During the last decade telecommunications market has become a very competitive environment, with a great many vendors, carriers, and services provided. The Intermediaries are one of the main players in the telecom network market, and they may acquire network capacity from a set of telecommunication Backbone Providers that offer different pricing and quality of service schemes and sell purchased capacity to end-users in order to make a profit. Unfortunately, the bandwidth demands of customers are not known in advance. Therefore, Intermediaries may end up having either slack capacity or unsatisfied customer demands. In this paper, to overcome these problems, a two-stage stochastic integer programming approach is proposed. The performance of suggested method is validated on several randomly generated scenarios. In addition, the effects of different bandwidth demand distributions are examined by using well-known measures such as Expected Value of Perfect Information (EVPI) and Value of Stochastic Solution (VSS). The robustness of methodology is evaluated by carrying out sensitivity analysis on different problem parameters.