Abstract

Various methods have been proposed to identify promising AI use cases. However, it is not known if these methods work under real business conditions (e.g., limited time and personnel for method execution). We address this issue by taking an alternative approach to method development: Action Design Research at Germany’s leading e-commerce company “OTTO”. We co-designed our method with practitioners through iterative cycles of building, organizational intervention, and evaluation. Our method synergizes Design Thinking and Fit- Viability Theory to identify AI use cases that are relevant and realistic for the business. We make two contributions. First, we provide a method that accounts for resource constraints faced in business practice. Practitioners can use the method “as is” or adapt it for their purposes. Second, we abstract our learnings into four principles for identifying promising AI use cases under real business conditions: Human-Centered Technorealism, Co-Creative Pragmatism, Adaptive Standardization, and Innovative Continuity.

Download

Share

COinS