Expert Insights
Scaling Automation with the Flywheel Model
With most agile organizations already having begun their RPA journey, the question on many business leaders’ minds in 2021 will not be how to start an RPA program, but how to scale one. While we have written extensively about scaling in the past, we have recently noticed a pattern emerging that could become the key to organizations reaching a more advanced RPA program with even greater agility.
The automation flywheel refers to a pattern in which the institutionalization and continued use of automation leads to ideas about the development of new automations fueled by both the COE and the end users themselves, causing automation to expand on a company-wide scale with incredible momentum.
This pattern is undoubtedly related to the rising hyperautomation trend, in which organizations automate as many processes as possible and use artificial intelligence, machine learning, and process mining to continuously discover, analyze, redesign and monitor those automated processes. The flywheel framework takes advantage of how everyone’s efforts in an organization impact each other, and uses this impact to build a self-sustaining automation program that will eventually enable an organization to reach a state of hyperautomation.
Where to Begin
Phase one begins with introducing automation to your company and making a coordinated effort to institutionalize it early on. Acceptance of a digital workforce can go more smoothly if an organization has already identified its evangelists – individuals from the executive team or any other area of the business that understand the power of automation and can skillfully and enthusiastically communicate it to other employees. Automations in this phase are built and deployed by the COE and are typically unattended, running independent of human employees. Automations in this first wave also usually fall into the “low-hanging fruit” category – high-volume repetitive processes with known bottlenecks that are low in complexity and thus fairly easy to automate.
Show, Don’t Tell
Once your organization has adopted automation and educated the employees on the benefits of a digital workforce, it’s time for them to see those benefits firsthand. In this phase, the COE delivers attended automations to the employees. These bots will work in tandem with their human counterpart to help them complete tasks and can be deployed across various areas of the organization.
The most common applications of these automations are in customer-facing functions like contact center and help desk roles, but we have seen attended bots can deliver value in functions like invoice management, timesheet submitting, payroll, insurance claims processing, and a multitude of other functions in which customer data needs to be pulled from disparate systems. (See our RDA vs. RPA White Paper for more on this topic.) As the employees begin using these automations, they develop a greater familiarity with the technology and will soon see the returns firsthand.
Process Identification
With a growing understanding of automation, employees will begin to view their daily work activities from a new “automation-first” perspective, noticing inefficiencies and areas of their workflows that could be improved. This is the breakthrough point that causes the flywheel to spin even faster – workers begin identifying processes and tasks suitable for automation, and want to share their suggestions for improving those processes. The employees then submit their ideas for automated solutions to the COE to develop and deploy.
Citizen Development Begins
At this point in the flywheel framework, after seeing just how significantly automation enhances their work, many employees find themselves asking, “What can be automated next?” Inspired by the possibilities, employees may begin experimenting with building process automations themselves. These employees enter a training program like UiPath’s Citizen Developer Foundation courses to learn how to build automations across common applications in their tech stack including Microsoft Word, Excel, Outlook, File Manager, and others. These programs are primarily aimed at business users with little to no development or coding background, so citizen development tools like UiPath StudioX remain easily accessible to all who wish to leverage the power of automation.
The prospect of citizen development can be daunting to some organizations from a security and governance standpoint, fearing their automation program could become a “wild west”. It is critical to establish that within the automation flywheel model, as with any citizen-led development program, there is still a governance framework in place. From this point forward, employee-developed automations are submitted to the COE for code review, refinement, security control checks and approval before being distributed to the organization.
Organization-Wide Distribution and Continuous Discovery
The COE’s continued distribution of citizen-developed automations across the organization leads to the regular generation of new solution ideas and an ongoing discovery of automation-suitable tasks and processes, accelerating automation and spinning the flywheel even faster. Following this model, an organization can circumvent many of the top challenges that arise when trying to scale an automation program. (See our RPA Pitfalls article for more on this topic.)
The flywheel’s momentum becomes self-sustaining, powered by the company-wide consumption of automation, constant discovery of new processes, and end-user involvement in solution design and development. Following this framework will lead to the creation of more automations on a more frequent basis, turbocharging the automation program and enabling RPA at scale.