At UiPath’s ‘Reboot Work Festival’ this week, users including Ernst & Young (EY) and Xerox painted pictures of RPA ambition and scale that might silence a sceptic or two.
One of the most common criticisms leveled at automation is that it is very hard to scale. Given that at least two global organizations have thousands of automations and bots providing value every day… what are they doing right maybe others aren’t?
At this week’s online UiPath ‘Reboot Work Festival’ some intriguing answers came through—and not all of them align with familiar RPA cliches.
For a start, we’re often told to start small and go for small, big wins. But $7bn print and digital document products and services Xerox did exactly the opposite—choosing some hard, horrible workflows to battle-test the idea. “Right off the bat, Xerox and UiPath collaborated very heavily on some very complex use cases–a reflection of the legacy application landscape that we were dealing with,” its Chief Digital Officer, Steve Miller, told delegates.
Sparking a lightbulb moment over management heads
For example, the bots Miller and his team wanted to build first were for managing legacy billing systems to make invoicing and billing more efficient to customers. By setting the bar high and achieving demonstrable value from non-trivial business processes, he says, the narrative about the value of RPA was immediately changed at the company.
“I think our ability to come in and automate in some of those intractable places really was the ‘lightbulb moment’ here,” he states. Now, Xerox bots work with both the internal team and customers every day across every function, from HR to marketing to “create a digital way to work to make us more efficient, productive and make us easier to work with our customers.”
Another global RPA UiPath success case showcased at Reboot, consulting giant EY, where north of 120,000 attended automations is used by its teams and customers, perhaps started more conventionally, with around 30 initial bots. But, its Director of Enterprise Technology John Russo, stressed, these modest ideas were very quickly challenged:
“We did proof of concepts and pilots, about 20-30 bots, realized the power, and then said, Okay, how do we scale? So, we reached out for to our consulting organization and asked, said what are we doing with our clients? What’s best practice in the industry?”
By forcing the pace and moving rapidly to industrialization of bot use, he said, EY was inside 18 months able to scale to 500 bots and very quickly, 1000. “We really built out that environment, built out that process, and that engine,” he says—work his colleague, Simon Constance, was almost immediately used to provide direct value on actual client-facing projects. “When we realized the power of the opportunity, the next thing for us to do was to scale that up across the whole organization, starting in a division with 125,000 people and ran a program where we consistently rolled automation out across every back and middle office function, looking to find opportunities using end-to-end automation via the attended, the unattended, the chatbot tools we had, and the ML tools we had.”
For Constance, putting all that through the business on a rolling cycle of agile development soon turned RPA at EY into a global way of working, so perhaps failing to scale is more down to organizational confusion about what to do with RPA than any inherent limitations—demolishing a common snipe at the approach.
A dollar in, dollar back use case metric
Another interesting real-world contrast to pedestrian RPA thinking was how these two organizations measure its contribution, which is often seen simply as FTE reduction. At EY, for instance, automation is being used as a challenge to the very way the company makes money: “We’re working with our core executive leaders on how we challenge our end-to-end processes that the firm delivers on,” Rosso confirms. “The team’s asking local senior business leaders, what are the things that are giving you difficulty today? Now we’re starting from what really matters to them, we solve that problem, and then we get into the bigger issues that may require them to make quite significant changes in the way they work, the processes they run or even the operating model part of the business. I think that’s been a big shift to allow us to scale up.”
And at fellow global RPA user Xerox, RPA’s judged on a strict business basis: ROI. “If we’re going to deploy a particular bot, then its use case must stand on its own. And when I say ROI, that’s monetized value—there must be a dollar return on investment, though we also look at potential productivity we could gain, as it’s very important for us to be making sure that our employees have the latest technology to make their day better.”
