Margherita Caprara

My research focuses on marketing strategy, with an emphasis on innovation, machine learning, and gender dynamics. I have research that draws on various data types, from text to images, leveraging natural language processing and deep learning to explore key marketing insights. 

I’m Looking Through You: Automated Brand Audience Analysis with Instagram Images

This study presents brand audience analysis, a novel methodology that integrates deep learning-based computer vision with community detection analysis to capture consumer-brand interactions at scale. By detecting brand logos in Instagram images and combining this with hashtag-based community detection, our approach uncovers a richer, more comprehensive view of consumer habits, preferences, and lifestyles that go beyond the limited scope of explicit brand mentions. Leveraging the power of advanced Generative Pre-trained Transformer (GPT) models, we label and describe distinct consumer communities, providing marketers with actionable insights into both global brand themes and local cultural nuances. We showcase our methodology with Adidas, Nike, and Under Armour across eight global cities. This analysis demonstrates the utility of our methodology in identifying core brand associations and local variations, empowering marketers to craft targeted, context-aware strategies that are globally consistent yet locally relevant. This methodology redefines how brands can understand and engage with their diverse customer base in today’s dynamic digital landscape

WORKING PAPERS

• Caprara M. and Rubera G. “Innovating for Women? The Relationship Between Female Influence in the Top Management Team and Innovation Focus”
• Grossetti F., Cillo P., Caprara M., and Rubera G. “Assessing Entrepreneurial Activity through Deep Learning: A Geographic Analysis Using Instagram Data”