Imagine for a moment that you’re a mortgage analyst at XYZ Mortgage. Part of your job involves opening each loan application document emailed to you, opening the attached Excel file, copying the property address from cell A4 on the second sheet, and pasting that data into the United States Postal Service’s (USPS) address validation website. If the addresses match, all is well. If not, you need to find out how to reconcile the two. For the vast majority of emails received, the addresses match. The process seems boring, mindless, and repetitive. You find yourself thinking, “I’m pretty sure a robot could do most of this.”
If you’ve read any of the posts in this series, you’ll know that thanks to the power of robotic process automation (RPA), the answer to this question is an emphatic, “Yes!”
The above example is a textbook process that is ripe for RPA. It’s repetitive, simple, and very rules-based. A human is only needed for the infrequent exceptions. While this process may be easy to automate and save our worker a few hours a day, the real benefits of RPA come when the entire organization is full of RPA robots. Hundreds of employees are then freed up to focus on tasks that bring greater value to the organization and more fulfillment to themselves.
As beneficial as this transformation may seem, it can be difficult to carry out. According to a recent Deloitte survey of 424 companies, while 53% of them had begun the RPA journey, only 3% had actually scaled RPA to a level of more than 50 robots. This article addresses these issues, answering the question, how do you implement RPA at scale?
We at Celerity believe the answer lies in an application of our IDEAsm framework:
- develop clarity of purpose through stakeholder engagement,
- maintain unity of execution by collaborating closely with IT,
- empower teams through a Center of Excellence, and
- provide feedback loops through engagement with an implementation partner.
We’ll walk through each of these in turn.
Develop Clarity of Purpose
The first step in implementing RPA at scale is to ensure key stakeholders are engaged and providing clarity of purpose. This entails setting out the strategic vision for RPA in the organization. The same Deloitte article noted that the C-suite was the most supportive group among organizations that had implemented RPA at scale. Having executive leadership develop clarity of purpose for RPA implementation brings with it many benefits.
First, clarity of purpose provides credibility to the automation initiative. Individuals within the organization will see the implementation of RPA as a key step to digitizing operations and modernizing efficiently, allowing employees time to focus on higher-value tasks.
Second, clarity of purpose at all levels of the organization provides a vehicle of conflict resolution. At scale, RPA may involve the automation of processes for cross-functional business units, resulting in conflict. The Deloitte report notes that C-suite support can “cut through organizational barriers,” which is useful “where there is resistance from other parts of the business.” Rather than letting two business unit heads remain deadlocked while the project stalls, involving stakeholders further up the org chart can help resolve the conflict and keep the RPA implementation on track.
Maintain Unity of Execution
The next important aspect of implementing RPA at scale is the provision of unity of execution through a supportive IT department. A PwC survey of financial services firms implementing RPA found that the largest challenge to launching robots into production during the implementation phase was “technical issues with integrating with surrounding software”with over 80% of respondents claiming this was an issue. When we consider that one of the main roles of IT is to facilitate user interaction with the organization’s software, we see that having IT support of RPA implementation is key to scaling it quickly and painlessly.
While the address-verifying robot envisioned in the introduction may not require any IT support at the outset, robots that need login credentials to web or system applications would require IT’s help to get set up. As RPA scales and grows in complexity, we can easily imagine more of the latter scenarios occurring.
Further, while our address-verifying robot could be tested relatively painlessly and without any irreversible harm if things went wrong, a more complex robot might need a sandboxed environment in which to be tested to prevent disaster in case of poor test results. Here again is where IT can play a crucial role, setting up a unified testing environment for the organization.
Deloitte sums up this concept well, claiming “[t]he IT organization is essential in setting up a scalable and secure bot infrastructure—to ‘plumb’ robots into your existing systems—and can hold the key to testing systems, approving user acceptance testing, signing off go-live and incident management on live systems.”
Beyond C-suite and IT support, it is crucial to empower the teams that are implementing RPA on the ground. It may seem obvious, but before an organization can implement RPA, it needs subject matter experts. A KPMG Report lists several different models of RPA expertise concentration, but all have in common a unit solely dedicated to maintaining and providing best practices and governance.
