Decentralized Operating Model

The decentralized operating model is the simplest operational structure. In this model, all service systems are managed independently by their respective dedicated application teams. These teams are not only responsible for the design, development, testing, deployment, and O&M of applications, but also for the deployment and O&M of the IaaS and PaaS resources required by the service systems. Additionally, they must ensure the security of service systems and manage the costs of cloud resources. The central IT team is responsible for developing unified IT standards and processes, issuing relevant documentation to service systems for reference and overseeing their implementation. However, the central IT team cannot enforce these standards and processes on the service systems. In this operating model, there is no requirement to establish a dedicated CCoE team.

Figure 1 How the decentralized operating model works

The decentralized operating model offers several advantages.

• High agility: Each business unit can quickly deploy and scale resources based on their own requirements, accelerating innovation.
• Service-oriented: The application teams have a better understanding of service requirements and can customize cloud solutions accordingly.
• Clear responsibilities: Each business unit manages their own cloud environment, making it easier to identify and resolve issues and optimize performance.

The decentralized operating model also has its disadvantages.

• Lack of consistency: Since each business unit independently deploys and maintains their required cloud environment, the absence of unified IT standards or security policies can lead to inconsistencies. This complicates management efforts.
• Increased costs: Without centralized coordination, duplicated efforts and resource waste can escalate cloud costs.
• Lack of an overall view: Difficulty in obtaining a comprehensive view of cloud resource usage limits strategic decision-making and optimization.

The decentralized operating model is ideal for innovative service systems requiring full control over cloud resource creation and O&M. These systems need to rapidly innovate and iterate based on evolving service requirements.