Exclusive: ‘One step at a time’ – Paul Taylor, Thought Machine in “The Fintech Magazine”
Thought Machine Founder and CEO Paul Taylor believes Cloud-hosted banking infrastructure is irresistible. But how, if you’re an incumbent, do you achieve it? The answer is ‘carefully’
It’s been estimated that more than one in five adults have opened an account with a challenger bank in the UK, the majority looking for a higher degree of personalisation, reasonable and transparent fees, and superior customer service.
Not much to ask, you might think. And certainly challengers, almost exclusively born in the Cloud, don’t have to worry about how their infrastructure will cope with delivering it: it just does.
Legacy banks, on the other hand, have a tendency to ‘spend more of their time worrying about infrastructure than they do about banking’, according to Paul Taylor, CEO and founder of Thought Machine, whose software-as-a-service core banking platform, Vault, is designed to give them the head space to focus on the customer.
Cloud-native Vault is becoming a go-to for new-to-market as well as Tier 1 banks. Notable among the latter is UK banking group Lloyds, which bought a 10 per cent stake in the company in 2018 during a Series A funding round and came back to support a Series B raise, led by Draper Esprit with an extension led by Eurazeo Growth that closed at $125million in July.
Lloyds has adopted Thought Machine’s ‘incremental approach’ to migrating its 30 million customers to Cloud-based services. Banking giant Standard Chartered, on the other hand, chose Vault to power its new standalone digital bank Mox in Hong Kong, while UK challenger Atom Bank uses it as its core. The fintech recently sealed a deal with pan-European banking service Monese, too, and ‘all your cards in one’ fintech Curve has selected Vault for its new Curve Credit offering.
The Lloyds and Standard Chartered tie-ups are indicative of a rapid growth in partnerships between incumbent banks and infrastructure technology companies such as Thought Machine, revealed by a recent Banking Circle survey. It showed that 80 per cent of retail banks and 74 per cent of commercial banks in
Europe have worked with such providers. That’s good news for Taylor.
“Banks have got drawn into becoming infrastructure-heavy companies,” he says. “Banking is about providing credit to customers, managing payments, managing deposits, and being a secure store of value for your customers. Its principles haven’t changed since banking was invented in the early Renaissance. But it’s nearly impossible for a bank to get its energy focussed on this because the demands from the technology and the constraints from the technology, the issues with regulators, the issues with data security, are dominating.
“A big problem with traditional core banking systems is they really are designed for branches and, if not branches, then call centres; they’re really not designed for immersive, real-time, data-centric user interaction,” adds Taylor.
“Many developers of apps have become very frustrated working in traditional systems because they just cannot get access to the information and the features they need. A modern core, that has been built to enable the bank to do whatever customer proposition it wants, frees them up.”
That said, banks must select their infrastructure partners carefully.
“What you want is a suite of technology providers that you trust to be capable, of doing what they say they can do,” says Taylor. “There’s no point in using a know-your-customer verification company if you’re just as worried about them as you would be about your own system, for example. It’s the same with the core banking engine. Our goal is that the bankers should just be thinking about, ‘how do we want this credit card to operate? How do we want these fees to operate? How do we want interest payments? How do we want to extend credit? How do we make all this available through apps to the customer?’. They shouldn’t have to worry about how it actually works under the hood.”
Despite Banking Circle’s upbeat assessment of the sea change in attitude towards allowing infrastructure partners to take the back-end strain for a bank, Taylor says there is still a long way to go.
“I would like to see banks get back to the core business of how to lend, create, and deposit money and from that I think we’d all be in a better place.”
With so much of the world’s new technology being developed using Cloud-based systems, that’s the best way for them to achieve it, too.
“Cloud is the target platform for nearly every enterprise in a five-year time frame” says Taylor. “We also believe that it’s probably got a 15 to 20-year lifecycle. We’re happy we can build Cloud-native technology and not be looking over our shoulder for something radically different coming down the line in two to three years.”
The silent enabler
One of the clear advantages that Cloud has over on-prem systems is scalability.
“There’s technology scaling, and then there’s market and customer scaling. It’s not for me to say how a bank goes from a 100,000 to a million customers; they’re very clever at doing that,” says Taylor. “But the key benefit of the Cloud is what we call horizontal scalability, as in it should be able to deal with any level of traffic, with basically zero interference or worry from the bank.
“Look at Netflix,” he says. “The key reason why it works is that if you were using a traditional architecture, you’d be freaking out as to how to balance all the streaming technology every day – and, as everybody knows, streaming demand doubled as everybody went into lockdown. A traditional architecture wouldn’t cope with that. But the Thought Machine system does; it’s been designed to. A genuinely Cloud-native system, written with a microservice architecture, employing Kubernetes and horizontal scalability, will give you unlimited traffic capability. Then a sophisticated core banking engine will give you the ability to add in all the products you want, in just the way you want.”
Banks should look to the Cloud as a silent enabler, he says.
“I would be happy if not a single customer realised the bank was working on our infrastructure, working in the Cloud. What I care about is that it works, and that the end user is happy. If the end user is happy, then the banks will be happy.”
Getting there, though, if you’re currently operating on a legacy system requires careful handling. What he is not advocating is a ‘big bang’ migration; in fact, just the opposite. TSB’s monumental IT failure, which locked 1.9 million of its customers out of their accounts in 2018 ultimately costing it £330million, surely disabused banks of that idea.
“I think it’s impossible to contemplate a big bang scenario now,” says Taylor. “These things have to be done carefully. I would say at least do it in two steps. Firstly, prove that the new architecture can do everything it’s meant to do. For example, create a Cloud-based copy of a traditional bank, run it with some new customers, who onboard organically, try out the systems, make sure everything’s working correctly. Hybrid platforms do not seem to have worked very well – it seems to be nearly impossible to get an API microservice architecture to operate efficiently with, say, an old-style credit card system, so most of the discussions we’re having with banks is about creating a completely new bank in the Cloud, where everything, the payments, card, anti-money laundering, onboarding, credit scoring systems, etc, are all, in some sense, Cloud native and they interact with Thought Machine’s core banking engine.
“Once you’ve figured out how you get a genuinely Cloud-native architecture of all the components, then put together a product-by-product, or a region-by-region migration plan. Those two things together give you a good chance of getting it right.”
Launching a speedboat bank is, by comparison, of course, much easier, as demonstrated by Standard Chartered’s Mox. But Taylor believes there is now an unstoppable force for major banks to go Cloud digital, judging by the conversations his company is having with many of them.
“Those conversations are all versions of ‘within a five-year time frame we want to be in the Cloud and we want to be free of legacy. We want to be able to operate faster and operate without this enormous cost’,” says Taylor. Again, it’s not much to ask, is it?