Deloitte leverages RPA for audits

Deloitte is making use of robotic process automation to help its auditors and clients with their financial statements.

“We’re at an inflection point — I would say at a transformational point — in thinking about how finance in the future will look and what that will entail and the nature of work which individuals are executing, along with a change in the talent model of potentially what individuals will be doing,” said Scott Szalony, who leads the Digital Controllership service offering at Deloitte. “We see a future where the mundane routine of repetitive tasks of data entry and journal entries and things that are repetitive in nature, and highly standardized, low-level complexity become automated. Things that require cognitive ability and the ability to make judgments will still be done by humans and by individuals, but to make judgments we’re seeing a lot of technologies out there that quite frankly are helping with understanding large amounts of data to help make better decisions and augment those judgments.”

In addition to RPA, cognitive technologies such as natural language processing, natural language generation, machine learning and predictive analytics are coming into play within firms like Deloitte and other financial organizations. “It’s really an important time, if you’re in an organization, to think about the ability of your finance practice and what that’s going to look like,” said Szalony. “We professional services firms, audit firms, tax firms and public accounting firms are also using those types of technologies to think about how to execute their work as well. It’s across a broad spectrum, with lots of exploration and lots of use cases that are coming to fruition. We think about technologies like RPA and analytics. Those are adopted on a pretty widespread basis. We’re still building out use cases. We’re by no means to the finish line. I’m not even sure what that finish line looks like necessarily, but lots of people, lots of companies and lots of members of management as well as us are using those technologies today.”

Deloitte building in Ottawa
Deloitte's Canadian office in Ottawa

RPA and AI technology are helping firms perform audits more cost effectively. “Every year we’re able to provide a more efficient audit, and we’re always raising the bar on quality as well,” said Deloitte audit partner Bradley Niedzielski. “If you have administrative tasks that are repetitive in nature, sometimes you might have errors that occur. But if you implement robotics, it helps you to be more efficient and reduces the amount of errors that may occur. Clients are dealing with larger sums of data. But when it comes to processing time, the timelines aren’t increasing. They’re not being moved out. There’s still that same reporting period that needs to be met, and in some cases it actually gets moved up. So larger amounts of work need to be processed in a shorter time frame every year.”

There are trade-offs, however, in becoming overly reliant on technology. Several of the major auditing firms have run afoul of regulators in recent years over lax audits, leading to calls in Europe and the U.K. for more stringent regulation of the audit profession.

“Whether it be in an audit, or executing in a company, these technologies have both risks and rewards,” said Szalony. “Obviously the technology can be used to think about the effectiveness of our procedures and improve the quality of our work, but as you automate things you have a higher susceptibility of a higher volume of transactions going through that are being processed at a time exponentially greater than what we humans can do. Because of that and the large sums of data, there is an inherent risk just because of the pure volume and nature. You need to make sure you have the right controls in place to address the potential risks of error in whatever procedure you’re executing, whether that be in audit, or if you’re in a company, the processes that you’re using with that transaction. Think about the controls and the nature of the controls in the way you design them.”

It’s still critical for auditors to exercise their own judgment and professional skepticism, and not expect the technology to do all the work for them. “The focus here is more on the administrative tasks, but anywhere that there is still judgment involved within the audit or within the accounting function, this allows us to deploy the talent now to those higher-value tasks that we want to perform where there is judgment involved,” said Niedzielski. “It allows us to focus more on the areas with more risk and more complexity.”

The finance department within a client company can also leverage RPA technology in addition to the auditing firm. “The real benefit is within the finance department, within the accounting function of the clients that we have,” said Szalony. “When we go do an audit, now we can look at larger populations. For the most part, when you do an audit, you do a sample of a population, but for a finance department, they have to be comfortable with 100 percent of their population. That’s really where they’re getting the most benefit from implementing RPA and using analytics.”

The RPA technology can help the finance department with the process of preparing the financial statements. “When a company prepares their basic financial statements, a simple task might be when you do comparative financials, you’re going to take your prior-year numbers and move them into the column, because you went from 2018 to 2019,” said Niedzielski. “That’s something that someone in their department used to do manually, but now from an RPA standpoint, the process automation will automatically transmit that from the current-year column to the prior-year column. When the person was manually typing it, they may have what we call ‘fat fingered’ an entry, or switched a nine versus a zero. But when a robot is doing it, you have more comfort there. There’s still someone reviewing those financial statements to look over those numbers, but it’s more automated.”

RPA can help reduce the chances of data entry errors occurring. “If you have someone who’s in the finance department who’s processing 100 invoices a day, by the time they get to the end of the day on that 90th invoice, they’re mentally tired and if they see an invoice that’s for $46,969, they might by accident book $46,996,” said Niedzielski. “But when you use robotic process automation, which has tools to actually read that invoice, the bot isn’t going to get tired at the end of the day. It’s going to work just as effectively as it did at the beginning of the day. The human, when they’re working, might be working in the standard business day, but the bot can actually work during the off hours. The bot can work overnight. It can work after work, before work, and it also tends not to eat up server time when everyone else is in the office.”

For reprint and licensing requests for this article, click here.