Social networks provide entrepreneurs with a cost-effective way to acquire resources from people they know personally. In the current research, I develop and test a relational model of resource acquisition, which demonstrates the predictive role of universal social perceptions in resource owners’ decisions to provide resources (i.e., customer referrals) to entrepreneurs in their social networks. I draw on social exchange theory to explain that network members provide resources to dyadic partners because they expect the recipient will eventually reciprocate in kind. I then integrate social cognition research to explain how network partners’ expectations of reciprocity may be influenced by judgments of each other’s willingness to engage in reciprocity, which is manifest in perceptions of warmth attributes, such as liking. I test the model by analyzing data from over 4,000 entrepreneurial dyads in 37 bounded social networks. Findings provide clear evidence that the provision of customer referrals is a dyadic phenomenon. Further examination of isolated relationship effects of unique liking confirms the importance of interpersonal dynamics when resource acquisition is required for entrepreneurial success.