Despite the widespread use of the terms 'failure' and 'success' within social networking site (SNS) literature, determining benchmarks of either is a complex undertaking. Success and failure both are often identified according to the presence or absence of context-specific 'markers.' 'Snapshot' metrics such as network growth or activity are pointed to as indicators of a network's overall success. However, such reductive approaches oversimplify the underlying social processes. This paper presents a review of current literature, and argues for the conceptualisation of failure at the macro-level as the constraint of network actors, regarding 'success' and 'failure' as recurrent processes in the lifecycle of a social networking site rather than outcomes. It is hoped that theorising failure independent of specific markers will provide the basis for future development of predictive and analytical models to identify instances of failure in SNS case studies.