The RPA center should serve as a repository and enforcer of RPA best practices. Whether this takes the form of online training, third-party reports, internal trial-and-error, or a combination of the above, the RPA center should ensure that best practices are found and adhered to throughout the organization. As an example, the KPMG report lists the thorough documentation of both the pre-RPA manual process as well as the post-RPA process as one of the key best practices to which RPA centers can hold the business units in order to optimize processes in the long run.
The RPA center can also provide governance to the RPA implementation process. This may take the form of ensuring alignment between RPA implementers and corporate-level policies, keeping costs under control, and avoiding unnecessary risk. Again, such a role may not seem necessary or salient when one employee develops a bot to automate address verification, but with hundreds of team members running bots on mission-critical processes or with each other, having strong governance from the outset will serve to head off many problems down the road.
The experience of ANZ, one of the largest banks in Australia, may serve to flesh this out more. To adopt RPA, ANZ first brought in an implementation partner that helped train an initial group of RPA power users, which currently numbers about 70. This group serves as the organization’s RPA Center of Excellence, the chief robot builders and repository of “governance, compliance, IT, HR and of course RPA-specific” expertise.
This group then trained a second group of about 150 business users who could spread RPA expertise throughout the organization. Now grown to over 500, these business users have served as implementers of RPA throughout the organization. This allowed ANZ to start out with a bang, deploying about 100 robots in the first six months of implementation and over 2,000 by month 22. While we don’t see each element of the RPA center present, we can at least see that the RPA center is full of individuals who can not only apply their RPA expertise, but can pass it along to others well.
If developing C-suite clarity of purpose for RPA, readying an IT team to support RPA, and establishing a core group of RPA power users all at the same time seems like a daunting task, you’d be right. Enter the fourth critical element of a successful RPA implementation: the beneficial feedback loops brought by a strong integration partner.
The Deloitte report mentioned earlier notes that 63% of the firms surveyed plan to implement RPA with an integration partner. This partner can serve a variety of roles “from turnkey solutions to collaborating to up-skill in-house teams and build internal RPA delivery capacity”. This series of articles has contained a litany of examples of RPA successes, most of which have been enabled by implementation partners performing those very tasks. From the training of internal RPA experts at ANZ to EY’s partnership with a top 30 US bank for several RPA solutions, few companies successfully implement RPA well alone. Kieron Gilmurray, Director of Robotic Process Automation, Blockchain, and Process Excellence at Pearson, says outside help at the onset can make the biggest difference: “[r]emember your first process…just isn’t about making money. Boy you’re going to learn a lot when you start with RPA. Jump in, do it; but go and find someone who can help you right at the very start.”
The positive feedback loops can also be critical not only for the training of RPA experts, but also for training the IT team that will support the RPA implementation. Deloitte notes that “[i]f IT teams are struggling, consider engaging CIO support to create a small, handpicked team of agile, digital-minded technologists to support your RPA implementation and help you successfully navigate the wider IT organization.”
One of the most important ways integration partners help with feedback loops is to get others excited about and boost confidence in the process. Phil Fersht, CEO and Chief Analyst of HfS Research demonstrates the power of positive feedback loops this way:
“I’ve seen success in RPA where clients have generally taken one messy process, got the tools out, put it through, ran a script, and it worked. And they showed off. They shared it with everybody. And then they ran another, and another, and another. That’s how it works. It’s one thing at a time, build ROI, build consensus, build a sharing inclusive environment around you, and things start to roll.”
RPA at Scale: The Art of the Possible
Implementing RPA at scale is a difficult task achieved by few firms so far. However, organizations that can provide clarity of purpose through engaged stakeholders, unity in execution through a supportive IT staff, empowered teams through an RPA center, and positive feedback loops with a strong implementing partner will position themselves well to transform their organizations and dominate their fields for years to come.
Next in this series is, “Where is RPA Headed?” where we consider the rapidly approaching future of automation.