RPA’s Limitations Are Everyone’s Limitations

Every new technology that comes to the market says it will do (whatever it is that it does . . .) faster, quicker, and cheaper. That was never truer than with the robotic process automation (RPA) vendors that promise in virtually every sales and marketing outreach that their automation software will automate tasks faster, quicker, and cheaper than anything that came before. But like a lot, if not most, enterprise software marketing, that statement is only partly true.

If an organization automates a rules-based manual process then, by default, an RPA bot will most likely do that activity considerably faster than was previously possible. But, like any software toolset, it has to be deployed in the right way to undertake the right tasks—taking into account how RPA impacts the downstream process and the computing load on other systems. Plus, it’s critical to analyze and improve the business process(es) and manual activities before throwing attended and unattended bots at business problems. Afterall, a bad process that executes at lightning speed can make things far worse. Also, keep in mind that task automation is all too often a band-aid that patches together a broken or inefficient process that instead needs better improvement and automation than RPA alone offers. 

The meteoric growth of RPA has been something to behold over the past three years. But now that task automation software has moved into the mainstream and is being deployed at scale, its limitations are coming into focus. Deep Analysis is not against RPA, but firms must deploy RPA wisely or risk failure.

One major limitation is that the speed of processing transactions automated by RPA tools is not defined by the RPA tool alone, but rather by the slowest application it works with. That is not a problem specific to RPA but one that few buyers seem to fully grasp before it’s too late. For example, this past week the US government, Small Business Administration division (SBA) had to stop processing applications for PPP loans that were generated by RPA tools for just this reason—the applications surrounding RPA set the pace of automation. Though the SBA got a lot of criticism for this (you can see our quotes regarding the PPP situation in the media here), it was really just a public example of an all too common problem.

Almost nothing works in isolation; every piece of enterprise software is interdependent, and interconnected. Each product is a cog that plays a role in a larger set of operations. It’s why we embrace “information supply chain thinking”—which means acknowledging and trying to understand the bigger picture that each enterprise buyer is trying to paint. Decades of siloed IT automation approaches have left us with a broken and fragmented legacy that is difficult if not impossible to stitch together, especially if using RPA as band aids. (This reality is one of the key reasons that our research and advisory services at Deep Analysis are not siloed into the definitive product categories typical of analyst firms.) 

Moving forward into a post-pandemic world, there will be—indeed there is already—a surge of interest to “do IT better,” to work in a more connected fashion, and to recognize what is of true importance versus what is not. Some enterprises will leverage process mining to better understand what is really happening and to predict and produce more efficient outcomes. Then, hopefully, they will use digital process automation and task automation to execute and improve those newly mined processes. Others will look to automate currently manual tasks, while generating clean data for AI & ML from the point of capture. Others still will get serious about deploying blockchain to reduce intermediaries and build secure/efficient networks of trust. All of these approaches, in combination, point the way forward for organizations if they look at the big picture for how these technologies work together.

All established organizations must deal with their legacy IT, with decades of hodge-podge implementations, custom development, and little appreciation for the bigger picture. But the good news is that broken processes that became painfully obvious this year will put business transformation in play for the better. Using RPA systems to hit the organization’s head against figurative brick walls are but a symptom of the work that is ahead of us. 

Work with us to ensure you are a disruptor not one of the disrupted! 

Get trusted advice and technology insights for your business from the experts at Deep Analysis. [email protected]

Leave a Comment