By now you have heard many extol the wonders of Robotic Process Automation (RPA). It is true that RPA can provide much efficiency and savings within the processes for which they are deployed. However, it is also important to note that RPA should be just one tool, and not THE tool, in your process improvement belt. Knowing when to use that tool is just as important as the tool itself.
When starting to think about RPA there are multiple points to evaluate. In this article we will discuss key factors to consider when assessing whether or not to pull RPA from your toolbelt.
As I am sure you have figured out by now, the implementation of RPA without an accurate understanding of your process is likely to set your project up for failure. To do this, take the time to capture and document the process. You can even jumpstart this capture with process mining tools like Cognitive my-Invenio. These types of tools can help expedite the process discovery exercise and provide you with a real-world example of what your process looks like. They can also provide insight into the duration of various steps and the systems interacted with. Please note that these outputs should not be considered as gospel. They are merely the starting point. Process mining will capture the happy paths and the few daily deviations. Going one step further and outputting that data to a BPM tool such as iGrafx will allow you to review and add in the known variations for disaster recovery that compliance auditors will look for. I know that this is not sexy and takes time, but the end result will be a fuller understanding of the interactions, dependencies, and impacts that any automation will have on the business. Remember that your process is not likely yours alone. There may be up or down stream impacts, where a simple change may set off a ripple effect throughout the entire organization. If you want to show true return on investment (ROI) for your project, you must start with process discovery and capture.
As we discussed in a previous article, RPA does not automate the end to end process but tasks within it. Now that we understand the process, we have clearer visibility into which tasks are truly repetitive. We also now know and understand the complexity of those tasks, their durations, resource requirements, and how many systems are involved. A task that is a good candidate for RPA will be repeated many times, have very little complexity (requiring only a few decisions and system interactions), and high human touch time. For example, the data entry of application or order information from a PDF into a system, or the running of a report based on defined and unchanging criteria. These tasks are simple and may be repeated several times a day taking up valuable human resource time. You can identify these as ideal candidates for RPA.
Now that we have a set of RPA candidates we must remember that not all are created equal. We must have a way to prioritize this work based on strategic alignment, gains realized, and the cost to implement. Many have rushed to capture “low hanging fruit” but do not account for strategic alignment. As a result, they saved time and money on an activity no one cared about. By using a true BPM Platform, now we can map these initiatives to strategic goals and ensure the work we are doing is valuable to the business overall. This will improve visibility with tangible ROI and help to drive future projects.
Recently we conducted a study with the PEX Network wherein we asked over 150 business executives in 40 countries about how they decide which tasks to automate. You can download our infographic of the results here. If you want to read more about RPA and Workflow Automation, we have a whitepaper prepared for you here.
Now that we have the processes captured, the tasks assessed for RPA’bility, and prioritized for implementation, how do we control those implementation and bots to ensure governance and compliance. We will discuss that in my next article.