There is a longstanding debate on the advantages of quasi-markets for placement services compared to their public deliverance. During 2009, the German Public Employment Service (PES) implemented a randomized field experiment to investigate if intensive services for hard-to-place unemployed individuals can be provided more effectively by such private providers or by PES in-house teams. Unemployed persons were assigned to intensive services for a period of eight months. This paper presents the first results of this experiment; the observation period covers 18 months after assignment. Initial in-house provision reduces accumulated days in unemployment by one to two months. Approximately two thirds of this effect is attributable to labor market withdrawals. The effect on the share of individuals in a given labor market status (unemployed, employed, and withdrawn) occurs mainly during the program period of eight months and disappears by the end of the observation period. We conclude that in a particular environment, the public provision of placement services can be as least as effective as contracting-out - a simple comparison of effectiveness might, however, be misleading.