Atheoretical measure purification may lead to construct deficient measures. The purpose of this paper is to provide a theoretically driven procedure for the development and empirical validation of symmetric component measures of multi-dimensional constructs. We place particular emphasis on establishing a formalized three-step procedure for achieving a posteriori content validity. We then apply the procedure on the development and empirical validation of two symmetrical component measures of market orientation; customer orientation and competitor orientation. Moreover, the results suggest that average variance extracted is particularly critical to reliability in the respecification of multi-indicator measures. In relation to this, the results also reveal possible deficiencies in using Cronbach alpha for establishing reliable and valid measures.