insight magazine

The Bots Are Coming

Getting robotic process automation right should be CPA firms’ future focus. By LISA WILDER | Spring 2020


Does the possibility of robots in the workforce send chills down your spine? Whether you fear a robot uprising, or you’re just scared a robot might take your job, robotic process automation (RPA) can be a daunting prospect. But the reality is that embracing RPA is unlikely to result in a scene from “I, Robot.”

In your office, the tasks most likely to be automated are boring chores, like payments processing or account reconciliations— things you’ll likely be thrilled to hand off to an artificial mind. The bottom line on bots is that RPA offers the tantalizing possibility of saving money and time while allowing workers to focus on more challenging and strategic tasks that require human skill and nuanced decision-making.

At the most basic level, RPA is what it sounds like: a way to automate routine and repetitive work using software bots. So, it should come as no surprise this technology is gaining prominence and importance in the business world. The global market for automation technologies like RPA is growing by 20 percent a year and is estimated to reach $5 billion by 2024, according to Deloitte’s 2019 report, “Automation with Intelligence.” What’s more, Gartner’s 2019 paper, “Predicts 2020,” estimates organizations will be able to lower operational costs by 30 percent by 2023 by combining automation techniques with redesigned operational and human processes.

With its potential for increased efficiencies, what could RPA mean for you? Chris Denver, CPA, MBA, national assurance practice director at International Financial Group, says it’s worthwhile to begin with discussing “the elephant in the room”—what bots could mean for hiring and recruitment of future accounting talent.

“Can a bot replace a public accounting firm partner who has to exercise a lot of judgment in determining whether he or she believes a client’s financials are free of material misstatement? No. And the technology will never get to that point,” says Denver, whose firm consults on and designs tailored RPA solutions for a range of financial clients. “But could it mean the next hiring class doesn’t have to be as large? Possibly.”

His point: Future accounting professionals should consider building RPA skills, and current CPA firm leaders shouldn’t wait for new graduates to tell them about RPA. Here’s what decision-makers need to know now:


Maybe the better question is where’s a bot? You’ve probably already encountered bot software as a consumer or possibly when studying the customer-facing functions of various clients. The travel industry, for instance, has become a major user of so-called chatbots, which essentially execute predetermined conversations with customers to make reservations or solve customer service problems.

For accounting firms, a bot application can be as simple as directing a potential client to the right person to contact at the firm, or to perform data capture and reconciliation—replacing human taskwork that might take staff hours or days to do. What a bot isn’t is a solution for ambiguous tasks or functions requiring complicated judgments and decisions requiring weigh-in and review.


According to Jeff Aldridge, intelligent automation leader at EY, automation could help many organizations reach their long-term operational goals. Leaders should approach RPA with a specific outcome they hope to achieve. “What is your purpose in implementing automation? Is it efficiency and cutting costs or reducing head count?” Aldridge asks. “Or could it free up an organization’s best people to do more value-added tasks?”

Generally speaking, bots can automate repetitive, static processes and functions that do not require human judgment. With this in mind, large firms might consider an RPA approach among not just one but all departments burdened by repetitious tasks—even the smallest firms can find value in automating at least some repeat tasks. Why’s that?

Denver notes that as accounting firms and other professional service providers move toward more fixed-fee arrangements, strategic bot installations can cut costs in ways that allow them to focus on potentially more lucrative, high-value services.


Aldridge warns that even the simplest bots require customization based on the technology ecosystem they’re placed in. “Each organization has unique processes and supporting systems that such bots need to access,” he explains, which means your first step in the process should be designating “a strong C-suite champion” who leads a discussion about the purpose and goals of automation overall before leaping into buying or building bots. “It’s important to begin with a broader business case to justify the initial infrastructure investment and to see how such technology may be leveraged across the organization,” he adds.


According to Aldridge, an average bot can accomplish up to three discrete automations intended to be run separately, which is why large organizations may run hundreds of bots. “RPA works 24/7,” Aldridge explains, “but the one thing they can’t do is two things at once.”

Because of that, task scheduling is as critical as bot maintenance. Denver adds that bot software can be very “fragile,” as the smallest change in the application’s ecosystem can slow or completely derail its functionality. One example: If a bot pulls in information from a website that has recently been redesigned, even a seemingly small tweak like a form field name change could stop a bot cold. Denver stresses that bot function needs to be constantly monitored not just for accuracy but for any change in their interaction with other software. Simply put, when tasks change, bots need to change. “There needs to be a shelf life built into bots,” Denver cautions.


Since bots need to be designed for the software environment they’re living in, they require customization. EY, which in recent years has applied RPA within its own infrastructure, offers a rough average estimate of what bot startup costs: A bot that performs up-to-three discrete functions requires a hardware investment (mostly servers) of around $20,000, a software investment of approximately $15,000, and development costs nearing $50,000.

That said, Denver estimates some bots can be designed for considerably less—or more—based on purpose and needs. While current technology and budgets often require a focus on simpler tasks, that will also change in the future. The basic equation to justify the cost of investing in bots begins with accounting for staff time and compensation saved.


Reading about artificial intelligence and RPA technologies and attending seminars helps, but it also pays to watch RPA development among peers and clients. Sometimes the best learning is hands-on. Aldridge points to Gartner statistics indicating 60 percent of companies with more than $1 billion in revenues have already started their RPA journey, adding that by 2022, it will be close to 85 percent.

“One of my largest clients right now is a $20 billion manufacturer who is trying to be the most automated finance organization in the world by taking 50 percent of its current finance (human labor) hours out through automation,” Aldridge notes. Meanwhile, EY has installed some 2,000 bots throughout its global operation, 700 of which are aimed at making internal operational tasks more efficient, including in finance and HR, making EY one of the largest users of RPA in the world, according to Aldridge.

In other words, some of your own clients may be able to offer deep insight into bot technology.


Gartner describes the current state of RPA as “toolboxes where customers are expected to build their own automation.” But rather than focusing on the construction, Gartner suggests that organizations spend more time thinking about goals because technology will change faster than long-term targets.

In fact, Denver says that we’re not far away from a time when bots will become so sophisticated that they’ll be able to learn how to improve their performance and functionality. That means the future leaders of the accounting industry don’t necessarily need to become computer engineers, but they’ll need to better understand technological evolution and how it can improve their performance.

“One of the things that public accounting firms really ought to be looking for now are graduates with RPA skills at some level,” Denver suggests. “A good cultural fit is most important in hiring, but it’s also going to be important to recruit professionals who have a greater comfort with technology—and possibly some hands-on skills.” Further, Aldridge recommends recruiters focus on new graduates who arrive with business strategy and decision support skills that bots don’t yet possess and let bots do the rest.

After all, who doesn’t want a new, quiet coworker who does all the boring drudge work? So, hand over your mind-numbing number crunching and paperwork and welcome the bots to work.

Leave a comment