Online two-sided matching markets such as Q&A forums (e.g. StackOverflow, Quora) and online labour platforms (e.g. Upwork) critically rely on the ability to propose adequate matches based on imperfect knowledge of the two parties to be matched. This prompts the following question: Which matching recommendation algorithms can, in the presence of such uncertainty, lead to efficient platform operation? To answer this question, we develop a model of a task / server matching system. For this model, we give a necessary and sufficient condition for an incoming stream of tasks to be manageable by the system. We further identify a so-called back-pressure policy under which the throughput that the system can handle is optimized. We show that this policy achieves strictly larger throughput than a natural greedy policy. Finally, we validate our model and confirm our theoretical findings with experiments based on logs of Math.StackExchange, a StackOverflow forum dedicated to mathematics.