Fraud costs companies trillions each year and the vast majority of those deceptions are detected purely by luck—only a very small percentage are found by auditors.

Fraud in the supply chain can be particularly hard to spot because no business of any size can keep watch over every invoice, every order, and every consignment leaving a warehouse. The risk has grown as companies increasingly do business in parts of the world where bribes and kickbacks are commonplace.

Recessions typically increase the likelihood of fraud taking place; however, whether the economy is trending up or down, those intent on perpetrating fraud will do so. With these types of deception taking months, even years, to come to light, now is the time when many more companies are likely to find they are not selling what they think they are selling. Supply chain fraud affects businesses in every industry and at every stage, from raw materials through rigged supply contracts.

Using Data Mining

A major alcohol manufacturer was hit by a sophisticated counterfeit scam, with employees smuggling real bottles and real labels out of its factory to put the fake liquid into. The fraud was only uncovered when one of the company’s salesmen decided to do a bit of digging after noticing a drop-off in demand at one of his best retailers. The shopkeeper could not explain the decline in sales, so the salesman took a bottle away to test and found it contained the unpleasant cheap knock-off.

So, is it possible to protect yourself from this kind of deception? A conventional audit would not necessarily show such a fraud. However a specific fraud check, using data mining techniques to analyze all of a company’s information trails—orders, invoices, client and supplier databases, employee details—for indicators of wrongdoing is far more likely to succeed.

In more than 90 percent of fraud cases, ‘red flags’ will go up when such checks are done—these red flags may include supply companies with the same address as an employee, invoices with very little detail of the services provided or for very round sums of money, invoices that appear to be generic in style and structure, and transactions with jurisdictions other than where the vendor purports to be located. In the case of the drinks company, a discrepancy between the number of bottles and labels used and the amount of liquid produced could have raised the flag.

Technology is playing an increasing role in fraud investigations. In one case, a client in the hospitality sector could not understand why its volume of waste was so high and called K2 Intelligence in to investigate. Combining anomaly detection software with staff interviews, forensic analysis, and a review of communication patterns, we were able to analyze point-of-sale data in a wide variety of formats and from a variety of different locations to determine normal patterns of behavior. We then highlighted the anomalous patterns that served as leads for further investigation. Ultimately, the staff involved in the fraud were identified through the use of data analytics, allowing the company to take corrective action, reduce internal risk, and prevent further waste.

Small Signs That Something Is Wrong

Most fraud is discovered through small indicators that something is not quite right; be it inflating expenses or the same two suppliers always winning contracts.

An Australian client asked K2 Intelligence to investigate why one particular company kept winning $200 million contracts despite looking like an outsider at the beginning of the bid process. K2 Intelligence helped to trace payments from the winning bidder to a company owned by the wife of the CEO of the client. It turned out the bidder was making payments to the CEO each month in return for the lucrative contracts.

Within the corporate intelligence industry, there has been a substantial increase in preemptive fraud checks, particularly for companies with businesses or suppliers in countries where bribery is expected, and in some cases accepted. In these cases, companies may ask for a regular fraud check to be carried out every three or six months. The earlier a discrepancy is detected, the more likely a company will be able to stop fraud, minimize losses, and in some cases recover its funds before the money is spent or sent offshore.

Carrying out regular checks with the right technology designed to analyze a company’s financial data is a very powerful way to help prevent fraud before it occurs.