Robotic process automation (RPA) is a groundbreaking technology — one with the potential to generate a quick return on investment and save a business hundreds of hours saved across thousands of transactions. As with any technology with transformative potential, though, the deployment project is critically important. A successful deployment establishes the processes and goals that will determine an organization’s future success with RPA.
However, experience shows that smart, disciplined deployment is easier said than done. A significant portion of RPA projects initially fail. So, the question is: how can an organization avoid the setbacks, roadblocks, and glitches that keep it from realizing the value of RPA as soon as possible? Here are several of the best practices we’ve discovered for a smart and successful RPA deployment:
1. Spread it around, but keep it focused
An RPA implementation team should include people who may never interact in the normal course of their jobs. Because of this diversity in roles and expertise, it’s important to define a clear vision that creates a shared sense of purpose. The leadership team should explain the opportunity and outline how RPA fits into the bigger strategic picture. By introducing the new technology and outlining how it can be used, the leadership team can get people excited about the project and encourage team members to think about potential impact of RPA on their daily tasks.
2. Plan for an unpredictable future
Adopting RPA is a journey. It doesn’t end with implementation. Companies change. Processes evolve. Organizations expand, and the scope of RPA will expand with them. Although business leaders can’t predict the future, they need to plan for it — and that includes RPA deployment. Companies need a flexible RPA roadmap that defines major milestones. They need a plan for building the required skills as technology changes. They need clear guidelines for choosing which processes to automate. And they need ongoing executive sponsorship and oversight. Many of the best practices from the implementation should continue throughout the RPA initiative — such as conducting pilot trials that ensure RPA delivers as expected with each evolution. Additionally, an organization needs to be prepared to scale up RPA itself and expand its technology beyond RPA to integrate with emerging technologies like OCR, chatbots, machine learning, and artificial intelligence. Combining these technologies can broaden the scope of RPA, allowing organizations to automate more complex processes and uncover additional value through extending their automation capabilities.
3. Remember the big picture
Robots can help automate manual processes, which is the primary source of value from RPA. But it’s not the only source. By implementing RPA, a business realizes several other advantages. Software robots can improve processing time, meet service level agreement (SLA) requirements, and enrich the customer experience. They can also increase processing quality by reducing errors. Robots can also increase process traceability and help ensure compliance. In other words, they drive improvements that may not be easy to quantify in dollars and cents but are still important to highlight to team members and stakeholders.
4. Keep business ops and IT on the same page
RPA isn’t solely a technology project or business ops initiative. It’s both, and it requires close collaboration between the two teams. In a typical implementation, the business ops team selects workflows for automation and monitors RPA effectiveness. The IT team focuses on testing and maintenance, data security, and software support. But these two domains should overlap, and a well-designed RPA governance platform can help bridge any gaps between the two. The right governance plan mandates communication and collaboration when obstacles arise, such as potential integration issues of RPA with existing programs.
5. Get strong on governance
Speaking of governance, companies tend to underestimate the work required after initial automation. They assume software robots will continue to run autonomously without support or intervention. While it’s true that RPA doesn’t deviate from its configured algorithms, the targeted processes and their supporting software interfaces and data formats often change. Companies need the governance processes to address the moving parts within the business and the ongoing need for planning, communication, and testing. Without a strong governance framework, RPA won’t deliver the scope of improvement the business expects.
6. Target the right processes
RPA isn’t a silver bullet for automating every business process. It works best when it targets processes that have been rigorously optimized and carefully selected by IT and business analysts working together. However, when organizations first implement RPA, their enthusiasm can get the best of them, and they may focus RPA on complex, non-standardized processes that require frequent human intervention. Automating an inefficient process only magnifies the inefficiency. This means that organizations have to put in the work of updating and streamlining inefficient processes before RPA can do its job.
7. Think granularly
As mentioned, not every process is right for RPA, and the same is true for the building blocks of every process. If some components of a process are highly unpredictable or require frequent manual intervention, they can be challenging to automate. Some may not need automation at all. In other words, not every subprocess is a good candidate for automation. The business needs frank assessments and clear-eyed analyses of which subprocesses are best left alone and which are candidates for optimization before putting RPA to work.
An RPA implementation doesn’t have to be a high-risk scenario with an uncertain outcome. With a well-designed, well-executed deployment plan, you can avoid the missteps that can drive up costs and extend implementation timelines. For more suggestions, see our comprehensive guide to your RPA journey.
Principal, Capgemini Invent
|Chet Chambers, VP, UiPath