Centralization and formalization have previously been employed in research designs investigating organizational innovation. The results, however, have been mixed. Three arguments are raised toward explaining this inconsistency. First, innovation is a multi-phased process in which the influence of centralization and formalization could be expected to differ. Second, as innovations may vary with regard to their compatibility to individuals in organizations, the expected influence of centralization and formalization may also vary. Third, organizational innovations may be directed toward particular vested interests, such as the technical or administrative core. Accordingly, the influence of centralization and formalization may also vary. This paper reports on a study which examines the influence of centralization and formalization on organizational innovation given the arguments given above. The effectiveness of employing structural overlays to overcome the conflicting forces often arising between organizational processes and particular organizational behaviors—such as innovative behavior—is investigated as well. Managers of 49 software development groups responded to a questionnaire assessing the influences of centralization and formalization on the initiation, adoption, and implementation of six modern software practices (MSP) within their software groups. Three of the MSP were technical in nature and three were administrative. The technical MSP were viewed as being incompatible to the technical core. Two structural overlays were investigated regarding their ability to improve the diffusion process: the existence of a formal body responsible for identifying MSP and introducing them into the software group; and, the assignment of computer specialists to projects as “set” teams.