This paper describes an approach to adaptation to users’ preferences realized by an interface agency. Using an informed negotiation technique, agents as bidders as well as contractors compete with each other to meet users’ preferences. By learning from indirect user feedback, the adjustment of internal credit vectors and the assignment of bidders that gained maximal credit in respect to the user’s actual preferences and the preceding session can be realized. In this way, user adaptation is achieved without accumulating explicit user models but by the use of implicit, distributed user models.