Asia finance professionals discuss lessons learned in employing robotic process automation.
Finance teams at multinational corporations have, by now, heard a lot about the benefits of using robotic process automation (RPA) technology to help handle high-volume, repetitive tasks as the business pursues growth and scale. And while some companies are well down the road of adopting RPA, others are just starting to consider it or have it on their to-do list. That spectrum of experience formed the backdrop at recent meetings of NeuGroup’s Asia CFOs’ Peer Group and Asia Treasury Peer Group where members discussed their journeys in exploring and implementing RPA.
RPA is tactical, not a replacement for ERP. RPA should be used as a tactical tool to enhance automating cumbersome processes. One member used this analogy to describe the role of RPA in her company’s technology strategy: “If an ERP is like a subway network, then RPA is there to help in the last mile of the commute, like a shuttle bus service or a bike-share service.” One member shared about her company’s ERP implementation project and how RPA was used in data migration from the old legacy system to the new ERP system as part of the tedious phase of this project implementation.
Redesign business processes. Do not underestimate the importance of redesigning business processes before implementing RPA. If you use a robot to execute a process as humans would do it, the results will likely be suboptimal, and fail to leverage the full technology capabilities of the wider system infrastructure. To avoid such an outcome, one member’s company spent 18 months to redesign business processes in preparation for a robust RPA rollout.
Where to locate RPA teams. Several members said they have a centralized team within the region that focuses on RPA projects company-wide. For one member, the driver was to have this centralized RPA project team in Hong Kong, where the company has its regional management office. That way, decisions on prioritizing the various RPA-related projects benefit from having regional management input and more resources from the bench strength of a regional office.
Consultant considerations. For another member, the critical factor in locating the regional RPA project team was to work closely with an external consulting firm in Shanghai that had expertise to cover projects for the region and offered the most cost-efficient service. And while you can start RPA projects by using external consulting firms that provide experience and know-how, one member recommended building RPA development skills internally so that your team becomes self-sufficient and can expand and maintain all RPA activities.
Progress report. This member started the RPA journey in 2016 with their first group of robots with the assistance of RPA vendors and a consulting firm executing on implementation. Today, they have their own internal automation center of excellence team which evaluates business requirements for RPA and develops RPA solutions. They use 20 RPA licenses to operate over 500 bot programs every month and have downsized average RPA development time to 10 hours of programming per bot.
Start small and scale up when ready. With each RPA software application license costing about $10,000 annually, the barrier to entry is low and you don’t have to consider a heavy capital expenditure investment. The recommendation from several members is to start small RPA projects as a proof of concept, focusing on easy cases with quantifiable justification for RPA; then progress from these quick wins to expand RPA deployment in targeted areas while establishing a framework and methodology for an efficient RPA rollout as a proof of ability; that’s followed ultimately by a proof of value phase where you’re driving operation transformation with enterprise-wide implementation of automation initiatives.
Find your bot comfort level. Bots can be programmed to do any repetitive task that is logically sequenced. The bigger hurdle for humans is deciding what tasks we are willing to let go of and delegate to a bot. Members discussed their bot comfort level at the recent NeuGroup meetings. A few use bots for FX trading execution, with one using them for both execution and FX settlement. That led to a question on how to ensure adequate controls when RPA is used to execute financial transactions such as FX trades, revealing that some members only felt at ease with bots doing internal reporting tasks, like creating management reports that draw on information from various systems. Others are comfortable using bots to prepare tax filing reports and for central bank reporting. In this discussion, one member highlighted that her company uses an artificial intelligence (AI) tool to negotiate and administer legal contracts, which sparked surprise from others. We’ll look into that subject in greater depth at future meetings and in upcoming articles.