A Thing
The Green SheetGreen Sheet

The Green Sheet Online Edition

October 11, 2021 • Issue 21:10:01

AI-powered statement analysis

By Dale S. Laszig

Speed is fomenting more competition than price in today's tech-driven merchant services sector, where instant approvals, quotations and real-time payments are replacing traditional methods of underwriting and risk management. In this climate, merchant statement analysis, the industry standard for measuring acquirer pricing, performance and efficiencies, has evolved from a lengthy review process to instant quotes and on-the-spot proposals.

Advanced feedback loops, composed of human analysts and advanced automated technologies, enable merchant level salespeople (MLSs) to quote prospects during an initial sales call, bringing back the coveted one-call close. The Green Sheet interviewed payments industry experts for a firsthand look at machine and human collaborations that drive merchant statement analytics.

Infinite variations

Adrian Talapan, co-founder and chief executive officer at Fee Navigator, an AI-powered service designed to instantly analyze merchant statements, has seen numerous variations in statement fees, formats and nomenclatures. Within this complex landscape, companies sometimes need help analyzing their own interdepartmental documents, he stated.

"One company's retention department asked for our help to decipher what their own sales group was doing," he said. "This is the level of complexity the [payments] industry is dealing with."

James Shepherd, founder, president and CEO at ISO Amp, a technology provider that leverages AI and human experts to analyze statements and create proposals, described merchant statement analysis as a complex, multifaceted practice.

"What I realized early on is that statement analysis is actually two different things," Shepherd said. "First, it's getting statement data into a manageable format and second, it's what our company calls matching, a process that is almost entirely automated."

Data entry, whether manual or automated, is an important step because that is where issues come into play, Shepherd stated. Matching involves asking the simple question, what is this? over and over, which, he explained, is how AIs learn to recognize patterns and rules.

Data integrity

Monica Eaton-Cardone, founder and chief operating officer at Chargebacks911, a global chargeback management and remediation firm, noted that interpreting raw data can be a daunting task for humans and machines alike, especially when it comes to chargebacks.

"The crux of the problem with chargebacks is preserving data integrity," she said. "If you have bad data, it doesn't matter what you're looking at; you'll get a bad result. And this becomes challenging when not everybody uses the same words, when there are different languages being used and different interpretations of rules, which creates layers and layers of complexities."

The first goal in preserving data integrity, from Eaton-Cardone's perspective, is training AI models to interpret data correctly. Throughout their ongoing training, AIs that are unable to interpret line items are redirected to experts who help them master specific items and rules.

"You start with something simple, with the goal of getting as much data as possible on a given case, whether it's images or digital files or merchant information, and then interpreting all the data to make it as standardized as possible," Eaton-Cardone said. "And then you can leverage automation, which is not a perfect system but can accurately interpret 90 percent of cases on the acquiring side and quite a bit on the issuing side as well."

Text, image recognition

Shepherd has also seen highly accurate character and image recognition by his company's AIs but admitted not every picture is perfect every time, despite advances in technology. "You don't want to be in a position where you have to reject a statement because your image recognition can't recognize it," he said. "So, in our company, we try to automate as much as we can and we have a team of people doing data entry work when necessary."

On a typical high-volume day, Shepherd estimated, 94 percent of all statement line items his company receives represent some kind of a fee that text recognition correctly brought into the system. The remaining 6 percent needing additional analysis is possibly due to a subpar scan that requires a data entry person to manually enter almost the entire statement, he stated. He also acknowledged that a significant percentage of incoming statements need corrections, and if they are not corrected, would be virtually impossible to analyze.

Talapan concurred that human and machine collaborations boost accuracy, stating a fast 30-second analysis that is 80 to 90 percent accurate is a good starting point, especially if it gives humans the flexibility to go in and change a few things. "While the ideal solution would produce sub-second automated results that are 100 percent accurate, in reality, statements are very complicated and most human analysis is not 100 percent accurate either," he said.

According to Talapan, Fee Navigator partners are mostly happy with automated analysis, but some of them like to go in and change a few things. For example, he said, they may want to manually enter a few data points that were not retrieved from paper statements, but even with these add-ons, they still save hours and are able to get back to sign or service the merchant very quickly. Hence, a fast 30-second analysis with 80 to 90 percent accuracy may be a good starting point if it allows a human the flexibility to spend a few minutes to increase that accuracy to 100 percent, he added.

Cash discounting

Talapan mentioned that cash discount programs, designed to help merchants recoup processing fees, should make merchant statements simpler and easier to understand but that is not always the case. "The industry introduced more cash discounting and reduced the use of tier and bill-back, which should result in simpler statements," he said. "Instead, we have seen markups hidden inside interchange, dues and assessments, as well as additional fees being added."

Noting that hidden fees make statements appear simple when they are not, Talapan said lack of transparency and paper-based statements that many merchants still receive by mail can significantly impede statement analysis. However, he noted, Fee Navigator scans paper documents and applies best-of-breed optical character recognition (OCR) technologies to extract data and convert it to digital formats and gets excellent results despite these setbacks. 

