Background: The phenomenon of cross-channel return evolved through the increasing online presence of retailers, and the inability to evaluate products before online purchases. It has become essential to introduce cross-channel returns to manage the increasing online returns by taking advantage of retailers' physical stores. However, by introducing cross-channel returns, options for internal return processes increased. To perform cross-channel returns, actors need to manage serval presumptions of the system and modify behavior after observed changes. This results in a complex interplay among actors that retailers must recognize, yet understanding the behavior and interaction in the cross-channel return setting is insufficient. This multi-case study includes retailers operating cross-channel and explores the phenomena of cross-channel return through the lens of complex adaptive systems (CAS) internal mechanisms.

Purpose: The study explores CAS theory as a new theoretical perspective for managing and limiting the internal complexity within cross-channel returns. It aims to provide the research field with a further understanding of retailers' internal complexity by exploring the behavior and interaction among actors in cross-channel returns.

Method: A exploratory multiple-case research design was used to fulfill the purpose of this qualitative research. The data collection was conducted through semi-structured, in-depth interviews. Twelve interviews were conducted with employees in five Swedish retailers that operate cross-channel returns. The data analysis for this study was established by an abductive approach through systematic combing.

Conclusion: This study found that various behavior and interactions among actors affect the individual and system differently within cross-channel returns. By applying the theoretical frame of reference of CAS’s internal mechanism to cross-channel returns, empirical evidence was found that behaviors and interactions of agents either increase or decrease the emergence and self-organization on different levels. This study has developed a conceptualized framework visualizing the underlying internal mechanism of CAS and its dynamic relations in the phenomena of cross-channel returns. This study extends prior CAS research by studying how autonomy and control affect mental schemas and the internal complexity within a system with high connectivity such as cross-channel returns.