Narrating optimal distinctiveness: A machine learning perspective

Many new ventures fail to successfully establish themselves in the market or survive because they do not attract sufficiently positive evaluations from critical audiences. To ensure their survival, new ventures primarily depend on consumers, which they must convince in their multi-faceted roles of either investors, buyers, or users. However, how can new ventures try to raise their appeal to consumer audiences? One central and recent concept that aims to answer this question is optimal distinctiveness which postulates the need to be “as distinct as legitimately possible.” Entrepreneurial narratives are considered a primary means of achieving such optimal distinctiveness. Despite this widespread relevance, existing research provides entrepreneurs with little clarity on how and when they can strategically rely on different modes of narratives to address consumer audiences.

This thesis offers a more fine-grained perspective on how narratives can achieve optimal distinctiveness. It showcases by which mode entrepreneurs narrate their new ventures' distinctiveness, who and against what reference levels evaluates them, influence what level of distinctiveness is optimal. As a theoretical guide, this thesis builds on existing studies on strategic differentiation and entrepreneurial storytelling as a starting point. It extends these with insights from institutional logic, sensory marketing, and organizational learning literature. Using state-of-the-art machine learning-based natural language processing methods such as doc2vec, image recognition, and speech recognition; it also provides a methodological contribution that helps identify the underlying meanings in textual, visual, and auditory narratives as critical carriers of narrative distinctiveness.

Utilizing three different empirical studies, this thesis has essential contributions that add to the literature on optimal distinctiveness and entrepreneurial narratives. First, it highlights the increasing heterogeneity within consumer audiences who differ in their expectations for distinctiveness, depending on their role as either investors, buyers, or users. Furthermore, it shows how new ventures can deploy different narrative modes they strategically benefit from. It also outlines crucial contextual factors of reference levels' multilevel and dynamic nature that shape consumer audiences' evaluative processes.

For management practice, this thesis has important implications for entrepreneurs of new ventures who use narratives to appeal to consumer audiences, especially on online platforms. The results may serve as managerial guidelines that help entrepreneurs to decide on the right way and mode to narrate their products and new ventures to consumer audiences in various roles and online settings.