Shepherd also observed that merchant statements don't always indicate cash discount savings or credit card surcharge fees. In some cases, statements from cash discount acquirers do not even show a discount rate, just monthly fees and overall transaction volume, he stated.

As a general practice, Shepherd recommended skipping statement analysis when promoting cash discount programs and relying on it heavily when selling credit card surcharging. "If merchants process $7,000 to $10,000 a month and all I want to do is pitch cash discount, I probably won't even ask for a merchant statement, because I'm going to eliminate all their processing fees anyway," he said. "Surcharging is the complete opposite; some people sell it without statement analysis, which is insane because you have to differentiate between debit and credit cards."

Instant quotes

Sometimes it makes sense to skip statement analysis and go directly to an instant quote, Shepherd advised. ISO Amp provides a tool that estimates interchange costs, card brand costs and savings based on a merchant's volume and average ticket. Clients can use the tool without sending a statement, he stated. They would just ask the merchant how much volume they did last month, what their average ticket is and what they pay in total fees, he added.

"When it comes to surcharging, we have a tool that does a fully compliant surcharge proposal, even if the statement does not give us the details of credit versus debit," Shepherd said. "We use the average ticket size and other variables to make a guess of how much is debit versus credit."

To further illustrate this point, Shepherd recounted when a new client notified ISO Amp support that its surcharging estimator wasn't showing any savings. It turned out the merchant was a QSR running 70 percent debit card transactions, and the client was marking up debit, Shepherd said. Surcharging is not for everyone, he added, especially merchants processing 70 percent debit when you're charging 1 percent and 25 cents.

Talapan noted that even the best analysts can take up to 40 minutes to thoroughly review a merchant statement and create a competitive proposal. Full automation facilitates an agile review process in which a client or partner can literally upload or email a statement and get an analysis and proposal in seconds, right before your eyes, he stated.

"With Fee Navigator's Integrated Studio solution, our partners can inspect the analysis performed, as well as review and tweak the pricing and proposal to their liking, all in seconds," Talapan said. "The technology allows you to respond to opportunities quickly. It also enables digital customer onboarding—think of a potential customer visiting your site or marketing funnel, getting an instant proposal based on your parameters, and we'll pass all the information to your onboarding tech. All this is possible through advanced automation."

AI, human co-creation

Reflecting on the steady march from paper to digital services in recent years, Eaton-Cardone suggested that payment providers have much in common with other industries. "I look at chargebacks like payroll," she said. "Ten years ago, every company, regardless if you had one employee or if you had 200 employees, had a payroll clerk, maybe a whole team of payroll clerks."

Their job, she explained, was to manage the constantly changing government regulations to ensure that everyone got their paychecks, all line items were correct, and all transactions were reconciled, debited and credited perfectly. If it was done wrong, she added, "you had tax liabilities, you had payroll liabilities, you had government liabilities—it could become a nightmare."

As humans migrated to AI-powered solutions and automation, payroll companies began to replace yesterday's payroll departments, Eaton-Cardone pointed out. Payroll-focused companies that specialized in myriad aspects of payroll drove greater demand and adoption. Fast-forward to today, there's no such thing as having a payroll clerk with all this expertise and no business anywhere on the planet has the required level of competency to stay apace with the ever-changing payroll sphere, which is similar to chargebacks in so many ways, she added.

Pointing out that chargebacks can become a liability if they're not handled correctly, Eaton-Cardone explained why it's difficult to reconcile the debits and credits, stating, everyone speaks a different language; there are rules that get changed—twice a year minimally in different regions and if you don't do it right, you're going to lose money.

Digital transformation

"With COVID driving digital transformation across industries and companies and AI helping to streamline operations, more companies are looking for service providers that satisfy their needs so they can focus on becoming more competitive and selling to more customers," Eaton-Cardone said. "It may sound a bit cliché, but it comes down to dollars and cents. Merchants and acquirers need to understand their costs so they can make the right decisions."

Talapan has also seen significant progress with AI-powered tools and technology. Service providers have evolved beyond delivering instant analysis and proposals going deep with technology that shows interchange classifications, card brand dues and assessments and processor-specific data points, he stated.

"This provides an unparalleled level of detail and allows crowdsourcing of knowledge that benefits the entire community," Talapan said. "It's one more essential tool for understanding margins and profit potential that can be used to target different industries and custom cases, dramatically reducing time to market and accelerating sales." end of article

Dale S. Laszig, senior staff writer at The Green Sheet and managing director at DSL Direct LLC, is a payments industry journalist and content strategist. She can be reached at dale@dsldirectllc.com and on Twitter at @DSLdirect.

The Green Sheet Inc. is now a proud affiliate of Bankcard Life, a premier community that provides industry-leading training and resources for payment professionals. Click here for more information.

Notice to readers: These are archived articles. Contact names or information may be out of date. We regret any inconvenience.

Prev Next
A Thing