RPA: Scaling from 1 to 100 Bots

In this two-part series, we take an RPA adventure from the first bot to the one-hundredth. We look at the steps of growth and what to consider when embarking on your own Robotic Process Automation journey.

With its recent rise in popularity, organizations are eager to incorporate Robotic Process Automation (RPA) and other process automation technologies.

This is made easier due to approachable low-code platforms, free community software, and myriad training opportunities. But while many have dabbled, few have successfully taken their programs to maturity. Too often, they don’t even make it past a pilot phase.

There are plenty of resources highlighting the elements of scaled RPA programs, but that’s only part of the picture. How do you actually get there? In this two-part blog series, I will focus on the gradual progression to maturity – 1 to 100 bots – not just the end goal.

Each organization is different, of course, and the path to success will vary. Still, we’ll outline one possible progression based on our experience and observations. I didn’t intend it as a roadmap but simply a guide to inform your own journey and anticipate challenges.

For part one, we’ll focus on the first 1 to 10 bots.

1 Bot: Proof of Concept

Throughout this story, let’s pretend we work in the IT department of a typical company. We’re growing, looking for ways to optimize our business, and we know that there are process and technology inefficiencies everywhere.

Our IT team is tasked with exploring RPA, so we download the community version of a popular RPA tool and start fooling around. After some online training, we pick a simple use case from our finance team’s manual tasks. The finance team has been vocal about needing automation and process improvement, so it makes sense to start there. The goal isn’t to release anything to production, simply to get a proof of concept working and decide if the hype is real.

The first bot is basic and doesn’t handle errors, only the “happy path.” We don’t worry about infrastructure, security, or coding to best practices either. Still, we’re new, so it takes longer than estimated to build, it’s not fully production ready and it won’t generate a great return on investment. Nevertheless, we complete the automation and demo it to leaders around the organization. Just seeing the tool in action excites them about RPA’s potential. They can’t wait to expand the capability.

5 Bots: Opportunistic Solutions

IT Leadership assembles a small team to turn our RPA proof of concept into an actual capability. We call on the leaders wowed by our demo to generate ideas in their departments. Most of their candidates don’t fit RPA – not surprising since they don’t fully understand the tool yet – but we end up with promising use cases across two key departments. We start by targeting unattended bots that are easier to manage and run on infrastructure our team directly controls.

Next, we purchase RPA software, selecting a package with just the basics: a few developer licenses, a few robot licenses and an orchestrator command center to manage the platform. With help from our IT infrastructure team, we add the necessary servers and install the software. There’s a hefty investment in software and infrastructure now, so we’d better get busy delivering.

We hire an RPA developer who can hit the ground running and pull in a business analyst to help with requirements. The team builds out use cases on a shoestring but is able to deliver working bots. We engage functional users to validate the automations, then get help from our IT operations team to deploy them.

After a couple of months, we manage to launch five bots. There are speedbumps along the way, including missed process requirements, long wait times for system access, and not picking the best RPA use cases. Nevertheless, RPA is gaining steam.

10 Bots: Automation Center of Excellence

With more departments clamoring for automation, the need for structure and governance is clear. No more flying by the seat of our RPA pants – it’s time for a Center of Excellence (COE). We put new automations on the backburner and focus our efforts on key COE elements around intake, governance, delivery and operations. Since we can’t go from 0 to 60 with a small team, we prioritize the items that help us improve immediately:

• Solidify our company’s vision for automation and get alignment between the COE, IT and business departments. A clear vision helps prioritize how we spend our time and money.
• Create a formal process for delivering automation, from intake and prioritization to deployment and support. This way, we’re more efficient with delivery, more reliable to our business partners, and can set clear expectations for what the RPA team can do.
• Incorporate the right tools and technology to support delivery, including an intake portal, backlog and work management tracking and code source control. We also set up proper RPA infrastructure with multiple environments and security controls to match.
• Step up training to raise awareness across key areas of the business, not just the COE. By educating more folks across the organization, we’ll get better ideas for our backlog and get better support as we design, test and run automations.

Our biggest challenge is getting a commitment from those outside IT to participate in the full RPA lifecycle – an ongoing struggle we’ll continue to face. And though we still have a long way to go, the future looks bright. The COE emerged before our bots became unmanageable, creating a solid foundation to support what we had while building a backlog of future bots more deliberately. We can also leverage IT resources to support our growing infrastructure and ultimately provide quality service to the business users for whom we’re creating automation.

Conclusion

So far, our example company has delved into RPA using a lower number of bots, but each stage is important in our transition to a fully-fledged RPA program. Stay tuned for part 2, where our journey to maturity will continue as we scale all the way 100 bots.