Real-time collaborative editing systems such as Google Drive are increasingly common. However, no prior work questioned the maximum acceptable delay for real-time collaboration or the efficacy of com-pensatory strategies. In this study we examine the performance conse-quences of simulated network delay on an artificial collaborative doc-ument editing task with a time constant and metrics for process and outcome suitable for experimental study. Results suggest that strategy influences task outcome at least as much as delay in the distribution of work in progress. However, a paradoxical interaction between delay and strategy emerged, in which the more generally effective, but highly coupled strategy was also more sensitive to delay.