Robotics – short-term win or long-term strategy?

By Steve Young
7 January 2019

What if robotic tools become the new Excel? Powerful and able to solve many challenges but simultaneously introducing multiple and growing operational risks? As time moves forward, the benefits become lost and the management headache grows.

In the industry today, there are lots of discussion and buzz around the deployment of Robotic Process Automation (RPA) tools. It is hard to avoid press articles, webinars and event sessions covering this topic. Currently there is plenty of noise but very little substance around the subject. As with many emerging technologies, focus should be on outcomes and expectations need to be set and managed appropriately.

An area of focus for utilizing RPA tools is to manage low value and repetitive processes, freeing up human assets to work on higher value activities. This concerns me on two levels. Firstly, the industry has made great strides in automating obvious tasks and driving operational efficiency. Secondly, a level of domain knowledge and/or a reliance or involvement of legacy applications with limited integration capability drives much of the manual intervention.

The industry has already automated much of the operating model. STP rates have grown steadily to a point where only illiquid or unusual assets need manual oversight. Other tasks, for example reconciliation or reporting, can now be managed through workflow-based tools and exception processing. The concern with giving established operational teams RPA tools is that bringing automation and speed to a poor or aging process is potentially a step backwards. Inefficient processes are common in the middle- and back-offices of many firms. This is due to an ultra-conservative and change-averse culture, often aligned to aging and inefficient legacy technology. To hide this inefficiency behind RPA processes has the potential to increase operational risk and delay the critical process of deactivating legacy applications.

Excel is still a common answer to resolve complex problems. In most cases, the original author has, or will be liberated to, move onto other processes – in line with much of the current RPA rhetoric – and as a result a significant number of Excel solutions are poorly understood, poorly managed and extremely sensitive to changes to the operating model. Documentation on the requirements, design, data flow, coding and processes is poor, should it exist at all. Given most firms still do not have an answer to these problems, how are they planning to ensure RPA does not exasperate operational risk?

Governance and oversight are paramount for RPA managed tasks. Digital processes need to have at least the same level of scrutiny as manual processes. This requires management in an effective and efficient manner. Firms also need the capability to understand and interrogate processes quickly to evaluate outcomes and provide audit and oversight.

Another area for concern is the retention of process and domain knowledge. RPA processes require comprehensive documentation, regularly validated and constantly updated. This documentation must be both easily available and simple to understand by readers who may not be familiar with the subject. As firms become more agile and the speed of change increases, the ability to interrogate the processes and affect the evolving operating model will become a key competence, find out this here.

RPA processes need to encourage and facilitate change. Firms who wish to create an agile organization, both technically and culturally, need to be able to manage and implement change more regularly, more predictably and more efficiently. RPA tools could have a significant impact on this, in either direction.

The first step to bringing efficiencies and scale to any operation is to review the operating model and associated processes. RPA tools should be a part of this evaluation and must be part of an improvement journey. In isolation, they present a very real danger of bringing predominantly short-term benefits – but at what price?