Innosuisse Projektschlussbericht Front-End Engine (FEE)
Zusammenfassung des Projektes und der Projektergebnisse
The front-end of innovation fundamentally determines the later innovation success, but is still one of the greatest areas of weakness of the innovation process (Brem and Voigt 2009, S. 353). This weakness is mainly induced by lack of information and a high degree of risk, both contributing to the fuzziness of this early phase.
A key role within the front-end play the early activities, which include the identification of a problem or opportunity and the accompanying screening and evaluation processes (Gassmann and Schweitzer 2014, S. 5). Opportunity identification and analysis set the arena “Where to play” and lay the foundation for all subsequent activities, which in contrast focus on elaborating the way “How to win”.
“The new product development process (NPD), at its core, is an information processing activity” (Brentani and Reid 2012, S. 71). During this process information about markets, technologies and competitors is translated to design new market offerings. Therefore, remediating the lack of information in the front-end serves as key lever for improvement, especially for opportunity analysis.
In this project, funded by Innosuisse, the authors have developed a functional prototype
(front-end engine – FEE) which leverages vertical crawling of Internet data and latent semantic analysis for semi-automated analysis of an opportunity. In a final step results are consolidated into KPIs. These KPIs are used to provide decision support on which opportunity to pursue, considering the firm-specific innovation strategy.
The conducted benchmarking between the FEE-prototype and human experts has proofed, that a time reduction of >90% can be achieved with semi-automated opportunity evaluation - compared to manual expert work. Along with the developed decision-support logic, based on the calculated KPIs, innovation managers can use freed-up capacity for downstream activities. The FEE-prototype enables a process transformation in the front-end: Formerly iterative activity processing can now be replaced by sequential process steps. Thus the FEE-prototype actively drives the “defuzzination” of the front-end.