In one of our earlier blogs we talked about the importance of having a CoE (Center of Excellence) and the abilities it entails. After an organization decides to have its own CoE, it is important to be aware of the mandatory steps for a successful integration, taking into account assigned resources and the organizational framework. In this blog series to help you accelerate your RPA journey, we have gone one step back and asked ourselves “how can we build a CoE successfully?”. This blogpost (divided in two parts) will dive deeper into the subject by giving advice on what needs to be done from a business perspective while touching upon organizational and technical aspects.
It all starts by creating the goals, benefits and implications of a CoE for the organizational structure and the team building.
Align IT with Business goals
The CoE needs to have a long-term vision and an RPA strategy in place which considers technological changes which may occur in future. Don’t focus exclusively on RPA, but also put efforts on the operating model. Don’t try to do it yourself, you need to get support and align with other departments responsible for managing the digital workforce.
Before harmonization takes place, RPA targets should be clearly defined, and fixed deadlines set. It is crucial that the individual automation objectives are subordinate to the objectives of the CoE and that these in turn are linked to the greater corporate objectives. A bottom-up approach can be used here, where the sub-goals are first defined at the lowest level and then inductively checked to see whether they are compatible with the highest objectives. However, with this approach the risk of a conflict of objectives is high. Alternatively, a top-down approach can be adopted, which implies that the overarching objectives are broken down into granular sub-objectives. But this case could lead to unrealistic objectives and this approach is not particularly democratic, so it could have an impact on the motivation of the RPA team.
As a conclusion, a mix of top-down and bottom-up might be the best way to mitigate any risks arising from using only one approach. Nevertheless you should try to align with your company´s goal setting approach .
Create a cross-functional team
A CoE is like a joint venture of people who come together from different departments and contribute their specific know-how to the RPA implementation. The formation of an “expert team” is indispensable for communicating the RPA integration appropriately and company-wide. Also this team will concentrate the expertise of various areas in the company in the CoE which can be utilized for process analysis. The creation of an expert team might seem obvious, but it is key for the implementation of RPA and should therefore be mentioned again here.
The mandatory duties to define before setting up a CoE include the formulation and implementation of best practices and governance rules, which must be established within the organization. Sources to identify best practices can be internal training sessions, third-party consulting discussions, or lessons learned meetings. Generally accepted best practices range from not hardcoding system login data (to minimize security risks), to high modularity of workflows to ensure code reusability for similar processes (to accelerate future RPA implementations).
Tight planning and control of the RPA implementation helps with cost and time planning, risk management, and the aligning of RPA with overarching corporate goals. Such measures are not necessary for a small, manageable team, but are necessary if the team is constantly growing and gets over a certain number of people.
Even though the team consists of experts who formulate and implement best practices and governance guidelines, that is not all. The success of a CoE depends largely on its organizational structure, which we will discuss in the next section.
Select the appropriate organizational framework
First, we need to ask ourselves which organizational model we want to use to develop RPA. Generally, there are three different possibilities: Divisional, Centralized and Federated.
A Divisional CoE initiates RPA implementation in each division or department, independently of the other departments. This means that low-cost automation projects can be carried out with low effort. However, a department specific CoE is not scalable, has redundant hardware infrastructure and fragments RPA standards. If the expert team is focused on only one department, the formulation of universal RPA standards becomes more difficult.
A Centralised CoE eliminates this problem by acting as a central unit for all divisions and controlling the RPA implementation from there. This ensures scalability and communication. However, this approach is only recommended if CoE structures already exist in the company. Otherwise, the CoE may be a significant investment for the company, which is faced with limited resources for cost reasons.
This could be remedied by a Federated solution with several Divisional and one Centralized CoE units. This enables low-cost, scalable automation that runs across multiple departments and is controlled from a central platform. This approach can also be problematic if the structures are not already in place.
The selection of a CoE organizational structure is essential. There is no “one size fits all” solution, you should consider these above points and choose the model that adapts best to your organizational structure and goal setting approach.
Establish a focused Benefits Management
The harmonization of the CoE with cross-company objectives requires the ability to adequately quantify the automation efforts and benefits. The principal indicator for this is the expected Return on Investment (RoI). The introduction of further KPIs can measure targets in terms of cost, time or resources.
At a lower level, KPIs should be introduced for each process to measure the benefits of automation. A good starting point is to determine how many people or FTEs are involved in the process initially. As the result of automation is usually a reallocation of (human) resources into higher-value activities, a decrease of the required FTE is consequently expected. This change alone can measure the success of automation. Other assessment bases are, for example, the frequency of running the process (per day, week etc.) or the average time required to perform a transaction. Now that a robot is responsible for the process, it is convenient to increase the frequency or minimize the process duration time.
We hope you realized some of the elements you need to consider before creating a CoE. But there are even more elements to take into account to make sure that your organization has the right capabilities to put this into production. These include Infrastructure and Framework. Stay tuned for our next blog post to learn all you need to know before running a successful CoE.