Numerous reward system practices are adopted in software development companies in order to motivate their developers to perform at best level and align the management and developer's interest. However, due to lack of a central mechanism for team formation on crowdsourcing-based software development platform, it is difficult for managers to adopt effective reward system strategies in order to align the developer's interest. In order to address this issue, we exploit an approach, to empirically investigate the existing reward system practices, adopted on the crowdsourcing-based software development platforms, to motivate their developers and their perception. Subsequently, we implement a crawler to mine the characteristics of completed tasks and related reward information from the TopCoder platform of Tech Platform Inc (TPI). The promising results suggest the applicability of the proposed approach in order, 1) to investigate the reward system practices across the crowdsourcing-based platforms and, 2) to help the managers in formulating and implementing an effective reward system to incentivize their developers.