In a traditional relationship, a salesperson is the sole gatekeeper in the supplier’s relationship with a buying firm. Supplier integration (SI) suggests that each domain expert (e.g., engineering) of a buying firm should directly communicate with the supplying firm personnel, without passing through the traditional boundary spanner—a salesperson. Existing literature argues that such a multichannel relationship generates significant degrees of benefit (e.g., better product design, product innovation). However, SI may fail unless the salesperson accepts the disintermediated communication. The multichannel communication structure of SI may limit his/her role of the sole gatekeeper thereby causing his/her behavioral constraints. This dissertation aims to extend the existing SI literature by understanding a multichannel relationship from a salesperson’s perspective. This study understands how the work routine of a salesperson changes under a multichannel relationship, especially when an engineer of his/her company can also directly communicate with the buying firm. With the aid of some in-depth interviews with eight salespersons in a display industry, and with an inductive research approach, we have developed several propositions. These explain how SI changes a salesperson’s work characteristics and in what way such changes might affect his/her behavior. Based on these propositions, a set of testable hypotheses is established for an empirical study. These hypotheses are tested using (1) the survey data from the salespersons, and (2) the performance evaluation data from a manufacturer. The empirical study tests how SI affects an engineer’s and a salesperson’s behaviors (i.e., an engineer’s opportunism, his/her inadvertent benevolence, and a salesperson’s barricading behavior). Our results explain that SI triggers an engineer’s inadvertent benevolence—an engineer’s willingness to accommodate a buyer’s request without proper consideration for the consequences of the accommodation—which in turn causes a salesperson’s barricading behaviors to block SI. The barricading behaviors damage the supplier’s performance. For the implications, this dissertation addresses the root cause of SI failure, which might occur due to traditional boundary spanners (salespersons). Also, this research explains that benevolence—which is essential for external collaboration—could cause internal behavioral constraints that damage the external collaboration. This means that SI causes internal behavioral constraints, which paradoxically, damage SI.