RPA, Harder Jobs and Better Metrics

Conventional wisdom has it that Robotic Process Automation (RPA) will make humans’ jobs easier and more fulfilling by eliminating the drudgery of routine and repetitive tasks. While that’s true in a sense, the real point of RPA is not to enhance a human’s job but – let’s be blunt – to eliminate it.

Consider RPA applications for claims processing in health care. The objective is to automate the activity involved in reviewing, analyzing and then approving or rejecting a claim. A high percentage of health care claims are routine and predictable and, as such, lend themselves to RPA solutions that follow clearly defined rules. For a potential duplicate claim, for example, if the provider ID, member ID and date of service for the claims are not identical, then the pended claim is not a duplicate.

Let’s say an insurer automates 60 percent of these highly routine, predictable and rules-based claims. The truly interesting work lies in figuring out how to manage the remaining 40 percent of claims that will fall into the category of exceptions. The challenge lies in identifying the tipping point where claims become atypical enough that automation isn’t economically viable. While you want to take advantage of RPA capabilities to increase efficiency, you don’t want to over-engineer the solution. Take, for example, a claim filed by a 20-year-old who lives with her father, but is covered by her mother’s policy, who lives in a different state. While it would be possible to write an RPA program that evaluates the intricacies of the rules that apply, such claims occur rarely enough that funneling them to a human claims processor who can consult company guidelines is more efficient.

Which leads to the key question of how RPA impacts claims processing staff. Again, there’s no denying the reality that RPA will reduce the number of people an insurer will require. A department’s headcount is often reduced by 40 to 50 percent, while staff who remain experience a dramatic change in the course of their daily work. Rather than spending five hours a day processing routine claims and three hours researching and resolving complex cases, reviewers now spend the vast majority of their time focused on assessing unusual cases that don’t fall into the category of typical case. This puts a premium on industry experience and knowledge of the insurer’s policy guidelines. In this sense, RPA does enhance the value of work by taking on the no-brainer part of the job (pun intended) and offloading the problems that require analytical skills and expertise to the humans.

Another issue to consider is how RPA redefines the metrics applied to operational performance. Traditionally, reviewer performance was based on claims processed per hour. In an RPA-enabled environment, that measure of productivity should be expected to decline significantly, since the routine, easy-to-resolve claims are taken out of the mix. In fact, in an RPA-enabled environment, a high claims-per-hour rate by human reviewers could suggest that the RPA solution is leaving money on the table and setting the exception bar too low.

It seems no one would argue with the notion that RPA is a game-changer. As the market matures and more and more solutions come on line, and as we benefit from experience and lessons learned, we’re discovering that it’s not just that the game is changing but that the rules themselves are being re-written.

About the author

Rod Dunlap is a business process automation pioneer with more than 30 years of experience. He first utilized Robotic Process Automation (RPA) techniques in 2008 to automate manual claims processing activities in healthcare.