Robotics Process Automation (RPA)
Which lessons learned can be determined after the first Robotics Process Automation hype? How do Robotics Process Automation (RPA) projects become successful and when do they pay off? A first analysis.
So far quite good
When the term “Robotics Process Automation” (RPA) first appeared on the stage, it was widely acclaimed as a miracle cure and panacea for back-office processes that were scattered among countless legacy systems. The expectations for RPA were high, especially since the «proofs of concept» indicated high performance capability and tremendous potential for cutting costs.
During its implementation throughout a company and the establishment of «centers of excellence», however, various difficulties arose. Frequently, there was a lack of RPA governance, the appropriate change management, and development approaches that would have been suitable for RPA.
What was the problem?
Despite relatively low implementation costs, the greatest challenge during the automation of processes continues to be the scaling of RPA. The reasons for this vary from one company to the next. Still, these are among the most frequent factors:
- Lack of RPA governance: Fundamentally, successful implementation of RPA requires prior consideration of three points: (i) how the processes relevant for RPA are identified; (ii) how the RPA bots should be developed (and with what software); and (iii) how RPA is to be operated following implementation. These questions must be answered in the form of clear governance and taken into account in the business case.
- Inadequate focus on change management: The implementation of RPA should be accompanied by transparent communication. Employees need to understand that the implementation of RPA will mean a lightening of their work load and, in most cases, will not have any direct impact on their jobs. The objective of RPA is the automation of simple, rule-based tasks, not the elimination of jobs. The «expensive» employees will have more time to concentrate on their demanding and analytical tasks.
- Lack of agility during the development of RPA: RPA is a relatively fast and flexible methodology to emancipate employees from rule-based tasks and, in consequence, to heighten efficiency in the company. In the real world, however, it often turns out that the installation of RPA productive operation takes longer and is more complicated than expected, resulting in disappointment in some quarters. RPA is not necessarily to blame for this; the root cause is much more likely to be found in IT processes that were not flexible enough to allow the adaptations required for RPA.
What now and what next?
The basic principle is this: Complex processes lead to complex RPA robots and increase the costs for both implementation and maintenance. RPA should be just one influencing factor within a holistic transformation. When the desired output is analyzed, it becomes clear what is needed and what processes may need to be revised or can be eliminated completely. The goal of RPA should be the heightening of efficiency in the company and not the sewing of a «patchwork quilt» of poorly linked systems. As is true of all digitalization topics, an end-to-end evaluation is essential and significantly more effective than using RPA as a kind of duct tape for makeshift repairs at isolated points. This should not be interpreted to mean that the RPA implementation cannot begin until all processes have been cast in complex target models.
In theory, the re-engineering of all processes takes place before the «robotization» and automation. In practice, however, there are obviously good reasons why many processes cannot yet be re-engineered, whether because of technical or other capacity restrictions. The advantage of RPA is that it makes improvements in efficiency possible even when processes have not been completely optimized and creates the time and capacity for re-engineering the processes at the same time.
Experience has shown that successful use of RPA requires a change in thinking: from experimentation with automation to transformation. The exploitation of the full potential of RPA requires a solid commitment from management in the direction of RPA. A key element is finding the right balance between central leadership and flexibility. A number of different aspects must be clarified in this respect:
- Establishment of solid governance: In organizations that begin by testing RPA in pilot programs, it is important (at the latest, when the company-wide rollout begins) that there is agreement about the tool provider, the architecture, lines of responsibility, etc. The early involvement of the company’s IT department falls under this heading as well. After all, the IT department is ultimately responsible for the integration and scalability of RPA in the current IT environment, for the conduct of incident management, and, finally, for initiation of the go-live.
- Change management measures: Another task of change management (in addition to the communication of the advantages of RPA mentioned earlier) is the installation of RPA skills. Employees should be given training on the topic of RPA so that they can later develop robot concepts directly from the business department. A «digital workforce» of this type can provide substantially greater flexibility to a company and lead to decisive competitive advantages. When all is said and done, the «digital workforce» can grow into a «center of excellence» and become the place where the RPA strategy in the company is defined and that makes best practices available. It can ensure that further use cases are identified and that their feasibility is examined, making possible the permeation of RPA throughout the company. «Relevant skills» are not limited to strictly RPA-related expertise, however. Project management as well as process and change management experience are also required.
- RPA development according to agile methods: Agile development approaches have proved to be successful methods in practice. The way development teams tackle a subject is fundamentally changed, whereby the advantage is found in the immediate feedback. Teams are constantly notified about successes and failures and can build on the experience they have gained to modify their approach. The application of agile methods creates so-called «minimal viable products» that generate added value at a very early stage and allow teams to make small, but continuous steps in the direction of automation. The breaking down of major transformation projects into separate units, so-called sprints, keeps stakeholders up to speed regarding progress and obstacles of the project, and they can if necessary, contribute.
If the full potential of RPA is to be mined, certain framework conditions in the company should be clarified. The core elements are the support of management and the implementation of an appropriate governance structure.
As is necessary in any digital transformation, the processes that will be automated should be subjected to an end-to-end analysis. The objective of the analysis should be the critical examination of the processes and, as necessary, their re-engineering. The actual automation itself does not happen until the second step. In reality, however, technical or capacity restrictions may require the re-engineering of the processes and their automation to take place parallel to one another.
When used properly, RPA is an effective tool for heightening efficiency and, in the end, for cutting costs. Nevertheless, implementing RPA for no other reason than to save money would be too short-sighted. The far greater vision is to free employees from simple and time-consuming tasks and to give them more time for more demanding tasks such as the re-engineering of processes or the analysis of the data.