Four key questions to help you identify fraudulent use of your business’s fuel cards

Fuel fraudsters continue to ‘skim a little off the top’. 

The post pandemic e-commerce boom has fueled a marked increase in demand for logistics services. As a result, global fuel and fleet markets have expanded their business operations to meet this increased demand. For criminals, this has created a unique opportunity to take advantage of the ‘new normal’ and ‘skim a little off the top’. The past year’s steep increases in fuel costs have also contributed to increases in fraud, with drivers 'stealing' fuel via their fuel cards for their personal use or on-selling.

There are many types of fraud that road transportation and fleet operators need to be aware of. By far the most common type is skimming and copy card fraud, which together account for most fraud observed over the last three years. Other examples of fuel fraud range from dual filling, where fleet drivers fill their vehicle, as well as a bladder tank, and charge both to the business, then sell the fraudulently attained fuel to third parties or associates; to sharing or selling fuel cards directly to their associates.

When fleet and logistics companies are busy and are operating at capacity, it can be very difficult for fleet managers to identify fraudulent behaviour and differentiate it from genuine activity. Cases of dual filling and sharing fuel cards are especially challenging to spot, because the outcomes look very similar when fleet companies are reviewing their costs for the month. Increased fleet sizes and the use of multiple fuel cards per vehicle also make distinguishing behaviours between drivers very difficult.

It is not the sole responsibility of fleet managers to detect fraud, but they MUST be aware of the risk and be involved in managing and mitigating the levels of fraud that are occurring within their fleet(s). Fraud detection strategies that are managed by specialised fraud teams without any input from fleet managers can be ineffective because of the challenges of distinguishing fraud from genuine activity. Insight and experience are invaluable and help to minimise the amount of fraud that is overlooked.

There are several questions that fleet managers can ask to help with the initial identification of fraudulent activity:

1. Has the fuel card been used by single or multiple drivers or vehicles? Knowing this basic information will enable you to measure or compare card volume and frequency of spend. Although the card could be used for multiple vehicles, the aim is to understand the performance and workload of each vehicle to determine its usage across the board.
2. Have your drivers seen any increase in demand or activity reflected in any increase in business? In a single card per vehicle type scenario, it’s easy to define the average MPG of the vehicle, as it is readable from the dashboard. You can then cross reference that with the fuel volume in any given period. In most instances, a vehicle rarely deviates from its assigned route/destination. If a deviation does occur, it would only account for minor increases in fuel usage. If you compare this to your fleet’s odometer readings, you can accurately define the distance travelled and increase the accuracy of your attempt to identify abnormalities in volume spent.
3. Are current values and volume in line with historical and expected usage? Fleet Managers can review recent activity to call out unexplained increases in expenditure. Following a pattern of historic behaviour can help identify if the activity is aligned with an increase in workload. Also, it would be worth exploring if specific spend brackets are offered at certain points in a calendar month. Comparing fuel card spends between vehicles or cards with similar spend can show if a card is being abused as the fuel delivery values will be quite different.
4. Are your drivers aware of card spending limits, and are they being reached regularly? Criminals often try to take advantage of monthly card limits. For example, if a fraudulent cardholder is aware of his/her limit and has a rough understanding of how close they are to that limit at any point, they will avoid going over it and raising suspicion. This could translate to increases in the volume of spending leading up to the end of the period of the spending limit, meaning that the criminal knows how much in the ‘bank’ he has left to utilise before the card limit is reset.

Fraudulent activity will continue into the foreseeable future, as it is almost impossible to eradicate it entirely, but with improved education and awareness there is hope. If fuel vendors and fleet operators are more informed, they can help to control fraud via enhanced detection methods. Fraud solution vendors need to work with the ecosystem to educate fleet managers and their drivers too. All parties need to know what fraud is, how it is perpetrated, and understand the consequences of either perpetrating or ignoring fraud.

In addition to raising awareness, advancements in using AI to monitor fleet fuel card usage is helping to mitigate the effects of fraud, giving fleet managers the tools they need to stop fraud before it occurs. New technology and utilising new data streams are increasing the levels of insight fleet managers have over their fleets. For example, advances in utilising telematics data potentially offer highly valuable insight into many aspects of individual vehicles, including MPG, current speed, co-ordinates, engine RPM, oil pressure, oil level and engine temperature, all of which can help with fraud identification.

We believe that fuel card fraud could be reduced by utilising multiple data sources, including a combination of payment data and vehicle telematics data in the future. For example, telematics data can help to identify when a driver has navigated off an assigned route. This may be due to unexpected diversions, but it can also be a sign of a driver taking advantage of a work vehicle for personal use. Many telematics data providers install small boxes into a vehicle to obtain and transmit this information enabling fleet managers to view vehicle information in real-time. This sort of insight could help to better understand and monitor potentially fraudulent fuel card behaviour, including:

Card Abuse: By defining the expected behaviour of a card from distance travelled to the expected volume of fuel used, vehicle performance can be tracked, and fleet managers can use this data to understand if the fuel card data matches with MPG data from an individual vehicle. If this data is drastically misaligned it is indicative of abuse. This could result in a significant increase in identifying fraudulent behaviour in the fuel sector demographic, where the full history of a fuel card is available.

Compromised cards obtained through skimming or physical theft: By comparing the geographic location of a vehicle with the card usage location from the payments data enables the fraud or fleet manager to determine if a card has been separated from the driver. This is a simple but effective method for detecting this type of fraud, but it only works when only one fuel card is associated with each vehicle.

Ultimately the best way of fighting fraud is for the ecosystem to continuously evaluate their existing fraud detection methods and optimise their fraud prevention rules and strategies. While investing in advanced machine learning detection solutions to proactively minimise losses, enhance the overall customer experience, and improve their fraud operational capabilities.