Rafael Henrique - stock.adobe.co

Zopa’s emotional journey to becoming a bank

Zopa Bank’s CEO sees no reason why the app-based bank should not be as big as the high street giants

Zopa spent 15 years pioneering the peer-to-peer lending sector, only to drop that business to become a bank.

CEO Jaidev Janardana tells Computer Weekly how he helped make one of the first fintech firms even more fintech.

Janardana joined Zopa in 2014 as chief operation officer (COO), before becoming CEO a year later. At the time, it was a peer-to-peer lending financial technology firm, having pioneered the sector since its formation in 2005.

But within two years, Janardana, an engineer who moved into the business side of banking in the US banking sector, kicked off the process of transforming Zopa into a bank and closing down its peer-to-peer lending business.

Janardana grew up in India, where he trained as an engineer before initially working in the telco sector on network planning and design. After attending business school, he moved into financial services. In 2002, he joined Capital One in the US.

At Capital One, he was focused on data-based marketing and product design, where he got a taste for using data to personalise banking offerings, the lifeblood of fintechs. “Capital One was one of the first innovators in financial services, particularly around how to use data to create more personalised, customised offers,” he tells Computer Weekly.

In late 2007, Janardana moved to the UK with Capital One to manage credit risk for the bank’s UK business. This saw him hold the role during the global financial crisis, which hit the world in 2008. At this time, his tech background helped him with statistical modelling.

But it was at Zopa that he returned to his tech roots. “One of the things that attracted me to Zopa is that we could use technology in a far more meaningful fashion,” he says.

Early years at Zopa

When he joined as COO in 2014, he was focused on helping build a good lending business, making sure the company could use data and technology. This included using machine learning models to make credit decisions as early as 2015.

“One of the things that attracted me to Zopa is that we could use technology in a far more meaningful fashion”
Jaidev Janardana, Zopa Bank

He says this has provided the organisation, now a bank, with good foundations for the future, where artificial intelligence (AI) will play an increasingly important role in the finance sector.

“Today, we have eight years of data on the use of machine learning. A lot of businesses talk about [using machine learning], but very few can claim this much experience with it,” he says.

Zopa loans, where decisions are generated by machine learning, have performed well.

In 2016, with Janardana at the helm as CEO, the company was thinking about what was next for the business. “We decided that expanding our product set would be a very good idea,” he says.

This was the moment when Zopa set out on its journey to become an app-based bank.

Emotional goodbye

It was also the moment that spelled the end of its peer-to-peer lending business, which it had set up in 2005. “Lenders [on our platform] always made money, even through Covid, but ultimately, when we looked at the consumer sentiment, there was a lot of reticence. At the same time, regulation had become overbearing for peer-to-peer lending,” says Janardana. “We did not feel that business could scale and achieve compelling economics in the long term.”

“We have eight years of data on the use of machine learning. A lot of businesses talk about [using machine learning], but very few can claim this much experience with it”

Jaidev Janardana, Zopa Bank

He says closing the peer-to-peer lending business, which shut in 2021, was a very hard, emotional decision. “Zopa was the world’s first peer-to-peer lending company. We were very emotionally attached and thought it delivered great results for us and our customers.”

Janardana had his first meeting with UK regulators, the Financial Conduct Authority and Prudential Regulation Authorities, in November 2016 to inform them of the company’s plan to launch a bank. After a three-and-a-half-year journey to get its UK banking licence, Zopa Bank was launched in June 2020.

Today, Janardana sees no reason why a Zopa should not be as big as Barclays or HSBC. “I don’t see why not. We have been a bank for two-and-a-half years, and we are about to hit one million customers and about £3.5bn in deposits. If we can maintain that level of growth for five to seven years, we become significant.”

To achieve this, the company must stick to its roots, established in 2005. “The trick would be to make sure we don’t lose our customer centricity, ensure we continue to innovate and embrace new trends, technology and data, and transform those trends into good products for our customers.”

The digital banker

Janardana understands how technology can improve banking and uses his experience in the banking sector to ensure the company keeps up with evolving tech and consumer demand.

He applied these skills when he first joined Zopa, at the time a pure peer-to-peer lender. Although he had worked in financial services for a long time, he had not worked with loans.

“The first thing I asked was what people look for in loans. I was told they look at whether it will be approved or not, and what price they will get. So I asked when does the customer get this?”

The answer was that they get it three or four days after the process begins, which Janardana decided was not good enough. “At the time, Zopa did not make instantaneous approvals to the extent I would like,” he says.

When he joined in 2014, Zopa was instantly approving 20% of loan applications. “We upgraded our tech, and at the same time upgraded our credit modelling, so we could deploy our first machine learning-based models,” he says.

The proportion of loan approvals that were instantaneous increased to 60% after the machine learning models were introduced, and by about 2020, 99% were approved in less than five seconds.

Zopa Bank uses this technology to approve applications for car finance, credit cards and “buy now, pay later” (BNPL) products.

Making a fintech more fintech

It is inevitable that even the most advanced fintech firms will quickly lose their advantage if they fail to keep up with technology and customer demand, as well as changes in the banking sector.

“With technology, there is progression, and you need to embrace it and move forward. As a company, I believe we have done a good job of understanding the latest trends and how to use that,” says Janardana. “At Zopa, we have thought very carefully about marrying the ‘fin’ and the ‘tech’ properly.”

“With technology, there is progression, and you need to embrace it and move forward. At Zopa, we have thought very carefully about marrying the ‘fin’ and the ‘tech’ properly”
Jaidev Janardana, Zopa Bank

He says this paid off during the Covid-19 pandemic. “A lot of fintechs blew up, but we are still here. What we did was ensure we have a lot of people that are tech savvy alongside people that have experience of lending over a long period of time, through crises – people like me,” he says.

Such experience is vital, says Janadarna. “Models are just rear-view mirrors that tell you what happened in the past, but you can’t drive a car only looking in the rear-view mirror. You have to look forward, and that is where the experience counts. You can see where models might have loopholes and where they might fail. This helps inform where it might not be appropriate to use them.”

Zopa Bank now has about 710 staff, a third of whom are pure technologists, and about 100 work with products and data. These are mainly in the UK, with a small team in Spain and some customer service functions outsourced to a company in South Africa.

Zopa’s revenue was about £170m in its latest financial year. It has £2bn loans on the balance sheet and £3.5bn in deposits. It started off focused on savings and borrowing, but is expanding into areas such as BNPL, where it recently acquired fintech DivideBuy.

Read more about Zopa

Read more on IT for financial services

Data Center
Data Management