I had a conversation with a fleet manager in Texas about two years ago that stuck with me. He ran 60 trucks, mostly regional hauls, and he was convinced his fuel costs were higher than they should be but couldn’t prove it. His fuel card provider showed him reports. Everything looked normal. Per-truck spending was within expected ranges. No red flags.
Then he installed tank-level sensors and started cross-referencing the data against his fuel card transactions. Within the first month, he found that seven of his trucks were consistently showing purchase volumes that didn’t match what the sensors said went into the tank. Not by huge amounts. Eight gallons here, twelve there. Enough that it never triggered any spending-limit alert.
He did the math afterward. Those seven trucks, over the previous twelve months, had probably cost him somewhere around $35,000 in fuel that was paid for but never made it into his fleet. He’d been staring at fuel card reports the whole time and saw nothing, because the reports weren’t designed to catch what was happening.
That story gets at why fuel theft in logistics is so stubborn. It’s not that fleet managers are careless. It’s that the tools most of them rely on literally cannot see the problem.
What actually happens to the fuel
There’s a tendency in articles about fleet fuel theft to list the methods like a textbook. I’ll keep this practical instead.
The most common form of fuel theft in commercial fleets isn’t someone crawling under a truck with a hose at midnight. It’s a driver at an approved fuel station, using the company card, doing something slightly off. Maybe they pump 80 gallons but only 55 goes into the truck, and the rest fills their personal pickup parked one lane over. Maybe they swipe the card twice — once for the truck, once for a buddy’s car — and the second swipe looks like a split transaction on the monthly report. This is called buddy fueling and it accounts for a large share of internal fleet fuel theft. Unless you’re matching gallon amounts on the card against actual tank level changes on the truck, it looks completely legitimate.
Card skimming is the external version. Comdata found that it accounts for 75% of external fuel fraud in commercial fleets, and the FBI reported a 700% increase in skimming across all businesses. A driver’s card gets compromised at a pump reader, and then small unauthorized purchases start popping up across locations the truck never visited. The amounts stay low enough that they blend into the noise of a fleet fuel budget that’s already $30,000+ per truck per year.
Then there’s the really old-school stuff — physical siphoning. Someone pops the fuel cap (generic keys cost a few dollars online) and uses a portable pump to drain 50 gallons in under five minutes. This mostly happens at overnight truck stops, unsecured yards, and driver residences. A Swedish study of road freight companies found that fuel cap tampering was the dominant theft method, affecting nearly half of surveyed carriers at least a few times per year.
And then there’s the category that I think gets underweighted in these conversations: fuel waste that looks like normal operations. Four hours of daily idling burns 3.2 gallons per truck. Across 50 trucks for a year that’s about $145,000 in diesel that generated zero revenue. Whether a driver is burning fuel sitting at a loading dock for 45 minutes or siphoning it into a gas can at night, the money’s gone either way. Some industry estimates put overall fuel loss from theft, fraud, and waste at 5-10% of total fleet fuel spend. Motive’s survey of trucking leaders landed right around 5%. At scale, that’s not a rounding error.
The fuel card problem
I want to push on this point because I think it’s the core issue.
Fuel cards with PIN protection, spending limits, and transaction tracking are genuinely useful tools. They centralize purchasing, simplify IFTA reporting, they give you a paper trail. I’m not arguing against fuel cards.
But a fuel card tells you money was spent on fuel. It does not tell you that fuel ended up in your truck. Those are different facts, and the gap between them is where buddy fueling, phantom receipts, and skimming live comfortably for months.
Think about what a fuel card report actually shows you. Driver A purchased 82 gallons at the Love’s on I-44 at 3:15 PM on Tuesday. Okay. But did 82 gallons go into Truck #22’s tank? Was Truck #22 actually at that Love’s at 3:15 PM? Did the truck’s fuel level before the stop, plus 82 gallons, minus what the engine burned getting there, match the level after? Your fuel card can’t answer any of those questions. And without those answers, you’re trusting the receipt.
