Robotic Process Automation’s ability to automate mundane tasks has encouraged many companies to reap the benefits of intelligent automation as it transfers repetitive and costly tasks to computers so employees can focus on more productive work. However, the appeal to automate has rushed (or completely abandoned) strategic planning efforts, causing poorly executed implementations that decrease the initiative’s ROI potential. But while poor implementation is often blamed for these shortcomings, projects tend to go awry in the early phases well before implementation. The second most common reason why robotic process automation (RPA) initiatives fail is from selecting the wrong automation candidates. Selecting the right automation candidate is often taken for granted as people tend to approach RPA with specific tasks or automation objectives in mind.
Failing to critically analyze processes from end to end for automation potential can doom an initiative from the beginning, but with the added consequence of not realizing the problem until the project is underway or even fully implemented. So which processes and tasks should be considered, and which should be selected for automation?
Understanding RPA: What Can It Do?
To identify which processes to automate, you must first understand RPA’s capabilities. Robotic process automation software automates repetitive and predictable rule-based tasks that are currently performed by employees. RPA can read databases, log into applications & interfaces, copy & extract data from different sources, move and send files, fill out forms, and much more. While this list may seem all-inclusive, it is important to realize the limitations of RPA when considering which processes to automate. A process may be repetitive and monotonous, but if the steps performed are not consistent or do not follow a fixed set of business rules then the bot will frequently fail.
Elaborating on the diverse capabilities of RPA is a wide-ranging topic, but we will focus on how RPA’s capabilities and limitations should be used to determine the automation potential of a process or task. See below a list of considerations when prioritizing which processes to avoid and how to select candidates for your RPA initiative.
What To Look For In Automation Candidates:
- Make sure the process consists of repetitive steps.
- Processes that are rules-oriented and follow a fixed set of business rules.
- Processes that are performed by several employees or high-dollar executives multiple times a day, week, or month to ensure the automation yields enough ROI to justify the cost of its implementation.
- Processes with many pain points that provide qualitative benefits in addition to the ROI.
- Though it does not affect the ability to automate, be aware of processes interfacing with multiple applications and involving back-and-forth application interactions, as this increases the complexity of the automation
- Processes needing to have SALs (Service Level Agreements) met for a huge volume of transactions
What To Avoid When Selecting Automation Candidates:
- Avoid automating processes that are hosted in or linked to legacy applications that will be depreciated soon, as you are less likely to realize the full ROI and will be incapable of scaling your automation initiative in the long run.
- Do not automate a flawed process. While automation may seem like the quickest, easiest, and perhaps even the cheapest way to improve an operation, it is always better to optimize or reengineer a process before (or instead of) automating a broken system.
- A process may check all the boxes of what to look for in an RPA candidate and seem to be an ideal process to automate, but if it only takes 30 minutes for an associate-level employee to complete then the cost savings would never provide a return on investment.
- It is important to understand all the steps in the process as well as the impact it has on the department and/or the organization from end to end. While it may seem obvious, people often overlook the big picture, and increasing the speed and efficiency of a few steps in a process only to get bottlenecked at the next step/phase/department does not really improve your operation.
- While using OCR (Optical Character Recognition) software to read, transcribe, and format hand-written documents is a useful and common automation capability, it is another type of automation that is best suited for an experienced automation team and less ideal for an introduction to RPA.
- Processes involving human judgment should be avoided for RPA proof of concepts and first-phase implementations as hybrid automations are more successful when implemented when the organization has an established and experienced automation team or with the assistance of RPA implementation consultants.
While this article has served as an introduction to RPA and how to approach it, there are still several considerations to be made once you have created a shortlist of RPA candidates. Is now the right time to automate? Which RPA vendor is right for you? Who in your organization will lead this transformation initiative? What is your objective in implementing RPA?
There are several questions you need to ask yourself before implementing RPA but starting with realistic expectations and understanding which processes can and should be automated is the first and most significant step when undertaking an RPA initiative.
Process prioritization and selection can be difficult if you do not understand your organization’s processes end to end or if you are not familiar with RPA technology and its capabilities. Discuss the opportunity with leadership and with colleagues in your company to understand which processes have automation potential and to assess the appetite your team has for implementing RPA. Depending on the complexity of the initiative and the scale, you may want to consider using an experienced RPA implementation consultant to assist with selecting automation candidates, creating an RPA implementation strategy, and implementing the software in a way that ensures your organization can maintain and scale RPA independently. Either way, embracing automation can produce both quantitative and qualitative benefits when started correctly with the ideal automation candidates.