The Right (And Wrong) Way To Give Accounts Payable A High-Tech Overhaul

Most companies today have at least some kind of automation in place within the accounts payable (AP) department. Even when invoices flow in as paper documents, organizations are using electronic solutions or optical character recognition (OCR) tools to process those bills for payment.

Research from iPayables found that efficiency gains from electronic invoice processing are greater than those from OCR, however, a conclusion that draws attention to the spectrum of AP automation.

“There are a number of organizations that have integrated some automation in the last five to 10 years,” explained Brian Shannon, chief strategy officer at Dolphin Enterprise Solutions, in a recent interview with PYMNTS. “But that’s largely around the notion of capture — an invoice comes in, and they use a tool like OCR to capture information.”

Automation in accounts payable can go a lot further, though, he said. Unfortunately, because AP is a back-office function, it’s often not the first target for a technology overhaul. As organizations globalize, Shannon said the accounts payable department should get more than just “one kick in the can” when businesses upgrade their systems.

There are many opportunities to do so, especially when it comes to invoice processing. Whereas OCR may negate the need to manually re-key in data, more sophisticated solutions reduce the need for manual intervention even further. Shannon pointed to scenarios in which particularly complex invoices come in — cases in which a bill is not attached to a particular purchase order, for instance. Technology that can automatically code that invoice and send it to the appropriate department for approval shorten the chain of paper-pushing that can take AP professionals away from more strategic tasks.

In payments, Shannon explained, more evolved AP technology can automate the ability to address the nuanced needs of banks and suppliers across borders.

“If you become a global organization, you have potentially hundreds of banks in dozens of countries, and they each have their own payment formats,” he said. “There are statutory country requirements, so technology can really automate that process and make it easier for companies to adopt payment formats.”

That leaves organizations more room to decide exactly how to pay their vendors, via ACH, wire or other rails, by more efficiently connecting into payment networks like SWIFT.

Further, automation in AP can address one of the largest challenges corporates face today: fraud.

Analysis published last year by APEX Analytix said 65 percent of companies admit they don’t authenticate their vendors, while 79 percent said employee records aren’t cross-checked against supplier information.

“When we look at AP fraud specifically, the risks that these companies are facing is only increasing,” said APEX Analytix Senior Vice President Phil Beane in an earlier interview with PYMNTS.

Shannon said Dolphin’s own research into this issue led to a glaring need for enhanced AP security.

“The research indicates there are significant incidents of payments fraud [in] organizations today,” he said. “We hadn’t been doing enough for our customers to alert them to where this was. Organizations don’t want to talk about payment fraud unless it’s already public news, but they know inherently they’ve got an issue.”

Automation technologies can analyze data from credit reporting agencies and other sources, for example, to vet suppliers, validate and approve invoices and address areas in the B2B payments process prone to risk, from internal employee threats to external fraud.

Technology for Technology’s Sake

The key to achieving many of these goals is through machine learning, which Dolphin Corp. uses in its accounts payable tool, FS2, to automate invoice processing, payments procedures and fraud analysis. According to Shannon, machine learning, robotics process automation (RPA) and other such sophisticated solutions are helping to push the organization past their first step of digitization via electronic data exchanges and OCR and into the next generation of AP.

In fraud, for instance, algorithms can more accurately and quickly detect anomalies in the data; in invoice processing, bills can be automatically coded.

“You have the ability to have greater transparency, not only into the invoice or transaction itself, but also into the metrics of how a process works,” he explained. “You gain better insight into working capital, tools like fraud prevention, a much tighter integration into an organization’s cash flow needs.”

But as companies promise themselves and their customers a digital upgrade, there can be a fine line between investing in smart tools to enhance processes like AP and simply integrating advanced technologies for the sake of appearing flashy and modern.

“Politically, within some organizations, some executives will be particularly interested in [a technology] so the company pursues it,” he said. “I think we saw that with the cloud early on, where organizations looked into the cloud even though they didn’t have a good sense of what it meant for their business operations. You may see the same here [in accounts payable], but the smart organizations will look at it more holistically, to see how it all fits into a business and IT view.”

When implementing new technology, that tool must enhance a business process to achieve maximum ROI and effectiveness; in the case of accounts payable, OCR may no longer achieve that, whereas machine learning, AI and RPA, in many ways, can. Shannon noted that blockchain is another tool on the horizon in accounts payable, albeit further down the road.

And while there may be some organizations attempting to jump on the blockchain bandwagon, the executive described distributed ledger technology (DLT) technology as a “different monster” unlikely to see the same sudden implementations that the cloud saw in its early days.

“Blockchain is not specific to one company, and one company won’t be able to dictate how blockchain works for their industry,” he explained.

As the AP industry explores the potential of blockchain, other innovations like machine learning and RPA are in the here-and-now of the space. But there are two sides to this coin: While some organizations may be hasty to adopt tech, often they are reluctant to make drastic changes to processes that may be slow, but familiar. Shannon said organizations certainly have to be wise when it comes to technology adoption in accounts payable.

“Whether its robotics process automation or machine learning, you have to take all of this with a healthy dose of skepticism that it is not just a marketing hype,” he said. “Finance people are relatively grounded in terms of the reality of what’s possible. There is some education that needs to happen, but as we talk about different ways to apply tools and technology, then you can start to see the lights going on, and they start to see what’s possible — and there is an evolution toward acceptance.”