Point of View

The Raw Truth About RPA

October 12, 2017

The market for robotic process automation (RPA) and Intelligent Automation continues to be obfuscated by smoke and mirrors. If you listen to our friends at Gartner, satisfaction levels are allegedly at an unprecedented 96% while our own data rather suggest that roughly only half of the deployments led to satisfactory levels. So where is the market really at and what needs to be done to accelerate the journey? But more importantly what can be learned from the early deployments? Thus, our recent Summit in Chicago was a welcome opportunity to check what really is on buyers’ minds.


Buyers struggle to scale RPA projects


When we asked buyers in Chicago, how satisfied they are with their RPA projects, the results that can be seen in Exhibit 1 were astounding. The surprise was less around the low scores at the suggestion that their expectations were fully achieved, but more that many are struggling to scale projects and that they didn’t anticipate the impact on adjacent workflows and processes. Those suggestions are food for thought and are building on HfS’ much more detailed research on RPA satisfaction levels.


Exhibit 1: Polling question from HfS Summit in Chicago “Buyers, how satisfied are you with your RPA projects?”

Source: HfS Research 2017, n=36


One buyer succinctly articulated the implications of those concerns: “You don’t buy RPA, AI or Blockchain, you buy an outcome, yet providers’ organizational issues are pulling us back to technology.” It is here where the overselling of the supply side cuts in. Until compensation schemes and organizational issues change, we have to cope with an enormous amount of smoke and mirrors. Another buyer built on these issues in progressing on the automation journey, “AI is nothing you take off the shelf, it is a disparate set of capabilities, it is about orchestration, ecosystem, data.” We had heard similar concerns at our last Summit in New York: “We need to move beyond technology by being specific, in particular, specific about the use cases. And we have to move from bots to data.” This raises a plethora of questions from compliance to governance.


So, what holds the future for RPA? When we asked the audience in Chicago where they see RPA in 12 months’ time, we got clear answers. As exhibit 2 highlights, 34% reinforced the message that RPA will be all about transformation and not products. Slightly surprising 20% suggests that either Google, Microsoft, or AWS will enter and disrupt the market. While we have argued that around AI will see a shift toward mega ISVs, a direct involvement in RPA would certainly come as a surprise to us.


Exhibit 2: Polling question from HfS Summit in Chicago “Where do you see RPA in 12 months’ time?”

Source: HfS Research 2017, n=59


To get a more nuanced feedback on the issues surrounding RPA deployments, HfS did run two breakouts titled “The raw truth about RPA”. In those sessions we leverage a simplified Design Thinking method that facilitates constructive feedback on any given topic, using simple statements that convey feelings – I Like, I Wish, and What If? In both sessions, we saw a surprising convergence of thoughts, experiences and ideas by a wide range of RPA stakeholders – services buyers (both new to RPA and experienced practitioners), RPA vendors, and sourcing and automation advisors. We present the synthesized RPA experiences in the same design format below.


Constructive feedback on RPA to the services industry



  • That RPA works! It brings us efficiency and quality, increased throughput, and helps us in managing volatility.
  • That RPA has a low entry barrier. It has a relatively low entry cost, allows us to test quickly and fail fast, is relative ease of use, and has a good time to value.
  • That RPA solves stubborn business problems. Things we couldn’t address or bring up with IT before can now be dealt with by ourselves.
  • That RPA brings operations and IT together. Without change management, projects are likely to fail.
  • That RPA creates a new source of value to clients. It brings us (service providers) closer to clients. The branding alone is valuable – robotics sells.
  • That RPA documents undocumented processes! We can derive intelligence from automation, standardization, thus creating new levels of transparency.



  • We could stop calling RPA new. The basic concepts have been around for a decade.
  • We had more realistic expectations for all stakeholders – and definitions, offer insights as to what is reality and what is hype.
  • We had more RPA maturity overall. Maturity around change management and more education on the realities, use cases, examples of failures and successes.
  • We wouldn’t see Machine Learning and RPA as silver bullets or a cure. We need broader education around this. Machine Learning is not about having the best algorithm, it’s about the best integration into the fabric of the process.
  • We could get agiler. That is experiment faster, leverage bot libraries; furthermore, that we had reusable business knowledge, central business rules engines
  • We could rethink our RPA business case. Take a more strategic view, longer-term – albeit with softer criteria. We would move beyond narrow notions of cost. Fundamentally, it is not about FTE reduction, we need more clarity of the underlying costs including attrition
  • We would look at the investments as a strategic opportunity, in particular, that at least half of the cost savings would be used for transformational projects.



  • RPA was free? It is already commoditizing; RPA costs get reduced year-on-year.
  • We had common RPA standards? Furthermore, a knowledge hub for adjacent knowledge where FTEs get freed up.
  • There were people in the industry with functional expertise who were also experts in RPA? The talent that understands the impact of RPA on process chains and workflows is scarce.
  • We could share data across industries? The same applies for benchmarks and metrics.
  • We could measure less tangible benefits! Thus, the business case and communication internally would become easier.
  • Risk education was more advanced? We could overcome data breaches, security.
  • We don’t need bots in the future? Because of the speed of digital transformation, we could leapfrog legacy and have native automation.
  • We could look at this as a continuum? RPA/RDA could be part of something bigger (reengineering), thus RPA is a wake-up call!
  • We truly transformed to a customer-centric platform that is companies with legacy processes. OneOffice in all but name!


Bottom-line: RPA needs to support outcomes through orchestration of disparate sets of technology and data


The voices of the RPA community in Chicago were loud and clear: On a basic level RPA works and yields results. However, buyers are struggling to scale projects and often lack an understanding how they can advance to more data-centric models. On this journey, standards that can help with the communication, and case studies that convey the lessons learned would go a long way. While they acknowledge that RPA could evolve into a crucial lever for progressing toward the OneOffice, the buyers criticize that cost savings are not being reinvested for transformational projects. They are in agreement that in order to support outcomes, RPA needs to be integrated with other disparate sets of technologies as well as data. To succeed with those projects, the industry urgently needs a new breed of talent that blends functional experience with practical understanding of those innovative technologies.

Sign in to view or download this research.


Lost your password?


Insight. Inspiration. Impact.

Register now for immediate access of HFS' research, data and forward looking trends.

Get Started