The choice of a process to be automated is essential for successful implementation of robots using RPA. Proper analysis of a business process has a significant impact on the success of the entire project.

Critical assessment of automation potential for project success

One of the key factors affecting the success of robotization projects is the ability to assess automation potential. The process selection phase is one of the most critical moments during a RPA implementation project, therefore it is worth focusing on the right approach. 

Various assessment criteria

In recent years, many technology vendors and companies engaged in implementing RPA projects have analyzed the potential of process automation based on criteria whose fulfillment ensures maximum effectiveness of robotization. Effectiveness in this case should be understood as maximum working time savings as a result of automating a given process as compared to automation costs.

The assessment should also take into account the qualitative aspect of the automated process. In some cases, increased processing quality (minimizing errors occurring during performing a given task) becomes a prevailing benefit.

Assessment indicators result mainly from limitations of the available tools and the technology itself and take into account the economic aspect of the whole undertaking. When defining the selection criteria, one should remember about the foundations on which the RPA concept is built upon, as well as what business requirements it needs to address.

RPA aims at freeing employees from arduous, monotonous and repetitive activities not requiring creative thinking. From a technical point of view, the technology is a relatively cheap and easy to implement method of integrating various IT environments used as a part of a given process. It does not require direct interference with the systems. Instead, it operates in the user interface layer only. It is particularly useful when working with environments based on legacy systems.

Considering all these RPA differentiators, the process assessment criteria may be divided into three main domains. Criteria can be derived from the:

• identified specific project characteristics
• potential benefits ensured by the automation
• potential challenges that may occur upon developing the automation solution

Criterion 1. Project characteristics

The first group of criteria results directly from the processing path of a given task. Processes that are suitable for robotization have clearly defined execution rules. It is worth remembering that the created robots cannot make decisions on their own. They only execute programmed algorithms and every unexpected situation which occurs during their operation will prevent proper process execution. Therefore, it is important to clearly define the execution paths with the maximum number of scenarios, whereas the processes with few such paths are best suited for automation in the first place.

Criterion 2. Potential automation benefits

Another category of criteria specifies the automation potential of a process in terms of maximizing the benefits and savings resulting from RPA implementation. Automation should cover labor-intensive processes which are typically carried out by multiple employees. Robotization of such process will allow for mitigating the impact of human error on the final outcome as well as freeing up employees for more engaging and innovative tasks. Another important factor affecting the final benefits of robotization is the number of tasks regularly performed as a part of a given process. RPA allows for accelerating the execution of a single operation, whereas for large volumes, it translates directly into tangible time savings.

In addition to potential time savings, the issues related to the quality of the existing method of processing should be analyzed. The processes involving errors resulting from wrong input of data or broadly understood human factor are good candidates for automation. Robotization using RPA tools allows for minimizing such errors and, consequently, eliminates the time so far spent on potential adjustments.

Criterion 3. Potential challenges

The last group of criteria is based on analyzing processes in terms of potential challenges that might occur during implementation of solutions using RPA tools.

An important issue in this case is the format of the information provided. A process which is suitable for automation should be based on well-structured data that is stable over time. It is recommended that the provided information be stored in a consistent format using predefined templates. For example, input data stored in spreadsheets should always be stored in the same columns or rows. Each additional write method requires configuration of extra processing paths as a part of the bot being designed, which extends the solution development time and increases deployment costs. What is more, in some cases, the read and processing of information stored, for instance, in handwritten forms may prove ineffective or even impossible. These issues concern not only spreadsheets or document scans. Similar procedures exist for other data sources, such as e-mails, websites or third party applications. The analysis should also consider the issues related to the character of the processed information. It may turn out that so called sensitive (personal) data is used during processing which will require special attention. As a part of automation implementation, it is necessary to meet all the company’s applicable internal requirements related to data privacy and processing.

Not only quantitative indicators

Based on the presented assessment indicators, several different approaches to the analysis of process automation potential have emerged. It is becoming increasingly popular to use earlier prepared assessment sheets to estimate potential automation benefits based on a number of criteria by assigning numerical values to them – the total score enables process classification. The problem with this approach is that the assessment criteria are often described in a qualitative manner rather than quantitative one, which makes proper classification much more difficult. Therefore, the final assessment should not be based solely on the total score, as numbers do not always reflect the actual potential of a given process. Thus, assessment sheets are a very useful tool for evaluating robotization capabilities, but the final decision should not be based only on the assessment results.

The rule of five

An interesting approach to the issue of selecting a process for robotization is the so called „rule of five”, proposed by Craig Le Clair, a Forrester Research analyst. As he writes in his report, the processes that are best suited for automation follow three simple rules, based on the number 5. Robotized processes should include up to 5 decision points, use up to 5 systems in their operation, and the single processing algorithm cycle should take up to 500 clicks. This simple set of rules is aimed at rejecting the processes which are too complicated to be performed using RPA tools. Such approach to the analysis, however, takes into account technology limitations only while disregarding the aspects related to robotization benefits. Unfortunately, such an approach may result in rejecting the processes which, from a business point of view, are very good candidates for automation.

Significance of experts

When considering various approaches to analyze, it is impossible not to mention the significant role of experts. They are the main source of information on the processes subject to assessment. Whether or not the collected data is accurate, and the final outcomes of the analysis are positive, depends on the experts’ knowledge and engagement. Effective communication between the deployment team and process experts is a foundation of successful implementation of robotization.

Summary – a combination of various approaches

At the moment, there is no universal analysis path. Process selection is a complex issue, which has a strategic impact on the final benefits, and it may not be addressed with a simple table or a short questionnaire only. The main problem faced by teams which choose processes for automation is that selection criteria do not give us a clear answer about whether a given process is suitable for robotization. It is impossible to define specific limits which will ensure successful RPA implementation. Moreover, comparing different processes in quantitative terms is also problematic. For example, the fact that one process has so far been executed by 50 persons, and the other one by 10, does not have to mean that choosing the first one will be a better decision. It should be looked at holistically by focusing, in particular, on the potential challenges which may affect the final outcomes. To be successful, the analysis should adopt a combination of various approaches and selection concepts. In the end, however, it is the experience of the RPA implementation team, which analyzes the automation potential in a given enterprise, that will have the biggest impact on the final outcome.

Maciej Roszkowski – Intelligent Automation Specialist at Onwelo, involved in business process automation and intelligent document processing.