RPA Optimization

Robotic Process Automation (RPA) holds great promise. But business are still not receiving the benefits of RPA.

By Shashi Shashidhar

Head of Intelligent Engineering Operations at Centific

Robotic Process Automation (RPA) holds great promise. Businesses undertake RPA initiatives to improve operational efficiency and achieve potential cost savings. According to Deloitte, a company that deploys 500 bots at a fully loaded cost of $20 million could potentially realize $100 million in savings. According to this report, some enterprises are showing 200-percent ROI on their intelligent automation initiatives. But the reality is that for every enterprise that achieves ROI, many more do not. In fact, EY says that as many as 30-percent-to-50-percent of initial RPA projects fail. Why?

In fact, many businesses tell us that their RPA efforts aren’t delivering value and expected savings. RPA initiatives fail to provide ROI for various reasons. The top two reasons are related to:

Automating the wrong processes.
Inefficient bot operations.

Automating the Wrong Processes

Organizations rely on hundreds or thousands of processes, both manual and digital. These processes need to be reviewed and analyzed to identify the right candidates for process re-engineering and automation. Processes that are repetitive, rules-based, and static are prime candidates for automation. Processes that are dynamic and that have a lot of exceptions and variants are generally not good candidates. Moreover, processes with minimal organizational impact may not yield the benefits that are expected.

In addition, automating processes that haven’t been thoroughly optimized can result in an inefficient spend. Part of the problem is that organizations often try to automate complex processes rather than streamlining simpler processes that lend themselves to RPA. Moreover, automating an inefficient process will only make the inefficient process run faster. It’s like garbage in, garbage out: pick the wrong process, and you deliver an unsatisfactory outcome.

We suggest that organizations have a good understanding of the process prior to automating it. The chosen candidate processes need to be first analyzed and re-engineered to remove inefficiencies. From there, the business should remove all nonvalue added activities in a process to improve flow. The next step is to look for activities that could be automated. Process mining techniques are helpful in analyzing the operational processes and prioritizing them.

Inefficient Bot Operations

Organizations also need to carefully consider the efficiency of bot operations to ensure ROI with their automation initiatives. In particular, scalability of bot installation and infrastructure is very important. As more and more processes are automated, the infrastructure should be able to scale and support them seamlessly. Maintenance and operation costs of bots can also eat up the savings if not properly managed. Dynamic processes and bots related to dynamic applications need higher maintenance due to their very nature. Accurately sizing and managing the infrastructure and licensing is key to realizing the expected benefits.

Want to learn more? Contact us We’re happy to help you as we have for many businesses.