Beyond Platforms -- Growing Distributed Transaction Networks for Digital Commerce

Journal: arXiv
Published Date:

Abstract

Many of today's IT infrastructures are proprietary platforms, like WhatsApp or Amazon. In some domains, like healthcare or finance, governments often take a strong regulatory role or even own the infrastructure. However, the biggest IT Infrastructure, the Internet itself, is run, evolved and governed in a cooperative manner. Decentralised architectures provide a number of advantages: They are potentially more inclusive for small players; more resilient in case of adversarial events, and seem to generate more innovation. However, we do not have much knowledge on how to evolve, adapt and govern decentralised infrastructures. This article reports empirical research on the development and governance of the Beckn Protocol, a protocol for decentralised transactions, and the successful development of domain-specific adaptations, their implementation and scaling. It explores how the architecture and governance support local innovation for specific business domains and how the domain-specific innovations and need feedback into the evolution of the protocol itself. The research applied a case study approach, combining interviews, document and code analysis. The article shows the possibility of such a decentralised approach to IT Infrastructures. It identifies a number of generativity mechanisms, socio-technical arrangements of the architecture, community support and governance that support adoption, innovation, and scaling it. It emphasises the governance of both the evolution of the open source specifications and software and how this relates to the governance of the conduct of network participants in operational networks. Finally, it emphasises the importance of feedback loops to both provide input for technical evolution and to recognise misconduct and develop means to address it.

Authors

  • Yvonne Dittrich
  • Kim Peiter Jørgensen
  • Ravi Prakash
  • Willard Rafnsson
  • Jonas Kastberg Hinrichsen