The Xceptor Blog

Are these three things hampering your RPA business case?

27 March 2018

3 minutes read time

Building the business case for Robotic Process Automation (RPA) can have its challenges. But is it being made harder because the vision for automation arbitrage is too narrow?

The Robotics Division

RPA definitely has a place in the digital revolution but it can't be the sole answer. While the hype may well have started with RPA, one tool is unlikely to solve every process needing automation. RPA is about assisting or removing humans from a process. As Gartner says:

“Evaluate RPA opportunities where you have people acting as 'swivel chair integration' - rekeying data between systems - and where work is being performed by humans that involves structured, digitalised data processed by predefined rules”

Use RPA for what it does best and open up your kit bag to other technologies that can help with automation. Lots of activities in banking are complex. Lots of processes are high value. Lots of processes will still need human intervention. These are not what RPAs do best.

Step away from the Robotics Division and create the 'Process Revolution Division' or an 'Automation Programme' that welcomes other software such as AI, NLP, rules-based solutions that include workflow and reconciliation and can easily handle complex, unstructured data.

A bad process is a bad process

With constraints on talent, costs and an increasing focus on risk and data quality, automating existing business processes is essential. But existing processes are typically influenced by a raft of external factors that are a far cry from best practices. These may be legacy systems, mergers and acquisitions or the processes may simply exist because they have been done that way for so long and no-one has challenged it.

The most basic function of RPA is to replicate a human process. So the RPA copies the existing process. If it is a bad process, you'll just automate the problem. So rather than automating a bad process, you need to look to transform the process. We get it, simply replicating an existing process seems a lot easier than transforming a process but transforming the process is essential to fixing underlying issues.

Data-defined not process-defined

A single focus on RPA for your digital revolution will limit your automation capabilities. We work in an industry that has lots of complex documents with unstructured data. These need automating too and the RPA can help here but it might need some help first.

To automate processes effectively you need confidence that the data in the document can be pulled out accurately. Back to Gartner, RPAs do best with structured, digitalised data. By focussing on the data, we also can automate those higher value processes and generate better ROI.

This can be done by taking a RPA 'safety net' approach. By transforming, validating and exception managing the data before it goes into the robot helps guarantee the quality of the data and service. By taking a data-first approach, you naturally shrink, change and transform the process. And that is a good first step on your journey to digital transformation.