In 2018, the controllership team at Discover was beginning to build the roadmap for a significant technology transformation. Knowing that some of our manual processes wouldn’t change until we changed some of our tools and technologies, we started exploring the possibility of using robotic process automation (RPA) to help us free up time for our staff to do higher value work.
My colleague Chris Allmaras, senior manager of RPA, has been leading the team on this from the start, building the finance RPA program from the ground up. So, I asked him to share some of his thoughts and insights from what he’s learned along the way. He says it’s critical to establish a vision that’s shared broadly across the organization: “You want the organization to embrace the digital workforce (i.e., bots) as helpful resources. Therefore, getting buy-in from stakeholders is essential.”
If your corporate finance team is thinking of building and managing an RPA program, here are six strategies that Allmaras and his team used that may help your success.
1. Encourage Participation
To encourage participation, be open about your goal. For Discover, that goal was freeing up human time to do the work that bots can’t. To help motivate staff to think about processes that could be automated, Allmaras’ team requested a list from staff of all the tasks they hated doing. To further encourage participation, the team started with a pilot, recruiting staff that they knew had manual processes and would also be happy to share their experience with others. With that first automation, Allmaras’ team put together a roadshow to present the comparison of bot versus human doing the task, showcasing the positive experience the accounting team had and the time savings they reaped as a result of automation. The key messages included in their presentation focused on showing how staff gained the opportunity to do work that’s more engaging while the bots did the rote activities and encouraging staff to broaden their thinking about what’s possible with RPA.
2. Build a Backlog
To ensure the ongoing viability of an RPA program, it’s important to build a backlog early and to continuously add to it. Processes that fit well with RPA technology and provide the most value are highly repetitive, high-volume processes with relatively low complexity. For example, tasks such as moving data, reformatting, and summarizing can easily be managed by a bot since they require specific instructions. As you build the backlog, consider whether RPA is the right solution, or whether a script, feed, or other type of automation might solve the problem at its root. Notably, process readiness is important—you don’t want to be wasteful automating a broken process.
3. Share Your Prioritization Criteria
Maintaining ongoing engagement across the organization requires that people believe there’s an opportunity to automate and that there’s a tangible benefit in doing so. If your backlog is being sourced across multiple teams, you’ll want to share your prioritization criteria so folks understand how their proposed automation will likely fit into your queue. To help people see what goes into the process, consider building an intake form that captures the relevant information needed to support your prioritization framework (e.g., time savings, complexity of the process, number of systems used, potential risk reduction, frequency of running the process, and timing of when output is required). An early automation that Allmaras’ team produced was an example of something prioritized based on the timing of the required output. The bots were able to eliminate the need for a human to pull a daily report at 5 a.m. While the process was simple and took very little human time, there was a significant morale boost in the team by eliminating the early morning task.
4. Follow Automation Best Practices
Allmaras and his team participate in several peer forums that discuss automation best practices that folks have learned along the way. One widely accepted best practice includes building automations with reusable components that can be shared across processes. For example, if you program a generic access point to a particular system, it allows anyone building an automation that leverages that system to copy the component into their process. Another best practice is to consider the data being used and ensure that sensitive data is masked or encrypted as appropriate. When it comes to the digital workforce, it’s best to consider the level of access to applications and think through access provisioning and segregation of duties. Additionally, it’s important to establish relationships and communication with the people supporting the technology applications that the bots access regularly to ensure there’s a clear and repeatable protocol for communicating changes planned for those applications.
5. Manage Ongoing Automations
Governance, control, and maintenance are key elements of a good automation operating model. To ensure bots are built with best practices for reliability and auditability, you’ll want to set up a governance structure that requires documentation, code review, and testing prior to putting an automation into production. For maintenance, it’s helpful to have dedicated resources to schedule and monitor the bots and troubleshoot production issues. Also, you’ll want to consider the level of logging you want as your bots execute automations—too much will slow the processes down, but not enough may create troubleshooting challenges or limit audit evidence.
6. Enable Expansion
A development team focused on RPA will build technical knowledge that enables shorter lead times, faster development, and quicker problem solving when issues arise. However, they’ll still have to prioritize their backlog and may not be able to produce as much or as quickly as the broader team desires. Training citizen developers, or people embedded within an operating team, is a great way to expand the scope of what can be automated. Citizen developers can combine their subject matter expertise with a basic knowledge of RPA to develop smaller-scale solutions that may not get prioritized by your organization’s team of developers. The work of a citizen developer should be subject to the same governance, review, and testing as the team of developers, so quality assurance can still be maintained. Also, a routine protocol for review of automations and meetings with stakeholders can bring to light other ways to deploy a process that’s already built, whether applying it with other teams or similar tasks or reusing component parts in different processes.
RPA can provide speed, consistency, auditability, reliability, accuracy, and productivity at a reasonable cost. Just imagine affording your team the opportunity to have bots work in the off hours and provide outputs ready for their use at convenient times. By framing and communicating a vision and strategy for RPA within your organization—and supporting it with a strong operating model and well-defined roles—there’s no doubt you can successfully help your organization embrace the many opportunities that RPA can offer.