Case Study: Romance Scams

This publication presents a multi-chapter examination of social engineering and related scams, with Case Study: Romance Scams as one of several focal points. It frames the study as part of a broader effort to measure, understand, and counter online fraud, describing tangible data-gathering and analysis approaches used across several scam types. The work situates romance scams within a larger taxonomy that includes sales, rental, business email compromise, and other categories, and it emphasizes the need for empirical measurement, pattern discovery, and intervention points rather than generic warnings. Across the volume, the authors describe a common methodological backbone: automated data collection, pattern clustering, and analysis of attacker infrastructure and monetization. Notably, the sales and rental case studies rely on large-scale crawling of classified-ad platforms, automated honeypot interactions, and conversational engines to elicit and study scam responses. Findings from these chapters highlight that a small number of scam groups account for a large share of activity, that scams often rely on cross-site cloning and fake payment schemes, and that a sizable portion of scammers originate or route through certain geographic clusters. The work also documents gaps in existing defenses and points to concrete levers for disruption, such as targeting prolific scam groups, improving payment-based monitoring, and enhancing brand- and storyline–based detection techniques. Limitations in the presented material include reliance on specific platforms (notably Craigslist-like ecosystems) for measurement and the potential non-generalizability of certain findings to all romance scams or other online fraud domains. The volume argues for continued, methodical measurement and targeted intervention as essential components of fraud prevention, grounded in observed attacker behavior rather than generic advisories. The Romance Scams case study contributes to this empirically anchored, domain-spanning narrative.