Skip to Content

Three proven best practices for creating a center of excellence for RPA

Capgemini
2019-12-04

Robotic process automation (RPA) isn’t a one-and-done proposition. Instead, it’s a long-term, ongoing initiative aimed at helping organizations evolve strategically and increase business value. It’s also not the exclusive domain of a single group or department. That’s where a Center of Excellence (CoE) dedicated to RPA can help. A CoE can help an organization redistribute accumulated knowledge and resources across future automation deployments. It acts as the go-to source of expertise that RPA’s wide-scale transformation will require.

But there’s more to creating an effective CoE than just scheduling regular meetings among the organization’s RPA technologists. Here’s a quick look at some of the proven best practices that leading RPA adopters have used to get their CoEs up and running.

  1. Assemble a cross-functional team
    Because RPA requires significant technical, process, and management expertise, a CoE should involve a wide range of specialists from throughout the enterprise. Key roles and their responsibilities can include:
  • Sponsor – Establish RPA as an enterprise-wide strategic priority and underwrite corporate resources
  • Champion – Evangelize and drive RPA adoption across the organization
  • Change manager – Create the change and communication plan aligned to project deliverables
  • Business analyst – Create the process definitions and process maps used for automation
  • Solution architect – Define the architecture of the RPA solution and oversee its end-to-end development
  • Developer – Design, develop, and test automation workflows and support the implementation of the RPA solution
  • Infrastructure engineer – Oversee the infrastructure support for server installations and troubleshooting
  • Supervisor – Manage, orchestrate, and control the software robots as part of the operational environment
  • Service support – Act as the first line of assistance for the RPA solution during deployment
  1. Define a governance model
    So, that’s a quick summary of the roles and responsibilities that constitute a robotic CoE. The next priority is to define a governance model that will help the team identify opportunities and prioritize automation activities. The governance model provides the guidelines and templates for the assessment, design, development, and deployment of software robots, managing the demand pipeline. It helps ensure good collaboration and communication between units.
    One of the key questions regarding CoE governance is whether to adopt a centralized, decentralized, or hybrid CoE model. Here’s a quick summary of each:
  • Decentralized – In the decentralized or federated CoE model, RPA capabilities are spread across the organization. The CoE runs within separate business units (BUs), which oversee their own prioritization, assessment, and development of RPA processes. They also conduct their own operations.
    However, without centralized control, this model can be difficult to scale and it can complicate collaboration with the IT team. Different units will build expertise at different rates, and the lack of a shared platform can drive up costs. It also results in a lack of standardization across BUs that makes it difficult to scale.
  • Centralized – With the centralized model, a single CoE has all the capabilities needed to drive RPA distribution across the organization. It’s a top-down approach where the CoE provides expertise and manages the shared resources required to deliver RPA. A centralized CoE also sets priorities based on the larger enterprise strategy. It creates a centralized shared platform that can scale to host all BU processes, and because a single team oversees RPA, standardization and optimization are easier to achieve and enforce.
    Unlike the decentralized model, the centralized model is very scalable, with one point of contact for the entire initiative. One potential downside of this model is that it can be slower to distribute automation across the organization.
  • Hybrid – The hybrid model works best with mature CoEs. In this approach, the CoE centralizes operations, while the BUs define their own demand. The CoE provides operations and delivery support, while the BUs do their own prioritization, assessment, and process development. It’s akin to an RPA factory where the BUs use capabilities from the factory to automate their processes. When run well, the hybrid model offers the best of both worlds: the scalability of the centralized model and the better distribution of RPA capabilities of the decentralized approach.
  1. Clarify funding
    Like any business entity, CoEs need funding. As with a governance model, the organization can choose the funding approach that best fits its needs.
  • Enterprise funded – An enterprise-funded CoE will typically motivate BUs to embrace RPA, but it also limits the accountability of individual BUs. This model is suitable when you want to roll out consistent processes across the enterprise.
  • Enterprise funded, with a chargeback to BUs – This approach encourages BUs to automate on their own, but they are on the hook for the development cost of processes. This is suitable to encourage adoption on an enterprise platform.
  • BU funded – Under this approach, BUs fund their own automation processes, and part of the funding goes to maintaining operations and governance mechanisms. This is a suitable funding model for when the initiative is at a very mature stage, and the demand for RPA from BUs is strong.

Establishing a well-run, well-funded CoE is a proven way to make sure all the pieces are in place to help an enterprise embrace the transformative potential of RPA. No two businesses are alike — so no two CoEs will be identical. However, taking the time to think through the issues and challenges that most CoEs encounter will allow the business to land on the CoE roles, operating model, and processes best suited to its needs.

Blog Co-author
Chet Chambers, VP, UiPath