[ai]deation: GenAI-Based Collaborative Service Innovation

Abstract

Many new services fail within the first year as businesses often launch services without a deep understanding of stakeholder needs. Recent advancements in generative AI, particularly Large Language Models (LLMs), have enabled the automation of creative tasks. In this work, we focus on the ideation phase of service development, aiming to integrate stakeholder perspectives early in the service development process through a novel human-AI collaborative role-based ideation method. Based on this method, we developed a software tool and an evaluation concept to support this approach. By tailoring the LLM’s roles to meet specific stakeholder needs and incorporating diverse perspectives from the start, we encourage a more effective service development process, reduce the risk of poorly designed services, and facilitate coordination and understanding among stakeholders.