Any cocktail party conversation about AI in Silicon Valley inevitably comes around to the topic of SkyNet and the subsequent post-apocalyptic, dystopian future that will inevitably befall us as AI becomes more powerful. While the Terminator series does paint a bleak future for humanity, the focus on SkyNet distracts from another central theme, that AI (cleverly disguised as Arnold Schwarzenegger) working with mankind is the best hope for a prosperous future. However, having experienced multiple organizations struggle with automation, the idea of AI/human integration, while great in theory, is challenging in practice.
Organizations on the transformation journey should take the counterintuitive approach of embracing both change and the status quo. This idea is embodied by French philosopher Jean-Baptiste Alphonse Karr’s “plus ça change, plus c’est la même chose” literally translated “The more things change, the more they stay the same.” The notion that automation must work together with humans within many of the existing constructs of an organization, lays the groundwork for a successful strategy. This strategy must integrate the four realities of RPA, which will impact the majority of your enterprise processes.
Anyone that has worked with current RPA tools will tell you that they can be severely lacking, and that the chance of their automating one hundred percent of a process is between slim and none. A recent study by McKinsey echoes this, finding that on average about thirty percent of a process can be automated using today’s technologies. Even as this percentage creeps higher we have to deal with the fact that the activities automated are often not continuous, human judgement must often be interjected, and exceptions happen. All of this means todays automation technologies can’t support a fundamental rewrite of your processes, yet.
One the strengths of RPA is its ability to work with legacy systems, giving organizations the ability to get more miles out of their existing technology. It’s not just the systems here that matter, though. The overall enterprise IT architecture supporting the application of identity, change management, security, integrations, and infrastructure all are going to be around for a very long time.
If your processes aren’t going away, and your systems aren’t going away, your people can’t go away. While the actual number of people executing activities may drop, the groups involved in the process execution will likely increase. IT, Security, Risk, and Compliance will be more involved in the day to day operations, now that RPA is embedded in a process.
If people, process and systems aren’t going away, neither are the organizations that support them. While we can all agree that no organization is perfect, there are oftentimes decades of best practices embedded in their policies and procedures. Every component of an organization has evolved with a specific purpose that, while maybe not obvious, will most likely need to be addressed in the context of RPA.
When constructing our go-to-market strategies, we look at how can we lower the “switching costs” of our solution. A high switching cost means higher risks, higher expenses (impacting ROI), and usually a built in constituency that doesn’t look at change favorably. The same is true about an automation program. Every time the people, process, systems or org need to change, the switching costs are increasing. The key to the speed, and success of an RPA program lies in figuring out how to leverage components of the existing, while delivering change. It will at times be necessary to make large, transformational changes, but a land and expand strategy, while delivering value, presents the highest likelihood of success.