A firm’s ability to survive and prosper is often a function of its ability to design and develop new products that meet the needs of heterogeneous markets. The way in which a product is designed can have profound implications for product market structure and who is able to profit from an innovation, but despite this few industry studies have examined how and why product and industry architectures co-evolve and correspond across time. Notions of architectural co-evolution and correspondence are grounded in the modularity literature and assume a path towards increasing product modularity and industry specialisation. However, scholars have recently hinted that a reverse path towards increasing product and industry integration may be equally feasible. This research study contributes to the literature by proposing three stylised hybrid product and industry reintegration types that enhance our understanding of how and why reintegration may occur in product markets. Furthermore, the presence of a correspondence in the design characteristics between architectural layers (the so-called ‘mirroring hypothesis’) has also been suggested in the literature, such that product component design is often a blueprint for the way task, knowledge and firm boundaries are partitioned within a given product market. This research study finds that architectural correspondence is hard to sustain over time as firms often maintain a broader knowledge than task boundary for strategically important product components that offer differentiation opportunities or competitive advantage, contributing to the literature on contingencies that ‘mist the mirror’. Of particular interest to this research study is the UK personal pensions sector, a non-physical product, largely under-explored in the product modularity literature. By analysing the co-evolution and correspondence of a non-manufactured product over a 30-year period this research study breaks new ground. The research study makes use of a retrospective longitudinal research design, based upon semi-structured interviews with a purposive sample of 31 key personnel. The interview data was subject to a combination of matrix and template analysis.
|Publication status||Accepted/In press - 21 Sep 2016|