Banks can take remarkable steps toward successful automation by answering 04 simple, yet fundamental questions: Why, What, When and How?

With the potential to reduce monotonous, error-prone work and streamline various processes, Robotic Processing Automation (RPA) is making waves in the banking industry. Unfortunately, it is not a silver bullet for every problem of banking institutions. Robotic tools only drive efficiencies when applied correctly. By following the principle of 3Ws and 1H, banks can minimize possible risks and unlock the power of RPA.


Every solution must start from the problem. In order to reap the benefits, banks must make sure RPA is the right answer for their issues. The first and foremost question, hence, is Why RPA?

According to Gartner, one of the most critical determiners of success with RPA is correct application of the technology. Therefore, it is essential to establish and examine selection criteria to identify whether a task is ripe for RPA. The targeted processes shall repeat often enough and not change frequently to merit automation. Fair data quality and a large number of transactions per session are also crucial factors. Most banking operations, luckily, are rule-based and repetitive by nature, which makes them ideal for RPA automation. Still, banking institutions shall pay special attention if they intend to automate processes that require a high level of cognitive capability or cannot ensure the quality of data.

Instead of jumping into RPA adoption, banks shall always evaluate other alternative strategies, for example, redesigning the process or applying programmatic techniques like API-based (application programming interfaces) approach. In some cases, the long-term cost of upgrading a legacy system through API is much lower than automating with robots. More importantly, since RPA is a non-intrusive technology, it does not require complex system integration, and thus cannot create additional efficiencies in function.

Targeting RPA at the wrong process is among top common reasons that turn automation initiatives into disappointment, as revealed by Earn & Young. By starting with Why, banks can mitigate the risk of getting off track or being too ambitious with RPA-based automation.


Depending on the complexity of targeted processes, the robots will be designed with different characteristics. This is when banking organizations need to question: What type of RPA shall be implemented? (Read more: Keys to successfully choose the best RPA tools)

Generally, RPA archetypes can be divided into four levels (Dan Latimore, 2019). At its simplest, the robots replicate monotonous, repeated actions of humans at presentation layer of desktop-based application. These are known as task robots – perfect for rules-based tasks that rely on structured data such as, account activation, account closure, code issuance. Task robots can be upgraded into meta robots by leveraging API-level integration to create system-to-system automation, thereby conducting more complex, scalable and multi-skill processes.

Next level is IQ robots, which can learn and adapt over time to handle fuzzy rules, unstructured data, and are capable of making decisions based on accumulated learning and experience. The most sophisticated robots, considered as the Holy Grail of process automation, are AI-powered ones. Being operated by algorithms and integrated with smart features like computer vision, natural language processing, optical character recognition, AI robots are able to perform tasks that require language interaction, dealing with high amounts of unstructured data, and decision making in unusual situations. Popular use cases are trade surveillance, communications, transaction triage and data analysis.

Understanding what robots archetypes are most suitable to its current automation needs enables a bank to save time, money and efforts from unnecessary research and assessment. The chosen solution shall also be aligned with IT infrastructure and long-term strategy to ensure sustainability and highest efficiency.

(Read more: Breaking Barriers to RPA Adoption)


You have probably heard the phrase “Timing is key”, and RPA adoption is no exception. To achieve the highest value of automation, the process must be matured enough in term of architecture. As RPA only mimics human activities via the system’s existing interface, it is not in itself a cure for ill-designed processes. If the process is not well-architected, automation might even worsen the performance. Take debit card fraud processing for instance. It is nearly impossible for robotic tool to increase the accuracy of fraud handling with poor procedures and inadequate evaluation criteria. Therefore, ideal processes are those that have been refined and improved over time, demonstrate little to no implementation issues.

Another timing factor for banks to take into consideration when implementing RPA is the readiness of employees. In order to avoid the stiff resistance, human workforce should be informed about upcoming changes in the work flow and educated that RPA, instead of being a job killer, would allow them to perform more intellectual tasks. It is crucial to develop personnel transfer or restructuring plan prior to the deployment, as the work scope of staff will undoubtedly be diminished. A task force shall also be established and trained to ensure adequate in-house knowledge for RPA toolset management.


Now that the process to be automated and the level of complexity have been determined, your bank is totally ready for an RPA initiative, one last question remains: How will the automation be conducted? Normally, RPA projects consist of four major phases: assessment, Center of Excellent development, implementation, maintenance and support. The specific plans, however, vary significantly, depending on the partner banks work with.

Since banking is an industry of extremely complicated nature with tangled procedures and strict compliance requirements, it is crucial for vendor to possess solid financial knowledge and diversified experiences in the domain. 24/7 technical support is an absolute strength as several banking services must be available around the clock. In addition, the RPA vendor shall be able to offer cutting-edge products, yet easy and quick to implement that enable banks to keep up with constantly evolving demands of customers.

FPT Software has transformed operation process and improve cost savings for banking institutions via our RPA solution - akaBot. For further information, please visit here


  1. EY (2016). Get ready for robots. [Link]
  2. Dan Latimore (2019). Case study on Bank of America: How robots help serve and protect the bank. [Link]
  3. Oliver Wyman (2017). The new CIO imperative. [Link]
Author Pham Mai Ngan