Simply put, the processes that are most suited for automation are those that are repetitive and require minimal judgement.  Our Finance and Risk clients find there are many processes that match these criteria. But to help evaluate where robotic process automation (RPA) can make the biggest difference in banks, Accenture found that thinking through these questions can be effective in guiding how to evaluate potential RPA projects:

  1. What type of data is available?
  2. What is the nature of the data input?
  3. What magnitude of data volume is being processed?
  4. How much do rules drive the process?
  5. How logical are the exceptions?

A typical profile of a high potential RPA implementation includes one that has data available in digital format, where the data volume is high, and in a structured, consistent format.  In addition, the business rules that govern the data processing are consistent, and the exceptions follow a logical pattern so they are identified, isolated and handled manually. Some examples in Finance and Risk functions include the monitoring of key processes, general ledger account reconciliation or intercompany eliminations.

When will Breakeven Occur?

At the beginning of a bank’s innovation in an RPA program, Accenture often starts with low effort, high value projects.  These are generally determined by using parameters such as data location, number of apps and data fields used.  Additionally, rules that govern data access or validation tend to be few and simple, with minimal exceptions.   An example of a low effort project recently completed for an Accenture client is a Know Your Customer (KYC) automation.  Here, the necessary data was available from a single source, with few apps, and a couple dozen data fields to be accessed. The rules used for KYC validation were simple, and the few exceptions were simple to address.  This type of simple KYC process took approximately 3 weeks to automate and 1 week to implement.  A more complex process, such as managing limit breaches for example, could take roughly 8 weeks to automate with an additional 3-4 weeks to implement.

When will Breakeven Occur and how are these Benefits Replicated?

A single proof of concept project can create efficiencies by using technology and a solid methodology.  The real and material value to a business occurs when implementing RPA at significant scale and pace.   As for the challenges faced, they are similar in nature to those addressed during a typical transformation program, such as availability of skilled resources, internal resistance to change, or ability to deliver a cost-efficient maintenance model.

We believe the potential impact of automation in banks’ Finance and Risk areas can only increase as the capabilities of robotics grow and mature from RPA to machine learning and artificial intelligence.  These capabilities have numerous potential applications among production tasks and predictive analysis.  Robotics is challenging to categorize since it’s not all IT nor is it at all comparable to traditional workforce.  The paradigm shift driven by robotics should be in how companies integrate it with their existing human capital and IT capital.  The banks that capture the greatest benefits are those able to master the balance between all these parameters– both conceptually as well as in execution.

For more information on this topic, see our full presentation on Innovations in Finance and Risk – Robotic Process Automation and Accenture’s Robotic Process Automation page.

View the other posts from this three part series:

Part 1: Finance and Risk Innovations: Applying Robotic Process Automation in Banking

Part 2: Benefits of Robotic Process Automation Extend Beyond Just Cost Savings for Banks