Analyzing unstructured text, e.g., online reviews and social media, has already made a major impact, yet a vast array of publicly available, unstructured non-text data houses latent insight into consumers and markets. This article focuses on three specific types of such data: image, video, and audio. Many researchers see the potential in analyzing these data sources, going beyond text, but remain unsure about how to gain insights. We review prior research, give practical methodological advice, highlight relevant marketing questions, and suggest avenues for future exploration. Critically, we spotlight the machine learning capabilities of major platforms like AWS, GCP, and Azure, and how they are equipped to handle such data. By evaluating the performance of these platforms in tasks relevant to marketing managers, we aim to guide researchers in optimizing their methodological choices. Our study has significant managerial implications by identifying actionable procedures where abundant data beyond text could be utilized.

分析非结构化文本（例如在线评论和社交媒体）已经产生了重大影响，但大量公开可用的非结构化非文本数据蕴藏着对消费者和市场的潜在洞察。本文重点介绍此类数据的三种特定类型：图像、视频和音频。许多研究人员看到了分析这些数据源的潜力，超越文本，但仍不确定如何获得见解。我们回顾之前的研究，提供实用的方法建议，强调相关的营销问题，并提出未来探索的途径。至关重要的是，我们重点关注 AWS、GCP 和 Azure 等主要平台的机器学习功能，以及它们如何处理此类数据。通过评估这些平台在与营销经理相关的任务中的表现，我们的目标是指导研究人员优化他们的方法选择。我们的研究通过确定可利用文本以外的大量数据的可行程序，具有重大的管理意义。