By now most enterprise leaders are well aware of the benefits of robotic process automation (RPA) – financial savings, improved quality and a better customer experience, just to name a few. One of the more important benefits that often gets overlooked is the ability for bots to collect statistical and performance data for processes they are executing, especially when that data was not available when humans performed the process. In a world that is ever more LEAN and agile, continuously improving bots ensures optimization of resources. The best way to do this at scale is to use bots for the collection and reporting of the data they collect, which will require becoming familiar with the idea of “bots over bots,” otherwise known as BOB.
The concept of BOB is simple: deploy a master bot to manage and analyze the output of your bot workforce. A master bot could be configured to assess data points such as volume, error rate, business exceptions, process outliers, virtual machine utilization and processing time. Once the master bot analyzes the data, it sends alerts or notifications to human users.
Let’s use the input for a financial transaction as an example. Assume you have a bot that receives financial data on a specific form as the basis to begin its processing. If the data is not sent on the correct form, then the bot cannot process that work item. With BOB on the scene, the master bot can detect a higher process exception rate over time and notify a human of the problem. This scenario shows how BOB can easily recognize and translate data that was not being captured when a human was managing the process.
At first glance, it may seem that BOB has to work with unstructured data because information or data requirements are not clear, but with the right planning BOB can gain access to structured data. What does this planning look like? First, establish which data points are important to your automation program and include these data points in every automation. Basic metrics should include volumes, error reporting and processing times, but be sure to challenge yourself to capture metrics that are important to your business and will drive continuous improvement.
Once you have established the data points, build the bots so they can access the common module to capture data across your automation landscape in a reliable and consistent manner. This ensures all automations capture and report data in the same way, regardless of who is building them. Once implemented, you will have standardized data from which BOB can glean the insights you need to effectively manage your bot workforce and your business.
Just imagine the data that can be gained with this approach and how valuable it can be. For example, you will know how many transactions come in on the wrong form, how many transactions come in on a certain day of the month, or how many come in above a certain financial threshold. This data could point out seasonality trends that have not been noticed before or even identify a pattern of consumer behavior that differs from historical trends. Considering the wealth of data available as a result of automation, the value of these kinds of insights is indisputable. The key to making it happen is planning and implementation of a concept like BOB.
ISG helps enterprises configure their digital workforce and improve their data insights. Contact me to discuss further.
About the Authors
Tracy Lipasek is an experienced consultant with more than 20 years of experience in information technology, process automation, transformation, leadership and software development. Her experience includes work for EDS and HP. She currently is the Global Operations Director for RPA and Cognitive Automation for ISG.
Sam Best is a Consulting Manager in the RPA and Cognitive Automation practice at ISG. He is a leading expert on automation techniques, including automation configuration and architecture best practices. Sam’s expertise helps clients optimize their RPA efficiencies both in the short term and in the long term.