QFD) project is combining customers' preferences to create an optimum product design. George A. Hazelrigg has questioned the validity of this approach, however, illustrating that it may make conflicting and irrational demands on product design. Hazelrigg bases his argument on the Impossibility Theorem, first described by Kenneth J. Arrow. Arrow showed that in most cases utilities cannot be used to express group preferences. To apply the theorem and to minimize its impact on QFD program success, practitioners must understand the situations in which the theorem applies and how it relates to the methods used to measure and combine customer preference data when creating a house of quality matrix.

When each individual has a rational set of preferences, combining these preferences to create an expressible overall group utility function presents a problem. Optimization, which is performed only in regard to a specific measure of value like dollars, is impossible in this situation, which is described as intransitive. Hazelrigg argues that the situation is not rare but is the norm. He also claims that the more detailed the descriptions of individual preferences are, the greater the chance of encountering this kind of problem. Unless utility functions are the same, overall group decisions may exhibit irrational or intransitive behavior.

Successful QFD projects depend on identifying the true requirements of a group of customers and translating these requirements into technical targets for the design of new products. Quantitative measurements of the relative importance and satisfaction obtained by the current product to meet customers' needs are also required. The main approaches for measuring and combining each individual customer's perception of the importance of individual needs use information collected through surveys or questionnaires. They include constant sum paired comparisons mirroring the preference options described in the Impossibility Theorem and calling for each customer to select a weighted value for each requirement directly.

Since there is a large number of successful QFD applications, the occurrence and impact of the problem may be significantly less than predicted by Hazelrigg. Focusing on specific market segments to prevent the occurrence of noncompatible preference groups can minimize the impact of the problem. Accurate segmentation can be achieved by evaluating customer responses through a tool like the voice of the customer table. The concept of mass customization can be applied to avoid the necessity of compiling group preferences from individual responses altogether.

Another implication of the Impossibility Theorem gives weight to the application of QFD. Promoting effective communication among all members can maximize the avoidance of intransitive decision-making inside product development teams. The Impossibility Theorem illustrates that grouping individual preference data can result in an overall sub-optimal product design and that effective communication within product development teams is essential.


Customer satisfaction (CS),Customer supplier relationships,Voice of the customer (VOC),Optimization,Quality Control Department,Customer surveys

Browse QIC Articles Chronologically:     Previous Article     Next Article

New Search

Featured advertisers

ASQ is a global community of people passionate about quality, who use the tools, their ideas and expertise to make our world work better. ASQ: The Global Voice of Quality.