This paper considers a resource allocation mechanism that utilizes a profit-maximizing auctioneer/matchmaker in the Kelso–Crawford (1982) (many-to-one) assignment problem. We consider general and simple (individualized price) message spaces for firmsʼ reports following Milgrom (2010). We show that in the simple message space, (i) the matchmakerʼs profit is always zero and an acceptable assignment is achieved in every Nash equilibrium, and (ii) the sets of stable assignments and strong Nash equilibria are equivalent. By contrast, in the general message space, the matchmaker may make a positive profit in a Nash equilibrium. This shows that restricting message space not only reduces the information requirement but also improves resource allocation.