Binary Thinking is for Computers, Not People (or Robots)
People are excited about Robotic Process Automation (RPA), and for good reason. RPA has the promise to improve customer outcomes and the bottom line. RPA will likely even help engage remaining workers in higher-value efforts, thus increasing their engagement; and engaged employees bring more value to stakeholders. RPA is indeed exciting new technology, particularly when Artificial Intelligence (AI) comes into play; an RPA bot can do more than make simple binary yes/no decisions. Some people, however, are lured into thinking RPA is a magic bullet, and boil down the RPA promise to a simple binary approach: ‘We automate with RPA for quick and easy wins, and we don’t need to run all those process improvement projects or take on big system transformation efforts.’
It’s important to keep in mind that RPA is a specific tool/approach that has strengths and limitations, and is best applied to those processes where we want the same outcome over many repetitions. So choosing the process to automate must be done carefully, and you don’t want to have an RPA bot try to do too much. A bot should follow the rule of 5 (per another Forrester paper): 5 or less decisions, 5 or less applications interacted with, and 500 or less keystrokes/mouse movements. RPA bots do take ongoing attention as well, so application of RPA must be looked at holistically. Keep in mind that focusing on automating a piece of the process may sub-optimize the process as a whole; remember to see the forest of processes around you as well as the trees/tasks in front of you (Seeing the Process Forest, and the Trees has more of my thoughts on this). Traditional process improvement or process excellence efforts, customer journey mapping, system transformations, and so on are additional approaches to the opportunity of improving your processes. RPA can thus be seen as one of many solutions that a COE can bring to the challenge of more effectively delivering value to stakeholders.
I’d like to turn now to a new paper published by Forrester. At the request of RPA vendor UiPath, Forrester Consulting recently published “Harness RPA To Optimize Customer Outcomes”. There are some great insights from the survey of more than 100 decision-makers on, or influencers of, RPA buying decisions. The first-listed key finding in the paper was “Consolidate or coordinate BPM and RPA centers of excellence (CoEs)… [Many Enterprises] already have BPM centers of excellence. These have expertise in process assessment, critical to applying RPA in the right spot…” Here again we don’t want to fall victim to binary thinking; it’s not an Automation COE or a Process Improvement COE. We want to either have both types of COEs that work well together, or one COE that handles RPA as a tool in the Process Excellence toolkit.
The Forrester report has some other interesting findings as well. I conclude that a solution like iGrafx can help with Governance of RPA. For example, the report finds that “Many robot environments do not have clearly defined roles and responsibilities. An upgrade to a core system, for example, may disrupt robot performance” If bots are defined in iGrafx as resources supporting a process, and the process also lists what other resources/systems are used, then reports can be created that help understand the impacts of change before they happen. In addition, when choosing an RPA vendor or process to automate, an important factor is: “Assessment frameworks to find the right processes for RPA.” It will be key that the organization finds stable, repetitive, well-defined processes to automate with RPA, and BPA/BPM solutions like iGrafx can help organizations do that. It’s critical to understand the processes, and their context, before knowing you’ve identified good candidates for RPA, and—perhaps more importantly—whether applying RPA on a particular process will achieve strategic goals and deliver stakeholder value.