Achieving ROI from Robotic Process Automation in financial services
- Posted on May 24, 2017
Automation is a hot topic among banks and financial institutions today. Faced with agile competitors snapping at their heels and a rising trajectory to their cost-to-income ratios, Financial Services institutions are looking at automation to improve their productivity and efficiency. These organizations are showing particular interest in robotic process automation (RPA) and rightly so, as RPA can be scaled effectively within a short period of time and provide rapid return on investment.
What is automation and how does RPA fit in?
Automation and using technology to automate processes has been around for the last 70-odd years, with varying degrees of impact. Starting with hot keys and macros (remember mainframe programming and early Windows programming?), automation evolved through the 1990s with the emergence of messaging (MQ series), BPM (business process management) software and shared services (co-located or otherwise). Operations managers soon started promising the business 15-20 percent reduction in costs and efficiencies associated with straight through processing and shared-services model.
The Financial Services industry was the torch bearer and soon adopted these technologies and adapted their operations to shared services. However, one of the biggest challenges was legacy systems and the huge amount of time it took to make these new technologies (BPM, shared services) work with the decades-old legacy systems that were pervasive in Financial Services. This had a definite impact on time-to-market as well as the cost-to-time-to-market. In many cases, the cost associated with automation was a fraction of the cost associated with changing the legacy systems to ensure that automation software could work. In the late 2000s, a new adaptation of old technology came in to provide a balance between legacy systems and automation. This is the genre of software known as robotic process automation (RPA).
RPA is characterized by a set of software that is non-judgement-based and can automate human tasks involving at least two or more systems. The software is able to mimic human operations much faster and in a more auditable, repeatable manner, allowing for Financial Services organizations to reap benefits in terms of FTE reductions (between 25-40 percent), TAT reductions (between 30-70 percent) and significant reductions in variances associated with processes. These can also usually be implemented without having to change core legacy systems, essentially leveraging existing processes in a faster, auditable, machine-based way. RPA is also built for remote and global operations, which means it works 24x7 and also provides the benefits of shared-services operations.
How can RPA best be used in Financial Services?
RPA addresses certain processes in Financial Services and is unsuitable for others. To ensure your business case works, it’s important to be mindful of these distinctions. In Financial Services, RPA is most suited to three types of processes: in-country processes; middle office processes; and shared services or back-office processes. It’s also important to distinguish between attended and unattended automation. An example of attended automation is a mortgage process where parts of the process (usually between 40-60 percent) can be automated using RPA, but there is also a significant part that needs the attention of a human, e.g., payments and collection requests, mortgage processing and lending processing, PEP (politically exposed person) determination, reconciliation of trades, etc. Unattended automation is where roughly 70-80 percent of the process can be automated and the remaining is resolved by a human or using artificial intelligence, e.g., client due diligence, KYC data collection, AML verification, false positive reductions, fraud analysis, Finance reconciliation, procure-to-pay, etc.
The ROI of RPA for Financial Services organizations
Some banks have achieved up to a 40 percent FTE count reduction with RPA (even though they already had high levels of automation), as well as significant improvement of their TATs and variance measures. A typical RPA implementation for a process could be completed within 4-16 weeks depending on complexity and the amount of regulatory and legal approvals required. One important aspect to factor into your planning is the exhaustive testing and regulatory approval process that you will have to establish and follow as an organization before you release a RPA-powered bot into production. A benchmark that you could use for planning purposes is that if your bot takes four weeks to develop, assume anywhere between 4-8 weeks for testing and regulatory compliance. A significant amount of change management focus is also required to achieve a successful RPA implementation.
I’m pleased to share the news that Avanade is teaming with Blue Prism to implement RPA solutions for our clients and help them optimize business benefits from automation. You should know that Blue Prism is a leader in enterprise RPA software globally and considered a ‘Cool Vendor’ by Gartner. We have a series of upcoming joint seminars with Blue Prism on RPA implementation in Singapore, New York and London. Feel free to reach out to me for information on these events or on Avanade’s RPA offerings.