In this paper, we argue that corruption research can benefit from studying corrupt transactions as a particular form of social interaction. We showcase the usefulness of a theoretical focus on social interaction by investigating online user reports on the website Frontdesktip.com. Through this focus, we can observe users sharing experiences and tips on the best ways of bribing hotel clerks in Las Vegas for attaining room upgrades and other complimentary extras. We employ a logistic regression analysis to examine what factors (...) influence the “successful” performance of this bribery practice. Our study makes a twofold contribution to existing research on corruption. First, on the theoretical level, we show that the typified and scripted character of social interactions can help explain the occurrence of corrupt transactions. Second, on a methodological level, our study showcases online self-reports as a useful data source to study corrupt transactions in an unobtrusive way. (shrink)