Robotics Process Automation: Take Out Costs or Empower the Workforce?

April 29, 2016 | Post by Sam Gross | 3 Comments

Robotics Process Automation

Companies regularly explore new IT technology investments, and upon discovering the promise of new advances, quickly make their way to their traditional comfort zone: the knee-jerk reaction of leveraging technology only to reduce operating expenditures (usually starting with the CIO’s budget). This is increasingly true in the case of one of the newest technologies to enter our “technosphere,” known as Robotics Process Automation (RPA). Simple downward pressure on IT labor costs often generates a “CFO-friendly” return on investment (ROI) that justifies the business case for new technologies such as this. This is a “business of IT” pattern that repeats itself all too often as CIOs seek the tools necessary to modernize their company’s technology footprint, and their associated competitive enablement.

This thinking misses the mark.
Technology is one of the most powerful conduits available to enhance the experience of our true “customers” – the employee population, those employed to sell, deliver, manufacture and administer – who depend on technology to do their jobs more effectively. By extension, our “customer’s customer” and their experience can be directly correlated to technology enablement decisions! So, do we really win by only taking out cost, or do we win by empowering the workforce to be more effective and drive the top line? In the case of RPA, automating IT-related processes and systems will reduce labor costs. However, we quickly realize that the most significant impact of this very promising technology is on improving employees’ ability to do their jobs better. This, in turn, enables the competitive edge that differentiated customer experiences require.

We live in an economy where almost everything we do in the workplace is technology enabled, spanning all sectors from service through manufacturing. Yet still, IT systems and software remain complex. Even though we’ve been engineering software for 40-plus years, it is still a flawed process (imagine if we built bridges the way we engineer software!). Use of RPA as a technology to remediate the inevitable technology faults helps us to instantaneously execute fixes at machine speed. This minimizes the mean time to restore, which ultimately enhances the end-user experience of IT’s true customers.

Robotic Process Automation technologies can be transformational.
RPA use cases and applications are infinite – it’s mind-boggling. RPA is enabled by several artificial intelligence-based component technologies, and extends well beyond fault remediation. To get started with automating IT-related processes and systems, understanding the differences between scripting, orchestration and robotic process automation is good way to begin the ultimate journey:

  • Techniques such as scripting (and macro-routines) are often as diverse in their structure and syntax as the engineers that created them; this makes scripting a brittle solution that defies standardization by its very nature and remains tightly coupled to a specific component of technology that it services.
  • Orchestration introduces both structure and syntax and often involves participation and decision making on the part of the technology that is being orchestrated. Hence, the infrastructure itself becomes an active participant and collaborator – for instance, by publishing its configuration information that is then consumed by the orchestrated process. This characteristic makes orchestration “smarter” and so it can be loosely coupled to the technology that it supports.
  • RPA has a unique ability to be aware and adapt to changing circumstances, exceptions and new situations. It can manipulate data, trigger responses, initiate new actions and communicate with other systems autonomously. RPA is built on machine learning (a base component of AI) which makes it the “smartest” and enables a wide variety of use cases, building on its own knowledge and determining what to do, where to do it and when to do it.

Historically, the worry around large-scale scripting and even orchestration has traditionally been that if your automation doesn’t have the knowledge of what-where-and-when, to do it or not do it, you’re in no-man’s land and at risk of actually breaking more than you fix. However, the integration of orchestration and machine learning, which both fundamentally underpin RPA, is a way to ensure that you know when to initiate an orchestrated task, where and which one.

Will RPA and “AI-like” software just fizzle out again as AI has in the past?
The open source community, academia such as MIT, and industry giants such as Google and Facebook are all actively contributing the core technology components driving RPA advancements, as well as the cognitive and AI computing space. As the open source community continues to become the prevalent contributor of technology advancement, we are beginning to see a number of start-ups entering the marketplace. These start-ups leverage machine learning, RPA and cognitive computing, and apply these technology advancements in commercial software and services. We’ll see more start-ups emerge out of stealth mode in the very near future.

This will benefit organizations that are adept in capitalizing on start-up technologies. The organizations who successfully leverage RPA and cognitive computing create better experiences for their employees and clients and also drive increased revenues, while the rest find it increasingly harder to compete in the evolving marketplace. That’s a glimpse of what the current path looks like today, the “haves” and the “have nots.” What will alter this scenario is the broad impact of open source, the API economy and SaaS, where a whole new community of start-ups and venture capital firms begin to invest in this space aggressively enough until the technology becomes fully ubiquitous, commonplace and generally available to all.

Our Challenge.
To empower the workforce by improving the customer experience, do we need to first make the case for cost savings, to entice our leadership to buy in and take the leap of faith? Can we change our current mental model to one that emphasizes building a more effective workforce to drive revenue up, rather than simply drive cost down? Can we transform our corporate belief system to where if revenue goes up, cost cutting becomes less of a problem and would in turn improve the effectiveness and morale of the workforce, further contributing to increased revenue and margin?

