Looking beyond the hype of robotic process automation
Recently, one of the most talked about breakthroughs in robotics and automation that’s aiming to transform the modern business is robotic process automation. RPA is a software technology that replicates how human staff members use applications at the graphical user interface level.
In short, it mimics and automates common, repetitive software tasks that are normally performed by a person, like data entry and migration.
RPA is particularly appealing for companies that are juggling with millions or even billions of transactions a day. With such an overwhelming amount to deal with, they often struggle to effectively manage important tasks like addressing customer requests, processing files, moving information between different systems, allocating work and making decisions.
But RPA promises to help some organizations alleviate the challenge and operate more efficiently by automating, and thus accelerating, transaction processing. They can then provide greater customer service, which inspires continued loyalty and has a direct, positive impact on a business’ bottom line.
It all sounds very alluring, doesn’t it? The promise of RPA is creating massive hype in the market, which is being leveraged by RPA vendors to position their products as the “silver bullet” for any company looking to streamline and optimize operations. However, as history has shown–from the Sony Betamax to Google Glass–sometimes things don’t exactly live up to their hype. Not to mention, much of the hype surrounding RPA ignores one critical thing – not all transactions are the same. In fact, the cost and complexity of each transaction can vary significantly.
Automation by Design
RPA can be thought of as “automation by design.” This type of configured automation takes over high-volume, repetitive transactions or tasks like migrating data between different systems and formats, i.e. migrating customer records from spreadsheets into a database. Once configured, it can input the data far faster than humanly possible, without breaks and with zero chance of error.
What’s more, the higher-end “robots” are even smart enough to recognize if data is missing and fix it. For instance, if a customer’s address doesn’t have a zip code, the robot can determine and enter it based on the city on record.
While RPA can certainly help with repetitive tasks like data entry and migration, there are other common, but more complex transactions that require something a bit smarter. In such cases, automation by design can be done by configuring dynamic workflows with smart business rules using intelligent business process management (iBPM) software.
For example, a reinsurance company can use iBPM software and set it up so that when it receives a customer request for an insurance claim under $50, a workflow will be triggered to check and make sure it’s not from a serial claimer. Then, if everything is clear, it will issue the check automatically. This can significantly reduce the amount of claims that need to be processed by individual staff members.
Rapid, smart and accurate, automation by design should ideally make up the majority of a company’s transactions, as the cost of processing each one is very low.
Optimized Human Decisions
After you’ve automated all the transactions you can via RPA or dynamic workflows, sometimes there are transactions with exceptions outside of the configured automation and established business rules. It’s at these times that a human decision maker has to step in and determine how to move forward.
Fortunately, they don’t have to blindly decide on a course of action and just hope for the best. Using existing data and algorithms, predictive analytics can show them what the different options are and how each respective decision would impact the company.
For example, for work allocation, predictive analytics can help operations managers with high-volume caseloads intelligently delegate cases to their departments and teams. The system will take into account each one’s skill level, current work schedule, cost of work and performance history. Thus, the managers can make smarter, faster decisions with the confidence that they will lead to the best business outcome. Such transactions that require a human touch are more expensive to process, but because these are optimized decisions, the company is better off in the long run.
Overall, the points discussed above are not to say RPA is a poor solution. On the contrary, it’s highly effective for the low cost, repetitive transactions and tasks that would otherwise consume a large part of your employee’s time and efforts. It can be a valuable component of any organization’s enterprise architecture. But, it cannot cover the whole transactional spectrum.
In reality, companies need to incorporate other technologies beyond RPA, like dynamic workflow automation and predictive analytics. Then, they can take a holistic approach to address all of their transactions.