The "Everything'S Different, Every Time" Innovation Management Problem: A Promising Model Development

This study reports on the testing of a promising approach for aiding decision-making during innovation. By focusing on the effects of risk/action dyads on success (the Risk/Action/Success (R/A/S) framework), and because perceived risks do appear repeatedly even though they emanate from differing contexts, the model offers an opportunity to learn from what worked best before. Using Artificial Neural Networks, this novel approach allows for generalisation and applicability of specific innovation management actions that are context specific. For academics, the proposed approach contributes to the risk-management literature by proposing a new paradigm for understanding and analysing innovation processes and identification of the most frequently occurring risks as seen by managers directly involved in continuous innovation. In addition, the model offers the capacity to use quantitative techniques to model the overlapping risks and actions during innovation-related decision-making. For practitioners, it can provide specific recommendations in the form of success-sorted lists of actions taken by other innovation managers that faced similar risks. This paper presents the theoretical and practical rationales underpinning this R/A/S framework and reports on the viability of this approach using pilot data.