A fleet called Pure Freight Lines, based in Chicago, found that fuel card fraud was costing them between $4,000 and $20,000 per month. The swings were wild because skimming attacks come in bursts. They’d been running fuel cards with controls the whole time. The controls limited spending per transaction. They didn’t verify that the fuel being purchased was going where it was supposed to go.
How fleets that actually catch this stuff operate
The fleets I’ve seen get fuel theft under control don’t do it with one tool. They do it by connecting data that used to live in separate systems.
Tank-level sensors report fuel volume continuously. Not once a day, not when the driver fills out a log. Continuously. So when a truck parked overnight at a driver’s house shows a 15-gallon drop at 3 AM with the engine off, someone gets an alert immediately — with GPS location, timestamp, and engine status attached. AI-based monitoring systems go further by learning what each truck normally consumes on its regular routes. When consumption deviates from the learned pattern, even gradually over weeks, the system catches it. That’s how you find the slow-bleed theft that manual reconciliation misses entirely.
Then you pair that with geographic controls. You define which fuel stations are approved. If a card gets swiped somewhere else, that’s flagged. If the truck’s GPS says it’s in Memphis but the card was used in Atlanta, that’s flagged automatically. If 80 gallons were purchased but the tank sensor only registered a 55-gallon increase, the system catches the mismatch before the month-end report even generates.
And you tie everything to driver ID. Not just the card number. The driver. So when discrepancies show up, you’re not guessing. You have a specific driver, a specific truck, a specific location, a specific timestamp, and a specific gap between what was purchased and what arrived in the tank. That’s a conversation backed by data, not an accusation based on suspicion.
Fleets running this kind of integrated fuel theft detection see about 78% reduction in theft in the first year. The system keeps improving because the AI learns more about each truck’s behavior over time. The longer you run it, the better it gets at distinguishing between a legitimate consumption spike (heavy load, headwinds, mountain route) and an actual anomaly.
Nobody wants to talk about internal fuel theft
Sixty-two percent of fleet fuel theft is internal. I know that number makes people uncomfortable. It made the Texas fleet manager uncomfortable too.
But internal fuel theft usually isn’t a crime story. It’s a culture story. A driver fills their personal truck every other Friday because they see it as part of the compensation nobody talked about. Another driver regularly buys an extra 10 gallons and sells it to a neighbor at a discount. A third driver rounds up every receipt by a few gallons and skims the cash difference. Each incident is small. Nobody’s getting rich. But across 30 drivers over a year, it compounds into $30,000-$50,000 that the fleet never detected because per-truck fuel variance stayed within normal ranges.
The solution isn’t cameras in the cab or zero-tolerance policies that create adversarial relationships with your workforce. It’s transparency. When drivers know that every fueling event is automatically matched against tank data, GPS, and their ID — and when that system is explained to them openly — behavior changes. Not because people are scared. Because the opportunity for small undetected theft goes away. Same reason idle time drops 40-60% when you start tracking it. Visibility changes behavior.
Fleets without a fuel consumption monitoring system are operating on monthly spreadsheet reconciliation and trusting that the numbers are honest. That was fine when diesel was cheap and margins were wide. With fuel at 24-40% of operating costs and broker margins at 10-13%, losing 5% of your fuel budget isn’t something you can absorb anymore.
Before you buy anything, try this
Pull your fuel card transactions from the last 90 days. Match them against your GPS data, truck by truck. Look for the stuff that doesn’t add up: a driver fueling at a station that’s nowhere near their route, a truck that takes 90 gallons when it should have only needed 60 based on the distance since the last fill, transactions at odd hours when the truck should have been parked.
If your fuel card data and your GPS data are in different systems that can’t be compared easily — well, that’s the problem right there. The data exists. It’s just not connected. Most fleets sitting on undetected fuel theft aren’t missing the information. They’re missing the integration.
And look, maybe you run the comparison and everything checks out. In that case, you sleep better. But I’ve yet to meet a fleet manager who did this exercise for the first time and found nothing.