In the IT industry itself, across the major players, there is statistical evidence that 17-22 percent of the activities (incidents/requests) are being addressed using advanced automation. That’s estimated to go at least as high as 40 percent in the next three years, according to IT industry analysts. A doubling of automated fulfillment events in the next three years…is this a new expression of Moore’s Law, whereby processing speed doubles every two years?

Don’t Forget.
Regardless of your choice of technology strategy, RPA, cognitive computing and AI are all still dependent on the data that you have, generate and consume. Not a day should go by that organizations, departments and functions do not pay attention to the quality of their data and their ability to collect, aggregate and interpret it.

Nor should a day should go by when CIOs are not promoting the idea that employee job satisfaction is at least as good an argument for RPA adoption as the potential for reducing labor costs. Satisfied, productive employees stay loyal and drive more top line revenue. This saves on the cost of employee and customer churn and disruption to the organization. As to the initial costs of new technologies, including RPA, the old saying remains true: You can’t cut your way to prosperity.

What are your thoughts on RPA? Post a comment or question here!


The content and opinions posted on this blog and any corresponding comments are the personal opinions of the original authors, not those of CompuCom.

  • Sam Gross's picture

    Sam Gross

    Sam Gross is CompuCom’s Chief Technology Officer and is responsible for CompuCom’s strategic technology planning, technology portfolio management and product and services development.


Hi Sam,
Great article that sums up how people tend to engage on RPA.... Cost first and rest later... This is can be a very destructive approach and tends to reduce the spread of RPA technology as employee see robotics as a competitive form of labour. When organisations see RPA as a means to improve customer experience, employees in different departments can make use of the technology to improve their work and improve customer service and employee morale. This overtime can change the labour fabric within an organisation where employee-robot collaboration is used to improve efficiencies and ability to scale business services.

Thank you for echoing this important point of view. We have so become accustomed to years of cost cutting in IT, we are at risk of missing one of the biggest opportunities in IT history for the CIO to drive the strategic agenda!!!

A great practical view on RPA and its often overblown advance in business process automation.
Unfortunately, the reality is that RPA technology has been around for more than 25 years. It has more recently been reproduced by the marketing guys with a new set of clothes supposedly easier to use. The truth is very few of the pure RPA technologies, can consistently claim success and they rarely rely on anything new - keyboard commands, VB scripts, Javascript, etc. These are not new techniques, so you have to ask yourself why is this making a difference now. Well behind the marketing facade and over-hyped anecdotal results, the truth is it still achieve the same 8 - 12% cost shift that scripting and free code does for the IT guys; and did 20 years ago. 17%, and 40% are still the preserve of aspirations for RPA. There is very little smarts, AI, or orchestration in these technologies.. They automate simple sequences of repeatable tasks. From the singular promise to achieve entirely "robotic processes" this new marketing generation is facing complete failure.
However RPA is good as an assistive technology, a "sticky plaster" as one of our clients call it, in an ecosystem of automation where RPA enables access to systems via screen and keyboard for which API's are not available or simply are taking some time to get. The really smart technology is not in how you control the screen and keyboard, it is how you eliminate the screen, keyboard, and the person in the chair. This is known as "Intelligent Automation", a super-set of RPA, cognitive computing, AI, and orchestration. If you extend the analogy of the Robot, RPA is only the eyes and the hands...the interface to some of the outside world, whereas Intelligent Automation adds the brain and the coordination in the middle to provide cognitive ability to understand what the eyes are seeing, the knowledge and expertise to decide what to do, and the hands, legs, teams, armies, other machines to coordinate outcomes.
Intelligent Automation is already achieving >90% autonomous operations in places like service-desk, network management, and other digitally driven services. It is making strong inroads with reports of 70% - 80% levels of automation in SCM, Accounts Payable, Travel and Expenses, Claims Management, etc. This technology was only conceived around 6 years ago and has matured quickly.
Most importantly it is smart enough to know when to ask for human assistance... the human face of automation.. to truly accelerate the capabilities and value that an individual can bring to an organisation by allowing them to retain control.
Innovation in this space are new and exciting, every day a new technique or analysis; new patterns, new expert engines, or maybe IBM Watson capabilities are integrated into Intelligent Automation, and new frontiers are being broken down with no hands, no keyboards, no chairs, desks, and no bored or frustrated faces staring blankly getting a "screen tan".
The Robotic Process Automation tortoise may have started the race earlier but it is now dead... long live Intelligent Automation hare.

Add new comment

By submitting this form, you accept the Mollom privacy policy.


[x] Close

Sign Up for Email