RPA Challenges and Lessons Learned

Image of RPA Challenges

SIG University  Certified Intelligent Automation Professional  (CIAP) program graduate Sharon Shao shares her experience with implementing RPA projects and describes the different challenges and lessons she has experienced.

A successful Robotic process automation (RPA) streamlines workflows, which makes organizations more profitable, flexible, and responsive. It also increases employee satisfaction, engagement, and productivity by removing repetitive, manual tasks from their workdays. RPA is noninvasive, can be rapidly implemented to accelerate digital transformation, and potentially transform businesses of all sizes.

However, RPA does have its challenges. It results in duplicate work effort, disruptive processes, and employee frustration if not overcome. In this blog post, I would like to discuss two common challenges covered in the course and experienced from our end during the implementation of RPAs and the lessons learned to overcome these challenges.    

1. Limited IT resources

One of the biggest challenges of RPA is the limited IT resources. RPA is a software technology and requires a unique skill set that combines technical expertise with a specialty in a certain business area. This can be difficult to find. Meanwhile, a large volume of RPA pipeline might be waiting to be assigned to designers.    

To overcome this challenge, we outsourced some of the work when we had a backlog of the pipeline. It worked well for simple and straightforward processes. However, for more complex projects, the knowledge transfer could easily leave a gap. When the Bot process runs into an issue that requires a code review or fixes, our in-house expertise normally finds it challenging to fix the issue either due to the different design platforms used for designing or the knowledge gap that existed during the knowledge transfer.    

Another solution we applied is the federated model. The Finance System and Solution (FSS) team has been working closely with GPO, business, and IT teams on RPA finance portfolio management and taking part in the government role. The team has combined knowledge of system/technical aspects and finance/accounting areas. By providing training and technical support to FSS team members who are interested in RPA design, they are able to implement RPAs for the low complexity and low time saving yet closing critical processes.

2. Process Impact

Understanding the end-to-end processes includes evaluation of one, the preceding, succeeding, and parallel process to the one being automated, Two the, access to the system/application, and Three, data availability for testing and comprehensive testing scenarios availability. It is important to carefully evaluate all factors to determine whether the processes to automate are a good RPA fit.

When evaluating an RPA opportunity, the team may overlook some system limitations on implementation. For example, we had to withdraw an RPA project requested by the AR team due to challenges encountered by the third-party managed (Del-SAP team) application – Ariba portal. The process to be automated was to submit invoices and attachments in the Ariba Portal for the ones that do not require invoice details verification. The team had performed initial validation on portal access availability, lower instance, and data availability and provided a green light to move forward. However, during the design phase, the designer still faced multiple changes related to the Ariba portal and the testing data availability.  

  • As Ariba is a SaaS-based application, the designer has experienced unexpected behavior and frequent layout changes, which caused very unreliable UAT results.
  • Lower instance data is not being updated and has a big variation from the production data.  

As this application is a third-party managed (DELL-SAP Team), It was very challenging to get their buy-in to support the test data update.    

The only option left to us was to build, test, and deploy in the Prod environment directly; however, as directly building (fine-tuning the code) in production with edit access is high risk and the process has customer impact, after evaluating the risk and level of effort to mitigate the risk, the COE team decided not to continue this RPA.    

Other challenges may involve confidential data handling, business engagement, etc. It is important to be aware of these challenges and to take steps to overcome them. As an RPA GPO, I plan to apply the lessons learned about RPA challenges in the following ways:

  • I will work with the team on resource allocation and make sure to utilize IT resources and FSS team members’ skillset wisely. Our federated model helps to achieve the adequate assignment of the RPA projects and keep them all in-house.
  • I will work with my team to identify the right processes to automate. We will consider the factors listed above to make sure that we are automating the E2E processes that will take into consideration access control, lower instances, and adequate testing data availability.

The Certified Intelligent Automation Professional program is a six-week course delivered through SIG University ‘s unique education platform. Visit our website to learn more about intelligent process automation and enroll for the upcoming semester.

Sharon Shao

Sharon Shao

Finance Manager/Global Process Owner, VMware

I'm Sharon Shao, the record-to-report global process owner (GPO) and Finance RPA lead at VMware Inc. at Palo Alto, US. My major responsibility is to lead global accounting process assessments, to drive accounting process design, and to work on process improvements from the end-to-end by working with cross-functional partners. Our team has been taking the governance role for RPA finance portfolio management since 2018. I work with associated finance track GPOs to collect RPA ideas, evaluate the RPA automation fit, prioritize the automation projects within the finance RPA portfolio, and monitor the approval workflow.