Telling Your Story Like AI? A Processing Fluency Theory Perspective of Generative AI-Like Entrepreneurial Narrative Style in Crowdfunding

Abstract

Entrepreneurial narrative style plays a pivotal role in the mobilization of resources, yet the recent emergence of generative artificial intelligence (AI) as a new writing subject calls for a reassessment of existing typologies. Drawing on natural language processing (NLP) literature, we conceptualize a novel “generative AI-like” entrepreneurial narrative style, characterized by low perplexity and burstiness—indicators commonly used in AI-generated text detection. Through the lens of processing fluency theory, we investigate the direct and indirect effects of this stylistic innovation on crowdfunding performance in reward-based platforms. Our study reveals that low perplexity and burstiness enhance processing fluency, which in turn positively influences crowdfunding outcomes, consistent with the hedonic fluency model. Moreover, deviations from this AI-like narrative style strengthen the negative effects of System 2 cognitive processing of dual-process theories—specifically evoking by cognitive language style and concept-based rhetoric—on crowdfunding performance. Our research contributes to the expansion of entrepreneurial narrative typologies and their roles in resource mobilization amid the transformation driven by generative AI, while also enriching the conceptualization and operationalization of processing fluency through the introduction of two novel metrics derived from AI-generated text detection. The findings offer novel insights for entrepreneurship scholars and crowdfunding practitioners navigating the evolving landscape shaped by AI technologies.

Keywords: generative AI, narrative style, crowdfunding performance, processing fluency, AI-generated text detection