Extant literature characterizes the ties between buyers and sellers in online goods marketplaces–such as EBay or Amazon–as "arm's length" relationships. In such relationships, parties are effectively strangers, and exchange goods under the assumption that they will likely never again meet in the marketplace. The rapid rise of online service marketplaces, however, warrants a thorough examination into the applicability of these conclusions to buyer-seller relationships in the nascent markets. In this work-in-progress, we hypothesize that buyer activity in the marketplace will consist initially of "service encounters" whereby the uncommitted or inexperienced buyer gains exposure to a large number of potential suppliers. Increased experience in and reliance upon the market, we predict, will lead to a convergence towards "service relationships" as buyers seek trust, reduced risk, and close partnerships through repeated exchanges with a small set of suppliers. We intend to test this learning curve hypothesis using data collected from a leading online job auction site. The results will shed new light on the nature of buyer-seller relationships in online service-exchange markets.