Humans vs Robots: Who wins with RPA?

Mary Beth Folger

When you think about the future of work, what do you imagine? Is it robots in the place of humans, working away at all hours of the day and night in their cubicles? Or perhaps a robot assistant for every human, taking on all their mindless tasks and light decision-making?

It’s easy to get carried away thinking about all the potential applications of robots and sophisticated artificial intelligence enhancements that promise to give robots super-human capabilities. But, let’s all take a moment for a reality check. People will always be a necessary element within RPA because RPA itself was originally designed to help people. For this reason, people should embrace and advocate for automation.



Let’s start by exploring all the ways people can be influenced or impacted by RPA, and most importantly, how to measure it.

First, let’s understand why organizations want RPA. Spoiler alert: it’s to impact human work. For all the good things people contribute to an organization, like critical thinking, personalized customer service, identifying and remedying inefficiencies; the ways in which people can inadvertently have a negative impact on a business are just as plentiful.

This is born from our insistence that people should carry out robotic work. That is, the kind of work that demands speed, consistency, and accuracy on every single transaction, from every single worker. It sets a precedent that is impossible for people to meet.

Using people to execute mindless work can have a significant impact on a business, on the individuals themselves, and on downstream customers too.

But how do you begin to calculate that impact? We can start by calculating the cost of manual work at a process level. On its face, the calculation is pretty straightforward. Start with the manual handling time of the process – how long does it take the average employee to execute the process start to finish?

Next, break down how much the person is paid to complete that specific process within a particular timeframe. Multiple that cost by the number of transactions completed every month or year. And that’s your baseline cost for manual processing. You will need this number later, so keep it handy.

However, manual processing cost isn’t the only variable to consider. It’s important to understand the intangible impact of manual, repetitive work. What is the cost of mistakes? Of rework? Does the business incur financial penalties when something is processed incorrectly or outside of a specific SLA or timeframe? Is the volume of work inconsistent, making it difficult to normalize operations? Are you experiencing high staff turnover or frequent overtime costs? What about an influx of customer complaints or customer churn?

Such impacts should be measured as success criteria for your RPA program.

TCO – Total Cost of Ownership

We've covered the human cost, now to track the robot cost and bring them together. TCO, or total cost of ownership, is a calculation that can be used to measure manual processing cost, which we mentioned earlier, against the total cost of the RPA program. To calculate RPA's total cost of ownership, you must track upfront and ongoing costs related to owning and managing the software, as well as the program itself.


RPA software licensing

RPA vendor training

Infrastructure setup (on-prem or Cloud VM costs, or a hybrid of both.)

Consultancy costs

Third-party integrations, if applicable

Proof of Concept

Establishing the Center of Excellence (including the hire or reallocation of analysts, developers, and IT support for infrastructure and deployment)


Annual RPA software license renewal

Annual maintenance renewal

Annual third-party integration license renewal

Infrastructure maintenance and/or expansion

Ongoing training

Annual overhead for support personnel

Let’s dig in to that last point: personnel cost. HfS Research estimates that licensing costs represent only 25-30% of TCO. The remaining 70-75% represents the ongoing cost of personnel. Every Center of Excellence is comprised of stakeholders, process owners and infrastructure support personnel, but the majority will be business analysts and developers. Earlier we talked about calculating the human cost and the same concept applies to RPA personnel. Tracking any employee at 100% allocation to the project is straight-forward. For employees with partial allocation, use time-tracking and calculate an hourly rate based on their salary.

Tracking TCO is important for a variety of reasons. One, it is a quick glance into the ROI being produced by the program. Tracked all the way to the process level, clarity on ROI helps a business focus on what should, or should not, be automated.

For stakeholders, it is critical to understand how RPA is providing value to the business in order to report that information up to the C suite and retain support for the program.

As with any business or automation program there will be winners and losers. People who complete mundane tasks repeatedly most likely miss the opportunity to enjoy work and make a meaningful contribution to the business and its customers. Processes that costs businesses money; remove the opportunity to employ capital more effectively to generate wealth. Automation programs that do not return a meaningful ROI waste time and cash better spent elsewhere. 

With any automation program, understand, track and measure all of the program’s associated costs to ensure it is delivering the value you expected, for the business, for the humans, and for the robots

Latest Posts