Robotic process automation (RPA)-typically used to automate structured, back office digital process tasks-turns out to be the opening gambit in many organizations’ digital transformation strategies. It also appears to be a precursor to artificial intelligence (AI). In a recent research project on priorities in process and performance management , APQC, a business research institute, found that RPA was a nucleus of 69 percent of digital strategies. In another survey on i nvestments in process automation , anticipated RPA projects were right behind analytics and data management, and almost twice as likely as near-term investments in AI or intelligent automation. (See Figure 1) Only 12 percent of those APQC surveyed had no plans to invest in any of these technologies in 2018.
APQC also found that the number of RPA projects per organization doubled from 2017 to 2018. The average number of projects per organization was 8.6 in 2017 rising to 14.9 in 2018. (See Figure 2).
In a separate project, one of the authors (Davenport) conducted a study with a team from Deloitte-described in this — that found 71 of 152 early cognitive technology projects were RPA.
According to APQC’s Holly Lyke-Ho-Gland, who led the project, “Organizations spent the last two years getting smart and testing RPA through proof of concepts or pilot programs. Now they’re scaling up.”
What is fueling this early and rapid adoption of RPA? There are three major factors: ease of implementation, the proof from successful pilots, and the partnerships successful pilots require.
Ease of Implementation
RPA is a relatively easy entry-level strategy into digital automation of back-office processes. One consultant described RPA tools for structured digital processes as a “gateway drug” for other cognitive technologies. RPA is easy to configure and implement, and small implementations may not even require an expert consultant or much help from a vendor. RPA is particularly well suited to working across multiple back-end systems and doesn’t require re-architecting of those systems. It typically brings a quick and high return on investment.
The second catalyst for rapid adoption of RPA is the success of early pilots and proofs of concept. APQC’s latest report, Make Success Automatic: Best Practices in Robotic Process Automation found that over 75% of respondents said their early RPA projects had met or exceeded expectations (See Figure 2–Note: only 41% of respondents were far enough along to evaluate their satisfaction with RPA projects; Figure 2 includes only their data.)
For example, CUNA Mutual’ s pilot program focused on automating transactional activities for its claims adjusters. Not only did the pilot meet the strategic goal to increase capacity without increasing headcount, it also gave claims adjusters time to be more strategic in their assessments of claim payments and denials and allowed the finance team the opportunity to be more strategic in executing their process. This level of satisfaction is a rarity for many IT applications. Meeting expectations may be easier for automation and robotics given they often have a clear process to automate and a measurable business case.
At NASA, cost pressures led the agency to launch four RPA pilots in accounts payable and receivable, IT spending, and human resources-all managed by a shared services center. Shared services centers are often responsible for implementing RPA in many companies. At the space agency, all four projects worked well and are being rolled out across the organization. In the human resource application, for example, 86% of transactions were completed without human intervention. NASA is now implementing more RPA bots, some with higher levels of intelligence.
APQC has found across several research studies that one of the major predictors of funded and successful RPA or machine learning projects was the formation of internal alliances across functions to create and learn from the pilots.
Two factors had a statistically significant relationship with satisfaction. The first was having good selection criteria and the second was the inclusion of key functions in the RPA project planning and execution. Including representatives from information management, the target functions and especially HR (See Figure 3) is positively correlated with project satisfaction. According to Lyke-Ho-Gland, “HR is often included in organizations’ RPA steering committees, not only to allay fears and create buy-in but to create action plans and training for displaced FTEs. Ultimately this helps organizations use RPA as an opportunity to build capacity for sustainable growth rather than simply reducing costs.”
IT and process management participation is important too. “While not statistically significant, organizations need to ensure both IT and process management are equally involved in RPA efforts,” says Lyke-Ho-Gland. “IT ensures that bots are integrated smoothly with existing systems and process management helps reduce costly, post-production rework by re-engineering processes for digital execution and ensuring all process variants and exceptions are captured and understood.”
A Virtuous Cycle: RPA creates digital dust for future intelligent applications
One could also argue that RPA lays the groundwork for machine learning and more intelligent applications. It both gathers useful data and is being combined with AI capabilities. One of us (O’Dell) recently interviewed Eric Siegel, a predictive analytics expert and author of the book, Predictive Analytics: The Power to Predict Who Will Click, Buy, Lie, or Die . Siegel pointed out an often overlooked benefit of starting by digitizing processes with simple RPA: the digital bread crumbs it now leaves behind. “This data wasn’t amassed in order to do machine learning. It’s just a side effect of doing business as usual. The transactional residue accumulates and, lo and behold, it turns out this stuff is really valuable because you can learn from it. You can derive these patterns to help improve the very transactional processes that have been accumulating the data in the first place.”
Vendors and user firms are also combining RPA with AI tools like machine learning, natural language processing (NLP) and image recognition. Organizations that take a phased approach to their RPA efforts set themselves up for success as RPA continues to get smarter. One financial services organization accomplished this by categorizing its RPA projects into three categories:
- Simple-bots that gather input and output this data to other systems,
- Hybrid-bots that gather input and make simple decisions based on business rules, and
- Cognitive-bots that make decisions or interact with humans involving machine learning or NLP.
The RPA team started its work on simple and hybrid bots before experimenting with cognitive bots. This gave them early successes and a foundation in the mechanics of RPA, as well as experience with the semantic content of NLP.
These success factors make RPA a reasonable, low cost and lower risk entry-level approach to AI even if the technology is not very smart today. RPA nicely lays the foundation for more intelligent applications later. And even without the potential of more intelligent RPA, the ease of implementation and rapid ROI from many RPA projects makes them worth strong consideration for almost any firm today.
Carla O’Dell is the chairman of , a non-profit business research institute focused on benchmarking, best practices, process improvement and knowledge management for a global corporations and consulting firms. She has authored three books, one on competitiveness and two on knowledge management. She writes and speaks frequently on the impact of AI and cognitive technologies on how we share knowledge and writes an APQC blog and interviews series called Big Thinkers, Big Ideas .
Article by channel:
Everything you need to know about Digital Transformation
The best articles, news and events direct to your inbox
Read more articles tagged: Robotics