top of page

False Positives: The big problem you don’t know you have

Writer's picture: Justin PikeJustin Pike

Any merchant who sells online knows about payments fraud and chargebacks. This is a significant global issue costing over $40 billion a year and is only possible because the way we pay online isn’t bulletproof. 

 

In-store, we pay with a physical card and, depending on the size of transaction – a PIN. Card and PIN has been around for a couple of decades and has proven - without a shadow of doubt - to be the safest way to pay. No matter the sophistication of a fraudster, they cannot read minds to obtain someone’s PIN.  


Online however, there is no PIN. And there is no physical card. There’s only the information contained on the card, which anybody can enter. There is no categoric way to prove that the person entering the card details online is in fact the card holder. And this is why there are $40 billion in fraudulent online transactions taking place every year.  


To combat this issue, the average online retailer uses up to five AI systems to analyse a range of data to assess the probability of each transaction being fraudulent. The AI looks at things like:  


  • location of the device the customer is using 

  • the purchase history of the buyer, for example if the buyer has already purchased a similar item recently 

  • whether the item being purchased is in line with the buyer’s normal purchase profile 

  • volume and frequency 

  • and so on …. 


The payment process looks something like this: 



If the AI deems that the transaction is likely to be fraudulent, it will not allow the sale to proceed. And the merchant will receive a comprehensive report detailing all the ‘fraudulent’ transactions that have been stopped.   


You might wonder why I say ‘fraudulent’.  That’s because there’s a very high chance that the majority of the transactions AI deems as fraudulent (and therefore blocks), are not fraudulent at all. This is called a false positive.   

 

Now here’s the big question – how would a merchant know how many of the transactions the AI is stopping are in fact legitimate? The short answer is, they don’t know. In fact, I would say most merchants have no idea they are missing out on a staggeringly high volume of legitimate transactions that have been deemed as fraudulent. How do I know this? Well, a recent article by Mastercard reports that false positives are responsible for $443 billion in lost revenue each year. 


This is more than 10 times the cost of chargebacks

Now, while lost revenue is a substantial cost, there are indeed other problems that false positives bring: 


  • Diminished customer relationships: There’s nothing like being declined a transaction to erode the trust and goodwill of a legitimate customer. 

  • Reputational damage: Word of mouth spreads quickly and no doubt any customer who has had a transaction declined is going to talk about it.  

  • Time and resources: Time spent manually reviewing transactions that have been deemed fraudulent. The greater the number of transactions deemed fraudulent, the more time and resource required to analyse the information. Not to mention the significant time spent using that data to make decisions.  


Overcoming the false positives challenge

A quick internet search will give you resource upon resource on how to combat false positives, and there will be a myriad of (time consuming) recommendations to minimise the risk and impact to a merchant’s business. But these steps and measures all work within the current landscape. That is, they don’t get to the source of the issue, which as I talked about right at the beginning of this article, is the fact that the way we pay online, is fallible.  


I think you’d be hard pressed in your internet search to find a false positives workaround that actually addresses the root cause of the problem.


The answer, of course, is not to throw more brain power or fraud combatting AI at it. Surely it is to pay online as we pay instore.  Quite simply, to make card present payments possible over the internet. Reliable and trusted Tap and PIN.  The good news is that this is already possible, with Burbank’s groundbreaking technology and it’s called CPoI, (Card Present over Internet). 

 

CPoI turns consumer devices into the merchant terminal so they can facilitate card-present payments over the internet. During the checkout process the customer simply taps their card against their own device, and securely enters their PIN, creating a card present transaction.  This offers a range of benefits including liability shift (away from the merchant), lower processing fees, and a significantly easier customer experience. But the best part, is that there are zero false positives with CPoI, which in turn significantly increases merchant revenue.  And therefore, less reliance on fraud detecting AI systems. 


Let’s face it. Online payments were ripe for disruption. Merchants cannot keep increasing tools and technology for combatting fraud, nor can they continue to wear the cost of false positives. It simply is not sustainable. And now with CPoI, they don’t have to.   

 

If you’re interested in hearing more or seeing a demo of CPoI, reach out to the Burbank team.  

Comentarios


Los comentarios se han desactivado.
bottom of page