Getty Images Plus

RPA market booms as enterprises automate with bots

The RPA sector is expected to reach nearly $3 billion this year as new vendors compete with larger established RPA specialists and tech giants for enterprise automation business.

The RPA market is booming as enterprises look for new tools to create more efficient digital workflows.

Spurred in part by the COVID-19 pandemic, enterprises are moving quickly to automate time-consuming manual processes, both in the back office and in outward-facing areas such as sales and customer relationship management.

Forrester Research has estimated the size of the robotic process automation (RPA) market at $2.9 billion this year, with at least 200 vendors -- including tech giants such as Microsoft -- vying for the business of enterprises looking to automate tasks typically performed only by humans.

In an Aug. 13 report on the state of the market, lead author Craig Le Clair, a Forrester vice president and principal analyst, said AI will be a crucial part of RPA's future.

Gartner also published a report on RPA with a similarly expansive estimation of the strength of the market.

Le Clair, author of a book on AI, automation and RPA bots, also said intelligent RPA will be one of the key drivers in enterprises' evolution toward operating fully digitally.

In this Q&A, Le Clair talks about intelligent automation, digital transformation with RPA, and whether specialized RPA vendors such as UiPath and WorkFusion deserve their lofty valuations.

In this market, how do you distinguish between intelligent automation and RPA?

Craig Le ClairCraig Le Clair

Craig Le Clair: RPA is just task automation that takes some of the time out of back-office operations. When RPA looked at where it had to go to justify the value, it had to bring in more intelligent components because traditional RPA has no machine learning aspect. Intelligent automation is RPA bots that are directed by machine learning and conversational intelligence. Intelligent automation is the larger umbrella that includes not just RPA, but also traditional workflow automation or BPM [business process management] and chatbots, machine learning and text analytics for document extraction.

What would workflow automation in a digitally transformed enterprise using RPA might look like in, say, five years?

Le Clair: We're using the term workforce automation. It's a combination of human workers and digital workers. This is the way that a UiPath gets a $35 billion valuation, because of the notion that someday all workers will have a digital assistant, a bot, that will be more intelligent than RPA bots today, because it will link in with all these other AI technologies. It will be their own personal robot.

The idea of bot orchestration in this context is that you need to manage work in processes differently than you do today. Today, you have a simple workflow. In the world of workload integration with a hybrid workforce with digital workers, now you need to get a richer, more dynamic context of what needs to be done, and that's using more advanced analytics, and machine learning. You're not figuring out the routing logic for work. You're using the pattern matching of machine learning to understand what work needs to be done. And then you direct humans or digital workers to get the work done.

What are the advantages for enterprises to engage a specialized RPA vendor or a larger vendor that provides RPA services among many other enterprise applications and services?

Le Clair: The number of use cases for RPA, the reason is it has a lot of value, [is the variety of industries] that can use automation or the hybrid workforce to their benefit. If you are a company that sees that, your goal would be to try to have some kind of standardization of RPA technology and get the support for doing that. The reason you would use [a specialized RPA vendor] is if users are operating within specific domains almost exclusively with all the data being manipulated within that cloud data core. The enterprise architects are making difficult decisions about whether to automate using Microsoft RPA or a point solution. You might decide that you're going to let Microsoft do all the business developer stuff, the low-complexity automation done by the workforce at the edge of the company, but for the complex use cases, I might use UiPath. There's been a proliferation of the number of RPA tools and companies. And that's a chaotic situation given some of the risks associated with having bots that are operating in your core systems.

Technical debt is a tremendous barrier to transformation. The architecture that RPA has, to manage thousands of robots, is exactly the architecture you need to do workforce orchestration of the future.
Craig Le ClairVice president and principal analyst, Forrester

Regarding this huge market that you predict to be nearly $3 billion next year, are you saying RPA vendors are overpriced?

Le Clair: I don't think so. I talk to companies all day. They are really held back by the pain that RPA alleviates. There are 60% to 70% of applications that don't have any APIs at all. And a way to get to that is through the UI [user interface] integrations RPA provides.

Technical debt is a tremendous barrier to transformation. The architecture that RPA has, to manage thousands of robots, is exactly the architecture you need to do workforce orchestration of the future. There's a ton of automation to do. And the pandemic certainly made the whole market more valuable. Many have been calling for deflation of value in our community for a long time. But I'm still pretty bullish on the market.

Do enterprises have enough staff and the right staff to manage this explosion of RPA bots?

Le Clair: I think the answer to that is no, and that's why we have a surge in the RPA industry. Enterprises don't have that expertise currently. Do they have the staff internally that understands text analytics and NLP [natural language processing] for document extraction? Do they understand Python decision management? In terms of internally managing the bots, if they're taking out 100 FTEs [full-time equivalent positions], what we can say is take 8% to 10% of those to manage the bots. There's a replacement of line FTEs with support and maintenance and automation tracing, the people who do the governance, the automation checks and reviews. There's a cost to that as well.

Editor's note: This interview has been edited for clarity and conciseness.

Next Steps

Top 3 ways to overcome automation barriers

RPA market expected to plateau in the next few years

Dig Deeper on AI technologies

Business Analytics
Data Management