What to consider before implementing your RPA Center of Excellence (CoE)

Publish date:

A crucial step in moving from the proof-of-concept stage to distributing RPA across your organization is to implement an operational model. This helps to ensure governance and alignment to corporate strategy and operations.

In a survey by Hfs, only 18% have set up a dedicated CoE model for RPA. However, 88% of these said that an automation CoE is effective in delivering business value.

So what is an RPA Center of Excellence model?

The RPA CoE is a pool of capabilities required to drive all the functionalities of an RPA initiative. The CoE functionalities and their processes are:


  • Integration into the overall corporation organizational structure
  • Internal and external roles and responsibilities to support the functionalities of the RPA initiative
  • Training of new resources
  • Change management.


  • Escalation paths
  • Compliance to policies and procedures
  • System accesses
  • Process prioritization.


  • Architecture for the robotic operating environment (ROE)
  • Infrastructure support
  • Technology choice and licenses.


  • Assessment
  • Development
  • Testing
  • Deployment
  • Development standards.


  • Maintenance of existing automated processes
  • Support
  • Monitoring.

What do you need to consider when implementing your CoE? How should you change the model as your initiative matures?

The first consideration is where your CoE model sits. The two most common options are within IT or as a Business Services function at Group level. However, I have seen organizations place the CoE inside a business function. My suggestion is to drive the CoE as a Business Services function at Group level, with a close relationship to IT. This provides the freedom required for such initiatives and it enables proactive and strong leadership. If the CoE is set up within a business function, it will be difficult to drive the initiative across the organization. However, it will work well to drive automation internally.

Do we want a centralized, decentralized, or a hybrid model?

Decentralized – The decentralized or federated CoE model will have its capabilities spread across the organization. The CoE functionalities (listed above) will run from separate business units (BUs). BUs will have their own capabilities for prioritization, assessment, and development of RPA processes. They will also drive their own operations. By giving the BUs responsibility to drive their own RPA awareness and progress, the decentralized model is a good way to start your initiative.

However, as there is no centralized control, the model is difficult to scale and it can be difficult to easily liaise with the IT team. The maturity across the organization will be variable and it will be a costlier solution, as there is no shared platform. The decentralized solution will foster a tailor-made solution, instead of standardization and optimization across the organization. If you start out with a decentralized model, it will take significant effort to move over to a scalable hybrid or centralized model.

Centralized – The centralized CoE model will have all the capabilities needed to drive RPA distribution across the organization. All functionalities of the CoE will be driven from one centralized capability pool. This is an inside-out approach; the CoE provides expertise and manages the common resources required to deliver RPA successfully. It also prioritizes processes based on the overall project strategy, which is, in turn, aligned with the organization’s strategy. There will be one centralized shared platform, easily scalable to host all BU processes.  The processes pipeline is automated by a centralized team that distributes automation across the organization. This model works very well with a delivery factory model. With centralized capabilities, standardization and optimization becomes a natural part of the initiative.

Unlike the decentralized model, the centralized model is very scalable as there is one contact point for the entire initiative. This model is a bit slower to distribute automation across the organization. However, it is a better model for scalability and operations.

Hybrid – There are pros and cons for both decentralized and centralized models, and it is always recommended to create a model that best fits your organization’s needs. A mature CoE can have centralized operations, while the BUs are mature enough to drive their own demand. The CoE provides operations and delivery support, while the BUs have their own capabilities for prioritization, assessment, and development of automated processes. If deliveries come from a delivery factory, the BUs will have dedicated capabilities from the factory to automate their processes.

The hybrid model is good for mature initiatives where the BUs are capable to drive their own demand. The hybrid model will have the scalability of the centralized model without the limitation to distribute automation.

How should the initiative be funded?

  1. The first option is to have a CoE model fully funded by the enterprise. An enterprise-funded CoE will typically motivate BUs to automate but it also limits the business units’ accountability. This model is suitable when you want to roll out processes across the enterprise.
  2. The next option is to have an enterprise funded CoE but with a chargeback from the business units. This will encourage the BUs to automate as the continuing costs are removed but they will be liable for the development cost of the processes. This is suitable to encourage adoption on an enterprise platform.
  3. Another option is to ask the BUs funding their won automation processes. Part of the funding will go towards the maintaining of operations and the governance mechanism. This is a suitable funding model once the initiative is at a very mature stage, and the demand from the BUs are strong.

There is no one-size-fits-all when it comes to operational models. Every company is unique and needs to implement a model that supports company governance and culture. The size, geography, and function of the company also play a major role to determine what the best model is for you. Connect with me if you think there more things to consider before we implement an RPA Center of Excellence.

Related Posts


Leading the way to a data-powered future

Ramana Bhandaru
Date icon August 27, 2021

Proud to announce that Capgemini has been named a Leader in Everest Group PEAK Matrix®...


How to become part of the data-powered magic

Zhiwei Jiang
Date icon April 1, 2021

Proud to announce that Capgemini has been named a Leader in 2021 Gartner Magic Quadrant for...


House of the rising data

Pierre-Adrien Hanania
Date icon March 31, 2021

How strong data governance will benefit scaled AI in the public sector