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

6 Tips for Using Keywords to Detect and Prevent T&E and P-Card Fraud

on April 18, 2018

Whenever details are overlooked, problems arise. As the saying goes, “The devil is in the detail.” This idiom holds true in travel and expense (T&E) programs. Identifying suspicious T&E and purchase card (P-Card) transactions requires more than simply the price or location of the transaction. You need to know what the expense is for, specifically. You need the detail.

Identifying suspicious keywords plays a crucial role in determining if a card transaction or out-of-pocket expense warrants further investigation for potential fraud and misuse. By automatically analyzing 100% of transactions for suspicious keywords, Oversight flags expense reports, line items, card transactions, and attendees that need additional review.

Tips for Using Suspicious Keyword Match in Expense Monitoring

1. Keep a Restricted words list of suspicious keywords that indicate a potentially unauthorized purchase. Keywords such as liquor, giftcard, jewelry or .com should raise flags, but don’t overlook words such as services rendered, government, facilitation and cash that may signal potential bribery or other corruption violations. Oversight Insights On Demand® clients can select from a pre-built list of keywords, which includes-built list, including adult, boat, Bitcoin, cannabis, casino, cigarettes, drugs, firearms, kennel, iTunes, iPad, late fees, lottery, limousine, massage, nightclub, limousine, lottery, upgrade and Venmo. Clients can also add their own keywords.

2. Use keywords that match your policy restrictions. For example, if Hertz is your company’s mandated car rental company and Marriott is the mandated hotel chain, then you can have Oversight scan for non-compliant vendors like Enterprise Rent-a-Car and Airbnb.

3. Require employees to enter a description of the expense and/or a business purpose field in order to gain approval. By forcing people to describe the expense, you can use analytics to monitor all parts of an expense for compliance. A blank description should automatically raise a flag.

4. Integrate Level 3 Card Data in your data analytics for more detailed transaction analysis to quickly find suspicious activity based on detailed lists of products or services purchased on a card transaction. For example, an employee shopping at a grocery store for authorized items to stock a meeting room might add in some unauthorized items for later personal use: a visa gift card, cigarette/tobacco products, personal care items, etc. By analyzing Level 3 item detail for keywords, charges can be flagged as suspicious if they include suspicious items.

5. Remove language barriers. Are you testing for compliance in local languages of your business units around the world? Corruption has its own language and your library of suspicious keywords should include languages for where you do business globally. Oversight Insights On Demand maintains a keyword library in 13 languages.

6. Rank keywords based on associated risks. Not all suspicious keywords are equal. Prioritize the keywords that are highest risk to your organization and configure your analytics solution accordingly.

To reduce compliance risk, innovative companies are implementing artificial intelligence-based data analytics software solutions.  With an AI-powered risk management and compliance monitoring solution like Oversight Insights On Demand, you can analyze ALL expense reports, purchase card transactions, and accounts payable data to identify and prevent fraudulent, wasteful and out-of-policy spending, track repeat offenders, and find the root cause of behavior that needs to be corrected.

Demo Oversight to see how you can detect and mitigate risk in your corporate spend programs.

Julia Versaci

Julia Versaci is Director of Solutions Development at Oversight Systems.

best practices guide