The lure of social capital motivates startups to form in clusters with similar companies. However, having social capital is different from exploiting it, and there is conflicting research on the ultimate commercial success of cluster members. This work attempts to disambiguate the relationship between a startup's membership in a cluster and the startup's performance by modeling the availability of social capital separately from its use. Using the longitudinal Kauffman Firm Survey of 4928 companies founded in 2004 and the County Business Patterns from the United States Census Bureau, we compute a measure of relevant social capital available to a startup as the number of companies with the startup's 2-digit NAICS code in the startup's ZIP code, and the startup's use of social capital as collaborations that impact the startup's competitive advantage. We find that collaboration mediates the relationship between cluster density and firm revenue over its first eight years. This work suggests that the administrator of a critical mass of entrepreneurs, such as that of a business cluster or incubator, needs to promote the exploitation of its social capital and not just its accumulation.