Buy Now Pay Later is a huge industry. The ability to spread the cost of purchases over several instalments is proving incredibly popular.
Bank of England figures suggest more than three million households in the UK owe £2.7 billion to BNPL lenders. Over £10 billion has been lent in the past three years. The most active demographic is the 25-to-34-year-olds – who are asset poor and still working their way up the corporate ladder.
Klarna, DivideBuy, Clearpay, Zilch, and Scalapay are the bigger names in the UK, but there’s a parallel sector emerging in the business market too. Providers include Hokodo, Billie, Mondu, Resolve, Two, and TruePay allow companies to spread the cost of payment.
A survey by fintech company Marqeta suggests 43 per cent of European businesses have used BNPL to cover a business expense. Around half did so to buy stock or technology. The growth is impressive: these numbers are achieved with 30 per cent of European businesses saying they are completely unaware of the existence of BNPL.
What is really astonishing is the accuracy of the lending. Default rates are a fraction of credit card lending, despite a similar user base. At Klarna, default rates are less than 1 per cent, a full 40 per cent lower than credit cards. So, how do they do it?
‘BNPL lenders want to nudge their clients to behave obediently and pay back on time’
How Buy Now Pay Later works
The secret of BNPL is twofold. On the one hand, there is the intelligent use of data to decide who to lend to. On the other, investment in behavioural science. The BNPL lenders want to nudge their clients to behave obediently and pay back on time.
Let’s look at the first. Like all lenders, BNPL want to lend to responsible customers, who have both the ability and intention to repay. They also want to avoid fraud.
One of the largest global providers of consumer intelligence to the industry is Provenir. It provides services to Zilch in the UK, and until recently was the service of choice for Klarna, until the Swedish company developed an in-house function.
Provenir uses “more than a hundred” data sources, reveals Corinne Lleti, its director general of southern Europe.
“When someone gets to a checkout and selects Buy Now Pay Later, their information is sent to Provenir,” she says. “We take their personal information, their name, date of birth, address and so on, and we have what we call our marketplace, which allows us to bring in other data sources.”
Future of payments technology – Consumers demand speed, convenience and security when it comes to payments, which puts technology in the driving seat
One of the richest data sources are the credit bureaux, such as Experian, Equifax, and TransUnion. But until recently the BNPL brands refused to work with the bureaux, meaning the consumer debt position was obscured. Bad debts could be accumulated, which would remain invisible to other lenders. That’s changed. Klarna, for example, in May 2022 began to share payments with Experian and TransUnion. Zilch partnered with Experian. But the industry remains split on participation.
To make the picture richer, Provenir and other intelligence providers turn to both market data and publicly available information. Social media, such as Facebook, LinkedIn and Instagram are checked. The date of account creation, and relationship with other verifiable humans, are crucial. Fake accounts are spotted this way.
Another crucial method is to examine the transactions on the consumer’s bank account. This is made possible by the emergence of Open Banking standards. Open Banking is a protocol whereby account information can be shared with third parties, with the permission of the account holder. The BNPL provider, or a partner, can examine the bank account of the customer to see what they spend money on, and how their creditworthiness stacks up.
But the real trick is knowing what to do with the data, says Lleti.
“Open Banking data is, in its own right, stupid data or dummy data. It’s how you bring intelligence to the information. What do they spend on housing, on food, on gambling? You need the ability to interpret the information. Then you can build a picture.”
Artificial intelligence and machine learning are a big part of this. “The credit risk industry has always used forms of machine learning,” says Lleti. “Regression scorecards have been used since the 1970s. Now the algorithms are getting more complex. We also have a duty to explain our decisions. We can’t just create black boxes. Consumers need to know they are treated fairly.”
AI and machine learning are thus at the heart of BNPL. Over time, the algorithms are refined to better reflect the chance of default. Decisions, which must be made in a fraction of a second, become ever more reliable.
The front end
A big part of the BNPL story is the sheer convenience of usage. It’s so much easier to pay by BNPL on a shopping site such as H&M or Asos.com. The industry invests a lot of time and money into making the consumer journey ever easier.
“Our UX is award-winning due to our speed and ease of journey,” says Ed Massey, chief revenue officer at DivideBuy, one of the leading BNPL providers. “It takes less than two minutes from application to approval with a single-page form. That’s why we’re delivering conversion rates of up to 85 per cent – some of the highest in our industry.”
The UX plays a major role in ensuring consumers know what they are applying for.
According to Massey, understanding the full spectrum of lending options available to you, based on your individual borrowing circumstances, is key to making a considered purchase.
For example, DivideBuy’s Convert + tool shows you how much your monthly instalment would be, depending on how long you want to spread the cost for.
“Our Eligibility Checker tool lets customers run a soft credit check to see if they’re likely to be approved for the amount they want to borrow in advance – leading to approval rates of up to 89 per cent for some of our merchants,” says Massey. “Incorporating tools like these into our UX gives customers the insights they need to make empowered spending decisions, which is a big part of our mission.”
Another aspect is how consumers are protected by the BNPL providers. None of them wants customers to default, so are incorporating precautionary measures.
Klarna, for example, has introduced an “opt out” function. Consumers, in a quiet moment of lucidity, can request to be blocked from credit for a certain time period. Late fees work as a nudge to prevent missed payments.
“The fee is very small,” says Luke Seaman, head of public policy at Klarna UK. “The maximum fee is £10, applied both times a payment is missed. After then we get in touch and engage to set out a specific remedy to get them back in the black.”
Design and nudge theory are therefore as important to the industry as artificial intelligence. Informed consumers, who have the tools to manage their own accounts, ensure the borrowing is done on transparent and understandable terms.
The result
The impact of this investment is that default rates in Buy Now Pay Later are lower than comparable credit, such as credit cards. Activity in the industry is so frenetic the UK government unofficially delayed the introduction of regulation to BNPL. The Government was also worried overly harsh rules would curb lending in the midst of a cost of living crisis. For now, the fees and terms and conditions of this industry remain unregulated.
The industry is not against regulation – Klarna specifically calls for it to weed out bad actors – but the technology being deployed suggests this is an industry alive to the dangers of bad debt, and on a mission to forestall rogue consumers.
As the data sources get richer, and the AIs get wiser, Buy Now Pay Later can expand credit terms, and reduce defaults. It’s a remarkable picture of progress in a much-misunderstood industry.
More on payments technology
6 payments technology companies disrupting the sector – Here are some companies that are making waves in the payments technology space, bolstering transaction processes for businesses
