“If you can’t measure it, you can’t improve it.”
Robotic Process Automation (RPA), an innovative form of business process automation technology that was introduced in 2000, has witnessed immense adoption in past few years. As per a recent survey conducted, 53% enterprises have already started their RPA journey and this is expected to increase to 72%1.
The benefits of RPA adoption are significant and it is no surprise that RPA has become a buzzword in technology space, drawing a lot of interest from industry and investors. With the payback reported in less than 12 months, enterprises have observed improvement in compliance (92%), higher quality / accuracy (90%), enhanced productivity (86%) and reduction in cost (59%)2. While RPA is brimming with potential, a recent research by EY claims that 30-50% of the RPA implementations have failed3 in generating desired outcomes, leading to questions on RPA capabilities. It has also been observed that many CIOs have stalled their ongoing process automations to revisit their decision.
What drove CIOs to significantly invest onto untested and unexplored technology? The untapped potential of RPA to drive valuable business results in terms of higher accuracy and productivity were of course attractive and considered. The limitations of a RPA solution were probably ignored. To measure the results of a solution, it is essential to know the success metrics and this critical parameter was unfortunately set by very few enterprises before selecting a process for RPA adoption. Investing without the possibility to measure benefits is an absolute risk.
This is where RPA Analytics comes in and becomes an integral part of Automation Implementation journey. Integration of analytics with RPA, i.e. the information resulting from the systematic analysis of data and statistics, has now made such investment decisions more informed and calculated. Analytics in RPA can be applied in 3 stages:
- Pre-Process Automation: A RPA contender process, usually operated manually, generates significant amount of measurable data to assist CIOs in taking a sound decision on investments. For instance, a German bank realized the importance of analyzing such data only after few of its initial RPA implementations didn’t save enough to justify the RPA solution license cost. The bank established a core team to review processes before taking them for automation and the team with help of an external agency, created metrics to select a process suitable for automation and to measure outcomes. Process Discovery tools such as AssistEdge Discover help in analyzing and generating inputs for metrics defined for evaluating contenders for automation.Process Discovery tools are applied at pre-automation stage to pick the right use case. They feed on user key-stroke actions on screen and data generated by a process in the past few months/years to create an automation blueprint. This displays the data flow and highlights steps where exceptions are encountered in a straight through process. Analytics at such stage will bring-out the opportunity for automation in a pictorial format that can easily be construed by the senior management before making an investment commitment. The challenge process owners may face, can be around the data missing due to an immature process. In such a case it is recommended that the process continues to be managed manually for a while, so that necessary information is collected for analytics.
- RPA Execution: RPA products such as AssistEdge, constantly monitor and record process execution. The data generated during this stage can be displayed on a personalized dashboard for businesses to monitor users’ action on a process, Digital Worker’s involvement and reasons for a process failure.A leading American bank extensively used analytics capability of a RPA product during process execution to spot scenarios where the Digital Workers failed and to develop more robust automation capabilities. The personalized dashboard was integrated with alerting systems to generate automated alerts in case of any exceptions or failures.
- Post Automation: A process once automated is moved to production in parallel, where it is closely monitored and supported in case of exceptions & failures. During this period, even for an unattended run, unplanned manual intervention happens to address the initial hiccups. Once the process has matured in production and bugs are fixed in parallel, it is allowed to run as planned. The process runs for certain period on production environment will decisively let CIOs measure ROI from RPA implementation.The data generated during the automation run can also be fed to a visualization tool for business intelligence. The German Bank referenced above applied BI tools to analyze such data on quarterly basis to generate MIS and to present quantitative information before acquiring new licenses of RPA solution. Personalized dashboard or BI is applied to capture utilization of Digital Workers (DW), considering the DWs can run 24×7, and to augment utilization of DWs across multiple processes/businesses.Latest offerings by the RPA industry in monitoring space allow end users to apply analytics to perform load balancing across DWs, optimize resource utilization and create real-time reports. AssistEdge helps in measuring and enhancing the RPA experience by orchestrating and supervising Automation implementation. AssistEdge also helps provide a unified view and measure ROI across multiple RPA Implementations in the enterprise.
So, when should an organization seek analytics in RPA? The pre-requisite for analytics is the data generated for a process. To enhance the value of RPA, quality data must be analyzed using the RPA adoption tools at the right intervals. Analytics will help firms shortlist the right use cases and decide whether to proceed with RPA.
Analytics will not just bolster the enterprise RPA strategy, but also ensure a successful adoption of the latest technology.
References:
1, 2 – https://www2.deloitte.com/bg/en/pages/technology/articles/deloitte-global-rpa-survey-2018.html
3 – https://www.ey.com/Publication/vwLUAssets/Get_ready_for_robots/$FILE/ey-get-ready-for-robots.pdf
Amit Kumar Sharma
Senior Product Manager, EdgeVerve
More blogs from Amit Kumar Sharma >