Of course, the answer is that in many cases - that claim remains just about the best there is our there, and some issuers can't even claim that. Why on earth is it that the huge advances we have seen in the safety and security of retail transactions seem to have left the fuel card space in the stone age? You wonder.
"Fraud" is a catch-all phrase which throws a blanket over a wide range of risks and an equally wide range of misdemeanours, some of them sophisticated, others less so - but all of them endemic.
It could be professional crime: skimming, cloning, organised theft. It could be site collusion - deliberate transaction manipulation or the "blind eye". (I've even seen a large site near Zeebrugge openly selling large plastic jerrycans and hoses, for some reason). It could just be the classic "driver's bonus". But it is unrelenting - to the point where some transport operators accept is as a necessary evil and budget for it. Bonkers.
But at the same time, the fraud departments of fuel card issuers all around Europe seem set up to identify and manage fraud (blocking cards, limiting liability, handling complaints) rather than taking innovative steps to actually stop it happening. As recently as 2016, we encountered an instance where one fuel card was used for a transaction in Calais - and a little more than 30 minutes later used again (successfully) on the outskirts of Milan. Valid card number; limit not exceeded; online authorisation. Bob's your uncle.
Really?
And yet the funny thing is this: if I fly to Ho Chi Minh City tomorrow and stick my debit card in the nearest ATM, there is every chance it will decline (whatever my bank balance). If, at two o'clock tomorrow morning, and in a fit of Pythonesque randomness I decide to buy five digital cameras from Tesco.com, there is - again - every chance that my efforts will be frustrated.
That's because the machine learning (or "neural") technologies increasingly defending retail banking transactions are - in a nanosecond - making informed decisions about how a transaction stacks up in the context of historical patterns and probability. I very rarely, if ever, go to Vietnam. I tend not to shop online at 2am, and I certainly don't buy multiple luxury electronic goods at that time of the morning. Looks dodgy. Decline. Job done.
So why is it that these machine learning technologies are not all over the fuel payments space and making it impossible to pay for fuel in Calais and Milan at the same time with the same card. According to Mark Goldspink, CEO of The ai Corporation, it would seem to be a matter of time:
“Even though Retail banking cards and closed loop, commercial fuel cards have different purposes, the underlying mechanics and processes are very, very similar. As a result, there are transferrable learnings that can be applied both ways. I would say it stands to reason that for a fuel card issuer or acquirer, an omni-channel fraud prevention platform has to be more beneficial."
Says it all. Service stations are retail environments, and will be managing a widening array of consumer and corporate payment mechanisms - plastic and card-less. Why on earth would there not be a single, integrated fraud prevention tool deployed across all payment types?
In the context of what is going to happen in commercial refuelling, this is a critical topic. Payment methods will evolve and become card-less; as licensees, dealers and franchisees replace company site managers, acceptance strategies will become liberalised; new products will come onto the market - e-wallets, fleet Mastercard and Visa variants; commercial sales will increasingly move through indirect channel partners. This hotch-potch of change is going to demand coherent and simple fraud prevention strategy, which goes to the core of issuers' competitive strategy and value propositions. Not an afterthought.
It's when and who, not if.
