In payments, there are solutions that can address the symptoms of a friction-filled process, and there are cures that can replace that process outright with something better. Take the simple example of receipt data in corporate expense management. Technology today can extract data from a paper copy of a receipt for input and analysis in spend management solutions, or it can push corporate payers off the digital receipt entirely by offering tools that automate reconciliation of a transaction without the need for a receipt at all.
The issue for the company is, which strategy do you choose?
“It’s not one or the other,” says James Thomas, founder and CEO of Itemize Corp. “It’s both.”
It’s an interesting position, considering Itemize’s solution relies on corporates’ ongoing use of paper. The company automates data extraction from physical documents for organization’s expense management and accounts payable processes. Yet, according to Thomas, this technology is actually an important part of businesses’ journeys to letting go of paper altogether.
That’s because even when documents are delivered digitally, someone — or some thing — must read it in order for data from that document to be used and analyzed. If a company’s existing technology has no way to automatically capture data, it may not matter whether that information is in a physical or a digital document.
Thomas identified this area of corporate data management as a significant opportunity for artificial intelligence (AI).
“One way to quickly get rid of paper is to harness AI technology that will reach information from paper documents,” he said, adding that AI can provide the intelligence necessary to “read” documents, digital or not — and that is where the true value in efficiency lies, particularly in expense management and receipt data management, a process Thomas described as “inefficient, boring” and one in which “everybody makes mistakes.”
Worse than innocent errors, however, is fraud. A recent survey from Chrome River calculated that expense fraud is costing U.S., U.K. and Australian businesses $1.9 billion every year, while a separate report from the National Society of Accountants for Cooperatives noted one-quarter of survey respondents admitted to having been caught committing expense fraud.
The Association of Certified Fraud Examiners estimates expense fraud to account for 17 percent of all business fraud.
Traditionally, companies’ first line of defense against expense fraud is simply to check filed expense reports against receipts — a job that AI solutions can automate, Thomas said, plus validate and verify both merchant and transaction data, plus extract and analyze this information at a speed unmatched by human talent.
Despite its potential, however, AI adoption may not come easy for organizations. Teradata’s “State of AI for Enterprise” report, published last year, found 80 percent of businesses already have some form of artificial intelligence implemented today, though a lack of IT infrastructure and talent remain top challenges. In all, 91 percent of survey respondents agreed that barriers exist to AI adoption.
Further, businesses struggling with the friction of paper documents may be generally technology-adverse overall, less likely to be able to overcome the hurdles of AI adoption. But it’s not the smallest of the least tech-savvy of businesses that are only using paper. Nearly half of businesses surveyed by Foxit last March said they continue to use paper on a daily basis — while only 2 percent said they never use it at all.
The survey highlighted the current mix of digital and physical in the corporate back office today, as 70 percent of respondents said they use both cloud solutions as well as boxes and filing cabinets to manage and store company data.
Implementing a solution that can digitize data on physical paper and then use AI to analyze it may ease much of the friction of adopting the technology — rather than overhauling processes with digital documents outright (at least in the beginning).
For Itemize, another strategy to promote AI adoption is through collaboration. The company’s latest partnership with Infor sees Itemize integrate its data capture technology into Infor’s platform that analyzes companies’ hotel spend.
Beyond AI, Thomas noted technologies like machine learning, computer vision, neural bets and intelligence OCR (optical character recognition) can not only capture data, but ensure its accuracy and quality. Added AI and human expertise mean fewer exceptions and errors over time.
“Of course, there are always strange outliers and edge cases,” he said, adding that a team of specialists “teach the machine how not to fail next time.”
Companies may not be able to overhaul their entire systems and processes at once for the sake of digitization, but a hybrid approach to digitization — and the adoption of technologies that can work with existing manual tools at play — offer the opportunity for companies of all sizes and sophistications to embrace artificial intelligence, particularly in expense management.
“There are lots of ways that AI is going to affect corporate T&E in the short- and medium-term,” Thomas said, pointing to the technology’s potential to reduce compliance costs for accounting and finance teams as a particularly bright opportunity. “AI is able to [take over] mundane tasks that humans are forced to spend time on, and allows them to focus on more important activities.”