Essays on Technology-Enabled Platforms

This dissertation consists of three studies that examine dynamics on Business-to-Business (B2B) exchanges and crowd-based design contest platforms. In the first study, we examine trust formation and development in global buyer-supplier relationships. Trust affects all business relationships, especially global B2B transactions due to the distances between buyers and suppliers. We use information signaling theory to examine how buyers' trust in suppliers in global B2B commerce is affected by (1) their perceptions of the national integrity and legal structure of suppliers' country, and (2) third-party verifications of suppliers on B2B exchanges. A survey of global organizational buyers finds that perceptions of national integrity, legal structure, and supplier verifications are all positively related to buyers' trust. However, the number of prior transactions between buyers and suppliers moderates the impact of perceived legal structure on buyers' trust.;In the second study, we examine how selling and buying activity levels on B2B exchanges affect multi-homing buyers' preferences for exchanges. With the proliferation of B2B exchanges, many firms are multi-homing or using various competing platforms concurrently. We find that buyers' preferences are positively associated with selling levels on the platforms. However, buyers' preferences are non-monotonically related to buying levels on the platform. An initial increase in buying level has a positive effect on buyers' preferences. This effect may derive from the principle of social proof, where individual buyers observe and imitate other similar buyers' behaviors. As buying level continues to increases, it leads to greater competition among buyers on the platform. Consequently, buyers prefer less of this particular exchange and would participate more on the other exchange. We also find that the impacts of buying levels on buyers' preferences attenuate over time. Our results highlight the need to correctly model buyers' homing behavior, and show how market factors and social information conveyed by users on the platforms affects individual buyers' participation and the competition between B2B exchanges.;In the final study, we look at how advertisers can acquire impactful ad designs through crowd-based design contest platforms. Design contests allow advertisers to acquire a large number of designs that they can consider for their advertising campaigns. However, the large number of entries in these contests brings along a challenge in measuring design distinctiveness: The number of pairwise comparisons that is needed to determine distinctiveness may be non-trivial as it increases at a quadratic rate with the number of designs. To tackle this problem, we develop a novel model-based approach to efficiently measure design distinctiveness in design contests. We further find that ads can achieve better click-through performance when they have more distinctive designs.;We also examine how advertisers can influence designers to create distinctive work in design contests. Clients often provide examples of ad designs that they like in design projects. Using a randomized experiment that involved experienced graphic designers, we look at how clients-provided examples influence creative processes and design outcomes in design contests. Specifically, we examine how the quality and design variability of these examples affect designers' exploration for design concepts and their design submissions in the contests. We find that designers deviate less from client-provided examples when these examples are (i) of high quality, and (ii) relatively similar in their design concepts. We also find that design distinctiveness is lower when designers deviate less from the examples. The findings in this study have strategic implications on what clients and designers should do in design contests.