Getting started with RPA is faster than implementing traditional IT systems and RPA initiatives can be implemented as rapidly as within 4-6 months on wellscoped processes.
RPA is characterised by being implemented in an agile step-by-step approach to ensure fast results on wellscoped initiatives with high benefit potential. Often a Proof of Value is initially carried out on a well-scoped area in order to assess the value of an RPA implementation.
The RPA initiatives are generally initiated, run and owned by the business units implementing RPA. In order to ensure RPA expertise in the initiatives, there appears to be two main tendencies on how organisations choose to approach this. The first is to facilitate an RPA Centre of Excellence in the IT department manned with the expertise of RPA specialists and IT specialists. The second is to root the RPA Center of Excellence as a part of the business units focused on RPA expertise with a limited inclusion of IT expertise.
The combination of expertise in business processes, RPA and IT is nevertheless essential.
The business units understand the business processes in depth. They are the ones who can assess both the potential and the possible pitfalls of the application of RPA software from a business point of view. This knowledge is crucial in order to ensure the right process design for the RPA solutions.
The IT department has important knowledge regarding the general IT portfolio of the organisation such as business systems and applications, IT architecture, existing IT infrastructure, data and governance on data management. This knowledge is essential in ensuring that the RPA software is applied in a manner where potential IT pitfalls are avoided.
An RPA implementation process will typically have five main activities, similar in nature to any other technology realisation process:
Discover, Pilot & Scope
Select RPA Software
Build & Implement
Monitor & Manage
Realise Benefits
Figure 5. The 5 main activities of RPA Realisation
Activity 1: Discover, pilot & scope
RPA is a new discipline for many companies, and also a discipline where the actual benefits as well as the content are unknown territory. Therefore, many companies have started their RPA journey with an open process where the purpose has been to discover and learn about RPA.
The main approach to the discover process is to perform one or more pilots on RPA. These pilots form the Proof of Concept or Proof of Value for an upcoming RPA initiative in the company.
The knowledge of the employees who are currently involved in operating the business processes where RPA is considered is essential to include in this process, in order to distinguish and understand the potential for automation.
Once the potential of RPA has been identified, a more structured identification of business goals takes place in order to form the basis for a more formal decision to start an implementation of RPA. The business goal identification is not generally a lengthy process resulting in elaborate strategy papers, but is important in order to be clear about what is intended to be achieved by RPA. This is, not least, important in terms of the subsequent change management towards the employees whose work areas are affected.
When the business goals have been determined, the scope of the RPA initiative is defined clearly, including an assessment of the organisation’s RPA potential in detail.
This can include creating an overview of the IT portfolio and processes in scope to identify the processes with a high potential for automation. Typically, the exact process handling on several areas are documented to a high level by step-by-step sketches – what is done, by whom, using which IT systems and how?
When the candidates for RPA are listed, the potential of each process is assessed according to the potential for automation, the potential benefits, and the expected effect on business goals. Based on this, one or more specific processes are chosen to support it.
Processes with an RPA potential are characterised by being:
- Rule-based and repetitive
- Manual
- High volume
- Time-critical
- Seasonal
- Triggered by a digital input and
- supported by digital data
- A low exception rate
- Causing rework
Subsequently, following the first activities, the companies feel comfortable that RPA can improve their business, they have learned about the technology, and they have identified the processes to be automated along with the potential expected benefits from the automation.
Activity 2: Select RPA software
The next activity for most companies is to select the RPA software platform. When selecting the software, companies typically evaluate the following factors:
- Product maturity
- Functionality
- Breadth of the portfolio
- Ease of use
- Scalability
- Reliability
- Security and compliance
- Availability
- Enterprise support and services
- Costs: both initial investments and running costs
The compatibility of the RPA software with IT standards and architecture is an important issue for most companies. Furthermore, knowledge of the IT systems involved is essential. This could be in terms of stability, how often the systems undergo updates and changes, etc. If the IT system in scope is undergoing recurring changes in the user interface, it is essential that the RPA software can handle this.
Investment costs, implementation costs and ongoing costs for maintenance, RPA licences, system licences etc., are also analysed at this stage.
Selecting the right supplier and implementation partner can often be as difficult as selecting the software, as most of the RPA software suppliers do not have operations and the consulting services supplier market is very fragmented.
Activity 3: Build & Implement
The build and implement activity for implementing RPA software consists of multiple steps. The companies, having implemented RPA, use different names for the steps but are all more or less going through the same processes. The approach is identified as agile, where steps are revisited in order to analyse problems further and find the right solution.
The detailed steps and their interaction are shown in Figure 6.
Figure 6. Build & Implement
Analyse
The first step is the comprehensive analysis of the processes in scope in order to understand each step and task in the process fully.
This includes detailed process mapping, identification of data inputs and variables, mistake patterns, possible process barriers, etc.
The analysis should be conducted in close collaboration between the business units and the RPA specialists. The results need to be documented on a detailed level in order to use it as a foundation to design an automated process with RPA.
Design & configure
Next, the automated process is designed in order to configure the software in terms of creating the business logic and rules, how to manage exceptions, etc.
Often, the software is configured simultaneously with the design step in an agile process between business experts and RPA experts. As such, the software is used to test the design of each process step in turn in order to create the most effective automation.
If more significant problems arise at this step in the implementation process, it may be necessary to revisit the analysis step in order to fully understand the problem and design the right solution.
Test
When the design is completed and the solutions are configured, testing of the full process automation is initiated, typically by performing a range of processes and monitoring the results.
If the tests show either problems or potential for further optimising, it may be necessary to take a step back and revisit the previous steps to understand the problems and perform the appropriate change in order to solve the problem.
Deploy
When the testing is completed, the solution is ready for deployment. It is important that the employees know how to handle the actual deployment and their tasks of interacting with the RPA robots and initiating adjustments if needed.
Data should be analysed on an ongoing basis in order to assess the stability and the performance of the RPA robot.
Many companies also define the KPIs of the process at this time. These KPIs are used to monitor the process performance accordingly. When the KPIs are defined often depends on the availability of data in order to establish a baseline for the KPIs prior to the RPA implementation.
Activity 4: Monitor & Manage
At this stage, companies organise the day-to-day operations of the RPA. These include factors such as:
- Deciding governance procedures where needed
- Setting up the monitoring of the robots – focus on the expected availability and performance of the robots
- Assigning the responsibility of running and supporting the robots
Activity 5: Realise the benefits
There is also a strong focus on the benefits realisation using RPA. The responsibility for the benefit realisation typically lies with the business units, perhaps assisted by the IT department or RPA experts.
The benefits realisation activities include:
- Continuously monitoring the KPIs
- Adjusting the organisation to achieve real cost savings
- Following up on the business performance of the RPA robots and initiating adjustments through the determined channels when needed
- Following up on how the processes and changes are realised in the everyday cycle of the business units
- Reporting on performance and benefits progress to management according to baseline and goals
- Recognising potentials for unexpected business outcomes and ensuring they are utilised into benefits supporting the business goals
This phase does not conclude as long as the RPA robots are in use, and the activities should be implemented into the routines of the business organisation as persisting activities.
