Robotic process automation revolutionises banks’ back offices

At a time when banks face a host of challenges in growing their business and rising compliance costs, they’re looking for solutions that can provide a quick return on investment (ROI). Robotic process automation (RPA) is one solution delivering those results, according to Liew Nam Soon a managing partner, ASEAN financial services at the Singapore office of accountancy giant EY.

While the top-line numbers are dependent on the sophistication of the solution, banks are already getting robots to operate at one sixth the cost of an offshore worker and they are achieving 50% to 70% cost savings. Moreover, as Liew notes, “you can get a robot up and running in six to twelve months.”

Benefits of RPA

RPA has advanced substantially in Asia over the past 18 months. It is enabling banks to fulfil several objectives and, most importantly, lowers the cost of serving customers and increases operational efficiency, which in turn frees up people to do value-added work. It also improves quality because, once the software is constructed and the right data is being input, human errors largely disappear.

While there was a clear benefit to outsourcing work to staff in India, China or Malaysia in the past, Liew notes we are now seeing organisations use robotics to complement those workers, supplementing staff by embedding robotics as part of offshore capabilities. “This is pretty disruptive,” he notes wryly.

Another key advantage is that installing a robot doesn’t require sophisticated programming. “Everything is workflow – and parameter-driven, which is a quick accelerator.”

Liew adds it is also important to clarify that these robots aren’t the cute consumer-focused mechanical men that people often envisage. “This is software algorithms, a piece of software that does manual activities really well. What the software does, as long as you provide the data, is the manual key-in, plus checking, plus reconciliation. It’s complementary to process workflow.”

A key advantage of these RPA implementations is that they are inexpensive. “Compared to core transactions or payment systems, they’re a fraction of the cost.”

Where to use RPA

“RPA is applicable to anything that’s manual and high volume,” says Liew. While it may not be appropriate for operations with only one or two people, it is very worthwhile in places such as financial institutions in Singapore where upwards of 50 people may be carrying out manual entry.

Many banks in Asia have started their shift to RPA with changes to trade and cash because they have many manual processes and high headcount. “They develop a proof of concept (POC). Once that produces results – throughput goes up, cost is one sixth – they’re ready for large-scale deployment.”

Overall, RPA is used the most in finance and operations, including trade and foreign exchange (FX) operations as well as retail banking. “Client onboarding can be automated,” notes Liew.

“All the forms that need to be processed can be automated, along with the chequing, product processing, payment processing, links to payment gateways, settlement instructions and reconciliation.

Additionally, RPA can be used to support analytics, whether it is for detecting fraud or identifying transaction patterns. It helps to get the data into the system, and the analytical system then taps on it to make the right analysis.

Success factors for RPA adoption

RPA is following a classic adoption curve, Liew suggests Most of the leading multinational banks have been early adopters and Asian regional banks have started working on RPA.

The banks that are more successful at RPA have appointed a group to understand the application and determine where it can be deployed. The group then tests the hypothesis and business cases; where they achieve savings, before proceeding to” industrial strength” applications.

Instead of just automating existing processes, banks are also re-engineering current practices. Where there are duplications or inefficient processes, for instance, banks will redesign the process before they put RPA on top. “Otherwise, it would automate a sub-optimal process. It is complementary to redesign workflows as part of their programmes.”

The imperative for good governance

The other imperative for RPA implementation cited by Liew is good governance. If individuals within a company implement RPA on their own, there will be issues with consistency and how things are being processed.

An important step, then, is setting RPA up in a controlled environment with governance that does not allow it to grow uncontrolled across the organisation. Indeed, banks have found that getting the governance right and fitting the automation into their larger plans is crucial.

They define robots, functions and the ownership of robots in terms of development, maintenance and updates. They then continue to track the ROI of the robots, and the benefits.

The machines typically operate in a secure environment, often at a centre of excellence. “There is a robust process for writing the programmes and updating them,” says Liew. The development of the software and the robots are in a controlled environment, with little human intervention once they are up and running.

What’s next?

With the benefits of successful implementation being so compelling, Liew expects to see mass adoption of RPA over the next 18-36 months.

That said, it is not all plain sailing. One of the key issues is the social one. While banks are achieving efficiency and quality improvement, they have yet to fully work out how to deal with the social ramifications – given that not everyone who is displaced can be upgraded.

Liew stresses it is also important to understand that RPA is no ‘silver bullet’ and cannot completely take over from humans. Given how rapidly RPA is evolving and the likely scale of its impact on banks’ back offices, though, people do need to evolve to be more value added.

While there are clearly issues to be worked out, achieving better efficiency and quality – while typically saving 50% to 70% – means that banks in Asia are likely to move forward rapidly to make the promise of RPA a reality.

2770 views

Related reading

Tags: