<img src="https://ws.zoominfo.com/pixel/BUJfPb8NrEnpjSiz8kRz" width="1" height="1" style="display: none;">
Artificial Intelligence

How CFOs Can Tell the Difference Between Little Fence vs. Big Fence Issues

on November 07, 2018

As any good CFO knows, there are significant risks in corporate finance that require the use of financial systems to monitor issues and prevent them from blowing up.

One way of evaluating risks is look at them as little-fence issues or big-fence issues.

Big-fence issues are things like fraud that must be addressed with vigor. They are clearly wrong and intolerable. An example of a big-fence issue is when an employee books an international business trip and doesn’t take the trip. Instead, he gets a coupon from the airline, sells it, pockets the cash, and stays home for a week.

The little-fence issues can be more difficult to catch because they occur less often, and when they do, they might be out there on the edges of your business. One example is an employee upgrading to a luxury rental car during a business trip

In past decades, many companies would let little-fence issues go because the money they’d spend chasing after these problems or the people causing them was more than what they’d recover. The juice wasn’t worth the squeeze.

Now companies can use artificial intelligence systems to monitor issues big and small. An automated auditing and compliance monitoring solution like Oversight Insights On Demand can flag the person who constantly books luxury hotels (little fence) and the person who approves payments for every invoice he generates (big fence).

It’s important that CFOs determine the size of these issues when the automated auditing solution delivers data because the employees causing little-fence issues can be educated on how to make better decisions. The big-fence actors should probably be fired.

Let’s look at a couple examples of big-fence issues we’ve run across during our work with businesses and then lay out a blueprint for dealing with little-fence issues.

An Expensive Dog-Sitter

Here’s a startling statistic: the Association of Certified Fraud Examiners states that three out of four people who commit fraud in travel and entertainment are engaged in other forms of occupational fraud. We saw this firsthand in our work with one company.

Using Oversight, we noticed that every time one employee traveled anywhere on business, there was always a booking at a Spa Paws Hotel. The booking was always at the same location, even though the employee visited different cities on his trips.

Next, AI alerted us that the employee hired the same ride service every time he went on a business trip, even when he had rented a car or used Uber. The ride service was always the same amount and the merchant category codes obtained from the credit cards did not make sense for the categories he entered in his expense reports.

The company’s executives knew we might be looking at a big-fence issue.

Further investigation showed that Spa Paws was a high-end dog kennel. This employee put up his dog at Spa Paws and paid a doggy limo to pick up his pet every time he left town on a business trip. Obviously, this was not company policy, but he’d done a great job disguising it, tagging the company to the tune of $12,000.

It went deeper than an expensive dog sitter, however.

The corporate security department figured out that the same employee had sold over $200,000 worth of company equipment on eBay. What the AI found in travel and entertainment was a smoking gun that helped uncover additional fraud.

A Secretary’s Credit Card Fraud

Another story involved a secretary gaming the system. The first thing AI found was a high number of out-of-pocket expenses. Instead of using the corporate card, she said she used a personal credit card. She faked receipts and approved her own expense reports because she had her boss’ login credentials.

When the purchase card department asked about these expenses, she claimed she’d lost her corporate card but had since found it, so there’d be no further issues.

Fortunately, her response was logged as part of the case management process.

Six months later, the same credit card issues resurfaced and so did the old excuse, but because the system had recorded her initial explanation, the issue was turned over to her boss, who figured out she was using his account to approve expenses.

Turns out, her fraud amounted to tens of thousands of dollars.

These stories illustrate two ways AI helps you deal with big-fence issues:

  • Robo-auditors can flag little, unnoticed issues that could be the tip of the iceberg, like the dog owner who was selling company equipment on eBay.
  • The automation side of an AI system allows you to address issues over time because every bit of data is recorded and can be analyzed later.

 Without an AI system, these big-fence issues don’t always surface.

Taking Aim at Little-Fence Issues

Little-fence issues are not necessarily unethical or criminal issues, but they are unreasonable. It’s not a case of employees stealing from you but a case of people not operating to the company’s standards. Still, they can cost your company money.

AI can help you spot these policy compliance issues, but you’ll want to address little-fence issues differently.

For instance, going to an employee about a problem with their expense report—say upgrading a car or hotel room—can come across as being overbearing.

But with AI, you go to that employee with seven issues that have occurred over the last two months and show how their decisions cost the company an extra $1,000.

The employee can see the impact, and that more readily influences their behavior. The employee will also understand that their actions are closely watched. The combination of those two messages convinces them to change their behavior.

Using AI to spot little-fence issues can help managers and executives see they need to educate employees on corporate policy. Instead of an ineffective, generic refresher course, AI can produce five examples of policy violations that can be shared during training. That message is more direct, personal, and meaningful.

As you roll out the AI system and get comfortable with it, you can adapt how you react to the little-fence problems. Ask yourself these questions:

  • Are you too strident, or should you be more strident?
  • What kind of feedback do you get from your employees?
  • Are your efforts having an impact?

Part of ramping up your skills with an AI system is learning the best ways to remedy problems and what type of course corrections will have the best long-term impact.

This is part of an ongoing blog series based on the recently released book, Robo Auditing: Using Artificial Intelligence to Optimize Corporate Finance Processes by Patrick J.D. Taylor, Manish Singh and Nathanael L’Heureux.

Manish Singh

Manish Singh is Executive Vice President of Sales and Client Success at Oversight Systems.

best practices guide