In this study, we introduce visual totality of a crowdfunding pitch video which considers not only visual segments with human faces but also segments without human faces. Drawing from Emotions as Social Information (EASI) theory and expression theory, we analyze more than 4 million frames in 3184 Indiegogo rewards-based crowdfunding pitch videos using the ResNet 50 deep neural network. Results indicate that the impact of peak negative affective visual expression on funding performance is stronger than that of its positive counterpart for both segments with and without human faces. Additionally, the influence of peak negative affective visual expression from human faces is stronger in the first half (vs. the second half) of the pitch video. Further, we found a substitute moderating effect between the peak negative affective visual expression from segments with and without human faces on funding performance. We conducted an additional data collection to ascertain that pain points serve as the underlying mechanism through which negative affective visual expressions related to funding outcome. We discuss the theoretical and practical implications of our study to the crowdfunding literature and the broader research on entrepreneurial resource acquisition.

在这项研究中，我们介绍了众筹推介视频的视觉整体性，该视频不仅考虑了具有人脸的视觉片段，还考虑了没有人脸的视觉片段。借鉴情感作为社会信息 (EASI) 理论和表达理论，我们使用 ResNet 50 深度神经网络分析了 3184 个 Indiegogo 基于奖励的众筹推介视频中的超过 400 万帧。结果表明，对于有人脸和没有人脸的细分市场来说，峰值负面情感视觉表达对融资绩效的影响要强于正面情感视觉表达的影响。此外，在推介视频的前半部分（相对于下半部分），人脸的峰值负面情感视觉表达的影响更强。更远，我们发现有和没有人脸的片段的峰值负面情感视觉表达对融资绩效的替代调节作用。我们进行了额外的数据收集，以确定痛点作为潜在机制，通过该机制，负面情感视觉表达与融资结果相关。我们讨论了我们的研究对众筹文献和更广泛的创业资源获取研究的理论和实践意义。