Robotic process automation is in the spotlight as the new big wave for business transformation. Office workers hand over boring, repetitive tasks such as data entry to software assistants, freeing them to focus on the more creative and rewarding aspects of their job.

But RPA is also exposing the dark side of business processes. “Process sediment builds up in businesses in the same way that sedimentary rock builds up,” said Guy Kirkwood (pictured, left), chief evangelist at UiPath Inc. “Digging through that so that you can actually become more efficient is very difficult to do.”

“All the accidents of history are really being brought up by RPA,” said Cathy Tornbohm (right), vice president of BPO research at Gartner Inc. “So, RPA is an opportunity for companies to link their digital dreams with their existing legacy nightmares.”

Tornbohm and Kirkwood spoke with Dave Vellante and Rebecca Knight, co-hosts of theCUBE, SiliconANGLE Media’s mobile livestreaming studio, during the UiPath Forward conference in Las Vegas. They discussed the payoffs and pitfalls associated with introducing RPA into the workflow, as well as whether the future of the RPA market will live up to the current hype (see the full interview with transcript here). (* Disclosure below.)

[Editor’s note: The following has been condensed for clarity.]

Knight: What are the trends and the themes that are most salient in the robotic process automation space right now?

Tornbohm: I think the most fascinating thing about RPA right now is that it’s really highlighting the problems that organizations have. Typically, less than 15% of the applications that they’re using have got some sort of application programming interface. So, if you don’t have a way of linking them, you end up with this long tail of applications that are linked together with people literally being swivel-chair integration between the applications.

Vellante: Gartner lists RPA as the fastest-growing software category. We have seen spending data that confirms that. Why do you think this is?

Tornbohm: It is the digital competition that companies are facing and the recognition that they cannot continue to be quite as bad at some of the things that they are bad at! So it’s really that business transformation story back again; business process reengineering. The same story that we had with business process outsourcing 10 years ago but now with just with robots instead.

Vellante: What are the metrics that matter when you talk to your clients? 

Tornbohm: Well, what I try and encourage clients to do is to really focus on business outcomes. I don’t really care how many scripts AKA robots you’ve built or how many runtimes you’ve deployed. What I care about is the business impact that you’ve managed to achieve on whatever KPIs are important to you. Are you managing to collect more revenue? Are you managing to make your customers happier because you’re managing to decrease average handle times or increase right first-time activities? Those are the sort of metrics that companies should be focusing on, not how many scripts they’ve built. That’s absolutely pointless.

Vellante: Sometimes things like Gartner’s magic quadrant and the trough of disillusionment are misunderstood. So, how should we think about the hype cycle?

Tornbohm: Gartner says is that all industries have to go through this type of growing pains. We’re seeing that UiPath expanded massively, and that’s always a challenge for companies as they grow very rapidly.

There are four reasons why companies are going to go into this disillusionment. These are the main challenges with companies trying to use RPA properly. One is they don’t know what the processes are. They don’t know them at a microscopic level, and they don’t know them at the macro level. They don’t know how to orchestrate their resources. Politics — people will lie to you about what they do all day, so they can sabotage your process. And there’s a lot of silos within organizations that hate each other and throw things over the wall. So that all needs streamlining. And the more you can do across silos, the more successful any automation project would be.

Then when you take a person out of a process, you take their eyes, their mouth, their nose; how you’re going to replace that. You need all these other activities replaced, replicated, supported. Then you’ve got the economics of production. So actually making sure that the scripts that you’ve built are actually worthwhile and are going to be cost effective is something that we’re studying at the moment. So you’ve got all these different barriers from all these different angles. And that’s why it’s great that RPA companies are looking at ways to mitigate that for their customers.

Kirkwood: So RPA is really good at dealing with structured data, rules-based activities, deterministic things. That’s why in regulatory, highly regulated environments, it’s very effective, and the regulators love this sort of stuff cause it’s deterministic.

When you look at AI, then we look at it in four ways. So you’ve got process understanding, which is [enabled by UiPath’s] ProcessGold acquisition. You look at conversational understanding, cause ultimately robots are going to be controlled by voice. So the system has to understand that, let’s say you’re sitting in bank and the robot doesn’t understand something. You say, “OK robot sticking it in the Wells account.” It has to understand that Wells in this case means Wells Fargo. It does not mean a hole in the ground with water in the bottom or the town in Somerset in the U.K., because they’re all Wells. So getting those ontologies correct is so important.

Another is document understanding. Because as Cathy said, companies are still wading around in paper. So, understanding what those different documents are and how to action them is going to be really important. And, finally, they’re looking at visual understanding, so understanding and viewing things in the screen in exactly the way that humans do. So it’s getting that combination right.

Vellante: Cathy, in your opinion, will RPA live up to the hype?

Tornbohm: I think the most successful companies are the ones that take the person that’s managing the data and analytics of how their processes are performing and doing that with their automation strategy. But there are very few companies that have actually worked that out. They’ve still got two walls, and they just meet up here at the CEO.

So, unless companies actually take more active business outcomes approach and look at their end-to-end processes of order to cash and source to pay, these problems will carry on for some time.

Vellante: How do you see the future for RPA companies like UiPath?

Tornbohm: Gartner’s new prediction series shows that roughly 20 to 30% of enterprise adoption of AI, machine-learning activities for process-based activities, will go through the RPA market and the intelligent business process management suites market combined together. Because RPA has managed cleverly to capture the imagination of the business person.

Kirkwood: One of my predictions for next year is that the challenger RPA vendors, and indeed the service organizations that are small, are going to continue to consolidate and get acquired next year. So that’s the 2020 prediction for us.

Here’s the complete video interview, part of SiliconANGLE’s and theCUBE’s coverage of UiPath Forward. (* Disclosure: TheCUBE is a paid media partner for UiPath Forward. Neither UiPath, the sponsor for theCUBE’s event coverage, nor other sponsors have editorial control over content on theCUBE or SiliconANGLE.)

Photo: SiliconANGLE