Why a Virtual Workforce Could Be the Deal of a Lifetime

Robotic Process Automation (RPA) software is being used in a variety of finance and tax processes. Here's what you need to know about the cost, implementation process, and ROI.

Adoption of robotic process automation (RPA) is rising and RPA tools will play an important role in helping companies improve key business processes.

Though many of us understand the benefits of RPA, like speed and efficiency, few of us are aware of the implementation process, actual cost, and the potential ROI of a virtual workforce.


RPA can be implemented relatively quickly, compared to previous digital transformations, as it requires minimal capital or infrastructure.

According to research by Capgemini, a key advantage of RPA is that, unlike IT transformations such as enterprise resource planning, RPA does not require a massive upfront investment. Another advantage is that RPA tools can sit on top of a legacy system, so a significant change to the organization’s current IT tools and processes is not required.


Sharda Cherwoo, the Americas Tax – Intelligent Automation/Robotic Process Automation Leader at EY says, “When you think about RPA, you start thinking about it as digital workers or a virtual workforce. When you think about a virtual workforce person, it typically costs under $10,000 for an annual license. The bot works 24/7, 365 days a year, no distractions, no vacations, no lunchtime or anything.”

Speed comes into play when considering the cost of RPA. Cherwoo explains, “365 days a year, 24 hours a day, we can have this bot work 8,000 hours for $8,500. But if the bot works four or five times faster than a human being, they’re doing $40,000 worth of work. One bot, under $10,000. So that’s a deal of a lifetime for people.”

Also of note, bots cost about a third of the costs associated with offshore labor or service centers.

Return on investment

The return on investment (ROI) for RPA, is 10 times that of ERP implementations, according to Cherwoo. “At 10 times more ROI, it’s obvious in terms of where we’re heading with RPA. And this is first-generation RPA. We haven’t even scaled the fine-tuning and doing all the fancy stuff one can do.”

The faster you get onto the RPA journey, the better, because the transition requires a major change in work style. Cherwoo describes those who have begun this transition as having an “early adopter advantage.”

“If I’m sitting and doing stuff the old fashioned way, I’m inputting a bunch of information from one spreadsheet, putting it into a system and taking hours to do it, and now you say I’m going to have a bot do that for you, I can now do something else. You have to rescale, upscale, and get that person focused on what else they can be doing that’s greater value add.

“The point is, our best learning doesn’t happen when you are inputting a bunch of stuff into one place. Our best learning happens when there’s a problem you need to solve. So one number doesn’t reconcile to another number, and you’re trying to figure out if it’s supposed to be 100, why is it 120? What is the 20 dollar difference? What bots help you do very effectively is reconciliation, and pointing out the disconnect. Because they’re doing the routine, they’re not thinking. They’re just highlighting, looking at five million pieces of data, out of them these 50 didn’t match.

“If I can spend my time figuring out why those 50 don’t match and understand the why, I’m a better problem solver and I can get to the root cause faster. The bot and human, working together, is a hugely powerful combination.”