Technovation 24 (2004) 553–561 www.elsevier.com/locate/technovation Complexity and learning behaviors in product innovation Ross Chapman a,* , Paul Hyland b a University of Western Sydney, InCITe Research Group, School of Management, Locked Bag 1797, Penrith South DC, NSW 1797, Australia b Central Queensland University, Faculty Business and Law, Bruce Highway, Rockhampton 4700, Australia Abstract Successful product innovation and the ability of companies to continuously improve their innovation processes are rapidly becom- ing essential requirements for competitive advantage and long-term growth in both manufacturing and service industries. It is now recognized that companies must develop innovation capabilities across all stages of the product development, manufacture, and distribution cycle. These Continuous Product Innovation (CPI) capabilities are closely associated with a company’s knowledge management systems and processes. Companies must develop mechanisms to continuously improve these capabilities over time. Using results of an international survey on CPI practices, sets of companies are identified by similarities in specific contingencies related to their complexity of product, process, technological, and customer interface. Differences between the learning behaviors found present in the company groups and in the levers used to develop and support these behaviors are identified and discussed. This paper also discusses appropriate mechanisms for firms with similar complexities, and some approaches they can use to improve their organizational learning and product innovation.  2002 Elsevier Ltd. All rights reserved. Keywords: Complexity; Product innovation; Knowledge management; NPD processes 1. Introduction Product Innovation is now accepted as a key require- ment for business success. It is also generally accepted that product innovation is a process within firms and, therefore, can be managed by manipulating the influ- ences at work (Johnston and Ireland, 1999). In the last fifteen to twenty years, companies have experienced considerable pressure to improve both the quality and speed of product innovation. One of the main conse- quences of this focus was the emergence of product innovation models almost totally focused on the manage- ment of the New Product Development (NPD) process. Integration among different phases of an NPD project and autonomy of the project team were considered syn- onymous with best practice product innovation. Concur- rent engineering was thought to represent a long-lasting paradigm for product innovation management. In the early 1990s, a new stream of studies emerged which enlarged these perspective. These studies demon- strated how focusing on single projects is not enough to * Corresponding author. Tel.: +61-2-46203245; fax: +61-2-266683. E-mail address: r.chapman@uws.edu.au (R. Chapman). 0166-4972/$ - see front matter  2002 Elsevier Ltd. All rights reserved. doi:10.1016/S0166-4972(02)00121-9 sustain competition. Success is also dependent on exploiting synergy among projects; for example, by fos- tering commonality and reuse of design solutions over time (Wheelwright and Clark, 1992; Meyer and Utterback, 1993). In this perspective, attention progress- ively shifts from single projects to a project family (Meyer and Utterback, 1993; Sanderson and Uzumeri, 1995) and to the process of learning and knowledge transfer and consolidation (Imai et al., 1988; Lynn et al., 1999; Bartezzaghi et al., 1997a). Many of these studies, however, still considered product innovation as occur- ring only within the boundaries of the product develop- ment process. Downstream phases in the product life cycle were still important for innovation but only as long as they represented valuable sources of information or constraints that should be anticipated and considered during development (Clark and Fujimoto, 1991). Recent evidence suggests that other phases in the product life cycle may actually represent additional opportunities for direct contribution in product innovation. This is a consequence of increasing pressure for more rapid pro- duct development and decreased time to market. Several companies, especially in rapidly shifting environments, purposely release to market products that are not fully optimized, followed by a rapid almost continuous stream