Visual brand presence in digital social media; first funding period: Setup and Design of Networks by Providers.

每天在推特、Snapchat或WhatsApp等数字社交媒体上分享的图片超过30亿张。尽管DSM中越来越多的品牌标识，但这些图像数据尚未在科学营销研究中进行调查。应用程序包括社交媒体活动和跟踪与竞争相关的视觉品牌存在。在第二个资助期，将调查视觉品牌存在的营销相关性。应制定关于活动优化和绩效指标的程式化事实。A2项目是先前研究的自然延伸，该研究仅限于文本通信与音乐行业的实证应用。结果表明，企业传播的具体内容类别对网络需求、网络结构和网络传播的影响超过了传播量。考虑到与此同时的发展，本研究将在更多的行业中调查与品牌相关的视觉传达。机器学习算法可以实现这一点。它们能够以越来越高的精度检测物体和人脸，并且可以相对容易地训练以检测品牌标识。从经验上看，Twitter似乎是一种合适的调查媒介，因为它在公众可访问的图像传播方面已经达到了特别突出的地位。目前已经收集了各行业270多个不同品牌的850万条推文。在岗位层面，这些数据可以研究哪些因素有利于品牌标识在DSM中的传播（例如，标识的大小/位置、周围物体、面孔、情感、对比、相关文本）。谷歌的搜索查询、网站流量和基于调查的品牌跟踪在品牌层面上可用，以调查视觉品牌存在与品牌成功之间的关系。研究结果将为视觉品牌在DSM中的相关性提供第一个科学指示，并指出DSM传播方面的关键成功因素。

More than 3 billion images are shared every day on digital social media (DSM) such as Twitter, Snapchat or WhatsApp. These image data have not been investigated in scientific marketing research despite the growing number of brand logos in DSM. Applications include social media campaigns and tracking of visual brand presence relative to competition. In the second funding period, the marketing relevance of visual brand presence shall be investigated. Stylized facts on campaign optimization and performance indicators shall be developed. Project A2 is a natural extension of the previous research, which was limited to text communication with empirical applications in the music industry. According to the results, specific content categories of firm communication impact online demand, network structure and network communication over and above communication volume. Considering the development in the meantime, this research now investigates brand-related visual communication across more industries. Machine learning algorithms are available to accomplish this. These are capable of detecting objects and faces with increasing precision and can be relatively easily trained to detect brand logos. Empirically, Twitter appears an appropriate medium of investigation, since it has reached a particular prominence with regard to publicly accessible image communication. 8.5 million Tweets on more than 270 different brands of various industries have been collected so far. On a post level, these data allow studying which factors are conducive to the dissemination of brand logos in DSM (e.g., size/ position of logo, surrounding objects, faces, sentiments, contrasts, associated texts). Google search queries, website traffic and survey based brand trackings are available at the brand level to investigate the relationship between visual brand presence and brand success across time. The results will provide a first scientific indication on the relevance of visual brand presence in DSM and point out key success factors in terms of DSM dissemination.