Robotic process automation (RPA) is software that sits on a PC or workstation and is programmed to mimic the activities that a member of staff would perform.
It will open applications, copy and paste data, and follow predefined rules. A robot will complete activities three to five times faster than a person.
Customer satisfaction can be increased significantly by RPA; for example, a firm requiring 30 days’ notice on redemptions was receiving a high volume of calls from clients the week before the 30 days expired. The clients wanted to confirm how much money was going to be transferred and when it would be in their accounts. RPA was applied, and now the clients receive a text a week in advance of the funds transfer with the required information. As a result, calls to the helpdesk have dropped to zero and customer satisfaction has increased.
Experience to date shows that it is unlikely RPA will result in job losses, more likely it will “take the robot out of the person” giving the employee more time to do value-adding work and increase their job satisfaction – what employee wouldn’t prefer to hand-over checking sanction lists and completing KYC/AML activities to a robot?
RPA is sometimes referred to as front-end integration as it connects two systems through the user interface that is accessed by the robot. Unlike traditional back-end integration, it does not require a significant IT project with systems vendors and developers and as such, RPA projects are much lower risk and a lot less expensive and time-consuming. Typically, RPA can be applied to a process in eight to 12 weeks and provide a return on investment of less than a year.
In addition to increased customer and employee satisfaction and fast return on investment, RPA improves the control environment. Robots don’t have “fat fingers” and are 100% diligent regardless of the time of the day or week; they have nothing else going on in their lives to distract them from the task at hand. Audit logs provide a full record of all the actions the robot has taken.
While RPA is part of a suite of technologies that generally get bundled into the digital transformation bucket, it is more tactical than many of the other elements of digital transformation. For example, a retail bank recently applied RPA to loan administration processes for a runoff book. The book will close in a year and, a strategic project did not make financial sense. RPA reduced the headcount by almost 20% and paid for itself in six months.
Reduced or redeployed Full-time Equivalent (FTE) is often the driving objective of RPA projects. Depending on several factors, we typically see organisations achieve financial benefit when 0.5 to 1.5 FTE can be released by automation, so the cost/benefit threshold is low.
Not all processes can benefit from RPA. The cost of implementing RPA is a function of a relatively small recurring license fee for the RPA software plus a one-off charge of configuring the robot to complete a process. It makes sense to invest in processes that are mature and not likely to change in the medium term. Also, the complexity of the process (number of clicks required to complete it and the number of applications involved) will impact the cost of configuring the robot.
Where there are digital and structured inputs the business case for RPA is improved, e.g. a structured and typed form is easy for the robot to deal with, an unstructured handwritten form requires additional investment in optical character recognition software.
Organisations interested in RPA usually start in one of two ways: a quick pilot to demonstrate the value to the executive team; or a “flash assessment” where processes are scored for automation potential and prioritised accordingly.
RPA is one of the many tools for transforming financial services. While at present, it provides opportunities to outperform the competition, it is difficult to see how organisations that ignore it can remain cost competitive in the medium term.
This article first appeared in Finance Dublin, July 2019