What makes successful equity crowdfunding campaigns? A machine learning analysis of information cues - Empirical study on machine-learning-based information cues

Aug 5, 2026

This study integrates signaling theory with information cue theory to evaluate how different information cues predict equity crowdfunding success. Using data from 144 campaigns on China’s Dreammove platform during 2014–2019, five machine learning models estimate cue diagnosticity. Results show quantitative cues provide the strongest prediction of funding outcomes. Narrative cues remain informative but less powerful, while visual cues show limited predictive value. SHAP analysis highlights team size, ambitious targets, concise accessible narratives, and modest visual elements.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages31

Reading Time1 h 18 min

AccessFree to download

Introduction

Financing constraints for MSMEs

Crowdfunding models and equity crowdfunding

Research design and method

Data and platform context

Machine learning models and cue diagnosticity

SHAP interpretability

Findings

Quantitative vs narrative vs visual cues

Key campaign features associated with success

The European Journal of Finance ISSN: 1351-847X (Print) 1466-4364 (Online) Journal [homepage: ](homepage: www.tandfonline.com/journals/rejf20)[www.tandfonline.com/journals/rejf20](homepage: www.tandfonline.com/journals/rejf20) What makes successful equity crowdfunding campaigns? A machine learning analysis of information cues Jinjuan Yang, Yan Zeng, Jiayuan Xin, Zhibin Lin & Xue Chen To cite this article: Jinjuan Yang, Yan Zeng, Jiayuan Xin, Zhibin Lin & Xue Chen (11 Nov 2025): What makes successful equity crowdfunding campaigns? A machine learning analysis of information cues, The European Journal of Finance, DOI: 10.1080/1351847X.2025.2585952 To link to this article: [https://doi.org/10.1080/1351847X.2025.2585952](https://doi.org/10.1080/1351847X.2025.2585952) © 2025 The Author(s) . Published by Informa UK Limited, trading as Taylor & Francis Group. | View supplementary material | | --- | | Published online: 11 Nov 2025. | | Submit your article to this journal | | Article views: 28 | | View related articles | | View Crossmark data | Full Terms & Conditions of access and use can be found at [https://www.tandfonline.com/action/journalInformation?journalCode=rejf20](https://www.tandfonline.com/action/journalInformation?journalCode=rejf20) THE EUROPEAN JOURNAL OF FINANCE [https://doi.org/10.1080/1351847X.2025.2585952](https://doi.org/10.1080/1351847X.2025.2585952) What makes successful equity crowdfunding campaigns? A machine learning analysis of information cues Jinjuan Yanga, Yan Zengb, Jiayuan Xinc, Zhibin Lind and Xue Chena a College of Management, Shenzhen University, Shenzhen, People’s Republic of China;b Faculty of Business and Law, University ofthe West of England, Bristol, UK; c Newcastle University Business School, Newcastle University, Newcastle upon Tyne, UK; d Durham University Business School, Durham University, Durham, UK ABSTRACT This study innovatively integrates signaling theory with information cue theory to examine the diagnosticity of various information cues in predicting equity crowdfunding success. Analyzing data from 144 campaigns on China’s Dreammove platform (2014-2019) through five machine learning models, we find that quantitative information cues are most predictive of funding success. Narrative cues also hold diagnostic value, albeit to a lesser extent, while visual cues are limited in their predictive capacity. Employing Shapley Additive exPlanations (SHAP) interpretability, we identify larger entrepreneurial teams, ambitious funding targets, concise, accessible narratives, and modest visual elements like youthfulness and clarity enhance campaign appeal. This research advances the understanding of how investors process information in equity crowdfunding, offering new insights by moving beyond single-dimensional analyses and providing practical guidance for entrepreneurs and investors in emerging crowdfunding markets. ARTICLE HISTORY Received 9 February 2024 Accepted 7 October 2025 KEYWORDS Machine learning; prediction accuracy; cue diagnosticity; equity crowdfunding 1. Introduction Micro, small, and medium-sized enterprises (MSMEs) often encounter significant barriers in obtaining loans from traditional banks, a situation that hampers their development (Lopez-Gracia andAybar-Arias 2000). While private equity (PE) and venture capital (VC) have emerged as viable alternatives, providing substantial financial and social resources to startups, these options are typically limited by the involvement of a relatively small number of professional investors (Blaseg, Cumming, and Koetter 2021) . Initial public offerings (IPOs), another traditional route for raising public equity, require companies to meet certain scale requirements and tend to favor institutional investors in share allocation (Aggarwal, Prabhala, and Puri 2002; Cumming, Meoli, and Vismara 2019a) . In contrast, crowdfunding platforms have emerged as a popular and inclusive means for entrepre

FAQs

Which information cues are most predictive of equity crowdfunding success?

Quantitative information cues are the most predictive of funding success, followed by narrative cues, while visual cues have limited predictive capacity.

How are the study’s models used to evaluate cue diagnosticity?

Five machine learning models analyze 144 Dreammove campaigns (2014–2019) to assess how different cues relate to funding outcomes, and SHAP explains feature contributions.

What campaign characteristics are identified as enhancing campaign appeal?

SHAP indicates larger entrepreneurial teams, ambitious funding targets, concise accessible narratives, and modest visual elements such as youthfulness and clarity can improve appeal.

Aug 5, 2026

Upload by Genevieve