AI and Automation technology is advancing at a faster rate than ever, and demand for these technologies in business is steadily increasing as well. Certain AI technologies such as Robotic Process Automation (RPA) can drive tremendous business value and are being heavily leveraged by businesses to digitally transform their processes.
RPA adoption is expected to grow significantly in the next decade. According to a Forrester report, the Robotic Process Automation (RPA) market is expected to grow to $2.9 billion in 2021. The market was valued at only $250 million in 2016. Many research firms anticipate that the RPA market will grow at a CAGR of over 36% and that the market size would cross $8 billion by 2023! This is game-changing technology for business.
The performance of any enterprise comes down at least to some degree on the quality of its processes. To succeed and remain competitive in any changing market, enterprises must constantly look to identify process improvement opportunities. Today, AI and specifically RPA can help companies digitally transform processes and drive improvement by increasing productivity, reducing inefficiencies, and enhancing accuracy.
As RPA proliferates, and implementation teams become more familiar with how to best utilize this technology, an important lesson has emerged. Successful RPA initiatives are heavily dependent on the Process Discovery component of a RPA roll out. It’s usually not difficult for companies to identify the first few processes that make sense to automate. Processes that are highly manual, repetitive, and consume lots of human resources are the obvious places to start. However, the question that quickly arises is, “where do I take this next?”.
A company implementing RPA is likely to have hundreds of eligible processes across many departments that can be considered for automation. It can be a challenge to qualify and prioritize processes in an automation roll out. But this decision is usually based on two factors: 1) the value achieved by automating the process and 2) the feasibility of automating that particular process.
But assessing lots of processes based on these two factors can be difficult and time consuming. Many companies today do it with the help of internal teams or external consultants. Technology firms or consulting partners help companies study processes and assess their potential for automation. This is mostly done through human observations, specifically by interacting with the internal stakeholders such as the Subject Matter Experts (SMEs) to understand the minute details of each processes. After gathering information about lots of processes from interviewing SMEs, RPA assessment teams then attempt to qualify and prioritize the processes to automate, and the right RPA technology to adopt accordingly.
The results from this traditional form of Process Discovery are usually mixed, and it has a direct impact on the success of a company’s overall RPA initiative. While RPA has tremendous market momentum, a recent study shows that almost 50% of enterprises implementing automation do not meet the desired outcome due to certain gaps in their implementation program. These gaps occur due to the difference in the way processes are defined vs actually performed in an enterprise’s day-to-day life. The gap is often between how SMEs think a process works and how it is really done in practice.
So why is the discovery phase a typical failure point for RPA rollouts? It has to do with the ‘human observation’ approach normally taken in order to discover which process to automate. Most enterprises have Standard Operating Procedures (SOPs) that define how key processes are followed. Over the years, SOPs go through many iterations and modifications. Logic for ‘why’ things are done a certain way, and changes to SOPs don’t get documented and are often long forgotten.
Thus, when a RPA implementation program is initiated, Process Discovery is often based on old information, information not in sync with how the process is really done, and highly dependent on information gathered manually from Subject Matter Experts (SMEs). Information from SMEs is useful, but sometimes limited and subject to human biases.
SMEs often ‘think’ they know how a process works, but may not know all the nuances or exceptions. Often important details of a process can be missed, which leads to improper assessment of a process’ readiness for RPA. This can also negatively affect the way RPA robots are configured to support the process. ‘Subjective’ information about the process is insufficient, thus creating a real imperative to identify empirical data about the process.
So how can companies add empirical data about processes to supplement subjective information from SMEs when evaluating them for RPA? There is now technology that automates the Process Discovery process itself, providing RPA implementation teams detail statistics and empirical data about processes. This technology can also perform analysis based on gathered information and help us judge process complexity, automation feasibility, exceptions, and potential ROI.
EdgeVerve recently launched AssistEdge Discover, a unique Process Discovery tool that automates the Process Discovery process. By monitoring humans non-intrusively while they execute a process, this important component of the AssistEdge RPA Platform helps create insightful business process maps leveraging user key strokes and sophisticated neural network algorithms. This in turn generates effective automation blueprint for processes based on empirical data gathered about the process.
AssistEdge Discover significantly increases the quality and accuracy of Process Discovery, especially for enterprises that are rapidly rolling out RPA and require agility and insights to drive RPA success. AssistEdge Discover is built to work across many applications in gathering process statistics, and can benefit companies rolling out any RPA tool on the market.
By simplifying and enhancing the Process Discovery portion of an automation journey, AssistEdge Discover can further help enterprises speed up implementation, better deal with change management, and realize the true value of automation.