In September I had an exciting opportunity to present one of the keynotes at the annual Reuters Insurance AI and Innovative Tech Conference. Attended by over 1,000 insurance executives, this event covered how emerging technologies are unlocking a new world of data potential and shifting the industry’s competitive landscape.
For my presentation, I focused on robotic process automation (RPA), and specifically why many automation implementations fail to live up to expectations.
Why focus on RPA? It is a technology full of possibility to help organizations improve their processes, but it’s not without its share of challenges. RPA has grown at an incredible pace over the past few years (Gartner estimates the RPA market will reach over $2B in 2021), but the complexity of scaling from pilots to broader enterprise implementations is often cited about the technology’s future prospects.
My presentation focused on this topic. Titled “4 Mistakes That Are Ruining Your Automation Efforts (and how you can fix them)”, I walked through an example of utilizing RPA for a claims management process, and discussed four common mistakes I often see that limit the potential of to improve that process.
Watch Skan's keynote demo and presentation on-demand
It is critical to understand the details of a process before starting to automate.
Why? One of the fastest ways to cause your automation project to fail is by not accounting for exceptions and variations in a process. Additionally, processes may need to first be redesigned before considering automating them. As the saying goes, automating a poorly designed process just makes bad things happen faster.
The challenge is that many of the existing process discovery tools and approaches don’t deliver a comprehensive view of processes. Manual process discovery and consulting interviews only uncover anecdotal views of a process and often miss process variations and exceptions. Process mining tools deliver a data driven view of processes, but miss many critical process steps and details. Task Mining is able to provide a high level of detail about individual process steps, but struggles with building an end-to-end view of a process.
(You can read about each of these approaches in the following blog posts Why Manual Process Discovery Doesn't Work, What's Wrong with Process Mining?, and The Pros and Cons of Task Mining. Or, check out the Process Intelligence Playbook for a comprehensive, end-to-end view of process discovery tools and technologies)
Without the full understanding of a process, it is hard to automate successfully.
The second mistake organizations make is not basing their automation strategy on a solid, quantifiable business case.
During my presentation I gave an example of a claims management process. Many insurance executives could look at the process on the PowerPoint slide I showed and immediately identify common areas for automation. Process steps like viewing a PDF template for a specific piece of information or repetitive data inputs into a spreadsheet are classic examples of where intelligent document extraction and automation can help streamline the manual repetitive steps in a process.
However, building an automation business case is about more than identifying areas for automation, it is about quantifying the impact automation will have on key metrics. That means understanding the exact time and effort that will be saved as well as the impact on customer experience and business KPIs.
Too often automation is treated as a hammer in search of a nail. However, the goal should not be to find as many automation opportunities as possible, but to improve the process in the best way possible.
Automation is one way to achieve that goal, but it is not the only way. Strategies such as process redesign, identification and implementation of best practices, training, and even the introduction of new digital tools are all possibilite opportunities for improving a process.
Don’t start with automation as a goal - let data inform the best opportunities for improving a process.
Quantifying results and ROI is particularly important in the early days of automation implementation or pilot to show early results and success to build the future business case for even more resources to scale.
While process metrics such as efficiency, throughput, and cost/effort are common methods for measuring the impact of automation, business metrics such as customer satisfaction and revenue also need to be considered. Additionally, the amount of time, effort and rework involved in building the automation itself also needs to be accounted for in evaluating the success of an automation strategy.
Next Steps: Overcoming these challenges and mistakes
Interested in learning how to overcome these challenges? Watch my presentation to learn how Skan helps customers build a data-driven automation business case and quantify ROI with AI-Powered Process Intelligence.
“One of the things you don’t ever want to do is to automate a bad process. You are just going to make bad things happen faster, and that is not what anyone wants.”