Choosing the right fraud models during the COVID-19 outbreak

The Covid-19 outbreak has meant that industries across the board have had to adapt to different ways of operating while countries around the world are in lockdown.

From a fraud perspective, companies will be facing the same problems as before the pandemic, only on a greater scale as financial pressures cause more people to commit fraud. This can be broken down into friendly fraud and malicious fraud.

Friendly fraud occurs when a consumer deliberately disputes a transaction that took place with a merchant, despite that transaction being completely legitimate and the goods arriving on time and in one piece. With the amount of payments taking place online and the number of deliveries increasing, this type of fraud is rapidly growing.

In addition, we’ve seen a rise in identity theft and phishing attacks as fraudsters take advantage of increased online activity. Career fraudsters are playing on the fact that people are looking for new jobs and need financial support, by targeting them with phishing scams that offer financial aid or stealing identities through false employment websites.

Anyone doing business online, along with the financial institutions that service them, will be impacted by this, as the identities being stolen will be used to make purchases which will result in fraudulent transactions and increased chargebacks.

 Rising to the challenge

The speed at which fraudsters have responded to the opportunity presented by the global crisis is alarming but not unexpected – it now needs to be matched with urgency and expertise from online businesses, payment service providers (PSPs) and financial institutions that support the industry.

Companies in the business of preventing fraud are now the front line against serious criminal organisations. That’s a fact. But stopping ‘bad people’ is a cat and mouse game and right now fraudsters are staying out in front. Recent history shows us that the fraud industry has lost its focus and is quite publicly losing the chase - we critically need more relevant data points to predict, not just analyse, behaviour.

The worrying truth is that some tech advancements actually help and even harbour fraudsters. For example, the anonymity movement is creating a safe haven for fraudsters/criminals to collaborate and expand – that’s why we need to take a step back from things like machine learning and AI and take time to understand what’s really next on the fraud agenda.

 Moving towards the future 

The pandemic has brought a rapid mass move to digitization, across sectors, devices and consumer groups. As a result, it’s acting as an accelerator for companies throughout the entire payments value chain, who are having to adopt new and innovative ways to effectively handle fraud at this sensitive time.

Traditionally, to prevent fraud, businesses have used technologies that make smart decisions in the back end, through the analysis of information inputted during the payment process. This approach usually results increased friction during the user experience (UX) and customers dropping off before completing a payment.

Instead, they should use solutions that assess fraud on the front end of a transaction, such as whether the transaction is taking place on the same device it usually is, or whether the device has been used to make purchases in the past. I have seen this approach being used highly successfully during the outbreak as it makes security invisible and also tends to prevent false positives. In addition, it allows companies to collect compelling evidence that can be used to dispute fraudulent chargebacks further down the line. 

Light touch fraud models, where security is invisible to consumers and negative friction is absent, will be crucial for retailers and the financial institutions that serve them, when it comes to maintaining sales volumes and preventing fraud. The balance between fraud and friction is undoubtedly a vital element to businesses that want to preserve customer loyalty and trust, while also protecting financial data during the outbreak.