The use of robotic process automation (RPA) is gaining momentum in the financial services industry. Here, we focuse on the common pitfalls banks encounter when planning for and deploying software bots.
Decision makers need to be clear-minded about what they hope to achieve from embarking on an RPA journey; the first step is really to understand why to apply RPA at all. After that, it’s useful to know what hurdles that must overcome on the way forward:
- Siloed automation efforts. With vendor promises of drag-and-drop simplicity, a business or operations unit may be tempted to take the lead in RPA deployment. In other cases, IT may claim ownership. Either approach can fall short. The business or operations side doesn’t always understand technology scalability issues, while IT doesn’t fully comprehend business pain points or process optimization opportunities. It’s essential for both groups to join forces and understand automation in the context of the enterprise transformation strategy.
- Poor process selection. When considering RPA, many companies take the approach of selecting a simple process, establishing that the automation works, evaluating the benefits it can provide and determining whether it will scale. This approach may overlook a key question, though: Does the process represent the complexities that the automation initiative might present? It’s important to identify a critical mass of factors and then determine whether automation is the right approach. For example, a process with few touchpoints to web applications and simple data entry will not provide inputs for tool selection. On the other hand, if most processes use virtual desktop infrastructure (VDI) and involve calculations and seasonal volume, then selecting the simplest among these will provide better insight into tool selection and an overall roadmap.
- Insufficient organizational involvement in execution. Leadership may view automation as chiefly a technology play. But if you simply hand off a process to third-party developers for scripting, they’re unlikely to understand key factors adequately – how the process affects other areas of the organization, potential liabilities associated with automation, and whether automation actually supports business goals. The business unit and IT need to coordinate their involvement.
- Underestimation of ecosystem requirements. Automation can exponentially accelerate the performance of tasks such as data entry, straining systems’ capacity to scale and respond to demand. You’ll need to assess and implement infrastructure changes to effectively integrate RPA into the organization’s IT portfolio, as well as add safeguards to ensure robots operate within established bounds. For example, a system designed to handle human interaction may show signs of stress when bots begin executing transactions 100 times faster than a human operator.
- Postponement of planning until after a proof of concept or pilot. It’s essential to choose the right process to evaluate for its automation potential, but a proof of concept is just that – it validates a hypothesis. Waiting until this happens before deciding how automation will be deployed ignores how the solution fits into the spectrum of what is planned. Creating a solid business case for the proof of concept can help avoid this. Such planning extends to security, compliance and people requirements, particularly the skills needed to scale the solution. For example, a basic question of how to manage IDs for bots will uncover changes required for regulatory compliance and security.
- Underestimation of automation’s organizational impact. Views of RPA’s potential impact on jobs range from a bounty of new opportunities to dystopian unemployment. By one estimate, about 60% percent of occupations could have 30% or more of their basic activities automated. However, the future plays out, people will find themselves doing different things and needing new skills. Managing these transitions intuitively and guiding people to prepare for more challenging and fulfilling work can help smooth the shifts.
- Poorly defined success and acceptance criteria. A validated proof of concept can suggest that automation can succeed without human assistance. But automation is not necessarily intended to replace humans. Instead, it can fulfill a higher purpose by providing information that helps people perform their jobs better.
By understanding the seven pitfalls of RPA deployment, and deciding in advance how you’ll use it to achieve business goals, your institution can begin to capture the value of this transformative technology. If you're curious to learn more, please continue to read about in what areas RPA is applied in banking.