As digital, mobile-only bank Monzo passes one million users, Simon Vans-Colina talks to Andrew Hobbs about what a truly modern provider looks like in the age of Open Banking, and the critical role of APIs for companies such as this.

Internet of Business: Tell us about your role at Monzo…

Simon Vans-Colina: “I’ve been at Monzo since the very beginning, three and a half years ago. I was one of the first 11 people. I’ve had a range of roles since, but started off as one of our first engineers. Six months ago, I switched to doing product-oriented things, so I’ve been in the partnerships and revenue teams. I also run Monzo’s Open Banking API work, including being on the Open Banking Committee.

“After Open Banking launched and we introduced current accounts in January, I dredged up one of my dream ideas from the early days, ‘What would it be like if you had a bank that was built with an ‘if this then that’ (IFTT) conditional statement approach?’ And I managed to convince one of my colleagues to help me build it.”

What has enabled challenger banks like Monzo to emerge?

Simon Vans-Colina, Monzo
Simon Vans-Colina, Monzo

“The regulations changed to make it easier to start a new bank. It used to be that you had to have about £100 million capital upfront before you could start a bank, and it took a couple of years, so no one launched a new bank for 125 years.

“Then Metro Bank came along, though Metro did actually raise all the capital upfront and do a big-bang launch. Critically, the regulations changed in around 2013, which made it possible to get a banking licence first and then raise all the money.

“There’s also been a big technological shift. I’ve been working in IT in London for 20 years and there’s been a huge change in the way complicated computer systems are built, led by companies like Google.

“Google’s infrastructure is based on microservices and open source software, and the idea that you build everything yourself. This is very different to the way banks used to be built, where you’d go out to market, get a vendor, and they’d build the product and give you a warranty period.”

Following the arrival of Open Banking in January, you made an interim API, for third-party companies, available. Why is an ‘API-first’ approach important to you?

Monzo’s not a bank with an API, Monzo is an API.

“That means every single component in Monzo, every piece of software that powers Monzo, is an API. We have APIs that are exposed to the internet and we have others that aren’t, but broadly they’re all tiny little binaries that speak http and follow JSON API standards.

“It wasn’t very hard for us to make a public API, it wasn’t a separate project. It was just a matter of making our existing API available in a safe and separate way for people to consume.

“When I built the account information service provider (AISP) API, it was just a matter of taking some of our existing APIs and freezing them, so that they don’t change as often, and serving up the permissions to securely authorise those that had the AIS permissions from the Financial Conduct Authority (FCA), and give them access to those APIs.

The reason we did ‘if this then that’ was because I really believe in Open Banking. I’m a huge fan of giving people access to their accounts, data, and money through APIs.

“But to get access to Open Banking APIs you need to have a company and permission from the FCA. That requires insurance and penetration testing, which can cost £50,000. Therefore, only use-cases where there’s a clear commercial benefit are being built on top of Open Banking APIs.

“I wanted to create the ability for people to do things that don’t have a clear commercial incentive, which is why I pushed IFTTT. People who just want to build something cool, even if they’re not necessarily developers, can do so by using their data, accessing their money, or integrating their bank, without having to go down the AIS route.”

Do you see APIs being central to banking going forward, including incumbent banks?

“Yeah, I think so. It’s like the difference between a feature phone and a smartphone. Remember when you would buy a phone and it would have the features that it came with and that was all it was ever going to be able to do? And then smartphones came along and you could install new apps on them, and they could do anything. Anyone could write applications that would work on them.

“We’re at the beginning of this process now, in banking, but I think it will be the same sort of thing in a few years’ time.

You won’t want a feature bank account that does these five things, and that’s it; you’ll want a bank account that you can plug into everything – your way of life, your accounting software, how you choose to save and settle bills with friends.

“You’ll want a smart bank account. It could take a few more years but I think it’s inevitable that’s where we’re headed.”

The Open Banking API is expected by the end of the year. What’s your view on having a standardised API?

“I totally support what the Open Banking Implementation Entity and the Competition and Markets Authority (CMA) are doing, creating a standard API and making the big nine banks implement it. However, it’s not required for APIs to be a success.

“The majority of AISPs out there, like Yolt, Emma, TrueLayer and Token, are all connected to Monzo, even though we haven’t implemented the standard Open Banking API. Yet, I think it’s good for the market that the nine largest banks are required to implement it, otherwise I don’t think they would have done anything.

“That’s because there is a cost to banks, without an obvious benefit. They are having to give up some control. So, it was a good decision for the CMA to take the wheel and force them to produce the standard API. However, I don’t think a standard is necessary and sufficient to create the marketplace that the CMA wants to end up with.

“It’s one of those things where there should be a competitive market, where companies like Monzo can build an even better API than the standard. If it ever turned out that there was some great use case that Monzo customers couldn’t do – because we hadn’t implemented the standard – then we would do so immediately, but I haven’t seen that yet.

“That may change with payment initiation service providers (PISPs), early next year. If payment initiation takes off, there will be more of a push towards standard payment initiation APIs and I’m keeping a really close eye on that, and Monzo may just implement the standard API for that.”

You took a ‘ship it and iterate’ approach to Monzo’s development. Why is this?

“I don’t know how else you would build things these days. The lean start-up model, where you test a product’s market-fit as quickly as you can, listen to your users, and then ship and iterate, just seems like such an obvious thing. Even big enterprise companies attempt to follow agile development methodologies now.

“It’s not that they willingly choose to do something else, they definitely want to follow this model, but some of them have a history, a legacy, where they’ve built up organisations that aren’t able to do that. Everybody, I think, agrees that multiple lightweight processes – the lean development model – is the best model.

We deploy to production around 40 times a day at Monzo. Things very rarely go wrong because the delta between what we’re shipping and what’s currently live is so small. We ship tiny changes to features and test over and over.

“We might spend a week building something to test what the effect is on the market or on our consumers. Then we might ship it and watch the graphs and see if it works or not, and if it doesn’t we’ll look back. We won’t be emotionally invested because we didn’t spend six months trying to make something work. We’ll think, ‘Okay, that didn’t work, so we’ll try something else.’”

Why have some established banks been slow to respond to the opportunities and challenges presented by Open Banking?

“If you’re a big bank and you already have a huge proportion of the consumers, there’s no real upside to trying to change the world, you’re already incumbent, but there is a potential downside. If you make these APIs available, then FinTechs will come along and disintermediate you.

“They will build a better app on top of your bank and then you’ve lost the ability to sell more products through your own app. That’s what the incumbent banks are worried about. There’s been a lot of criticism recently that banks are very quick to consume each other’s APIs but very slow to push out updates to their own.

“I think that speaks to the economic model, that there is a strategic risk in Open Banking, which is why it was important that the CMA came along and made them march in lock-step to change this, because there would be a cost to being the first incumbent to produce an API. It forced banks to create a new marketplace together on the same date.”

Monzo has a close relationship with its customers, which is unusual for a bank. Why do you take this approach?

“When we started we had no banking licence, no technology, no customers, and no money. We had nothing and had to get it all at the same time.

“We had to test product market fit early on and we did that by forming a community. Tristan, our now head of marketing, was our community manager in our early days – he was around our 15th hire. We knew that having a strong, close relationship with our community was going to be critical. And we’ve just kept that.

“You’d be surprised how often you wrongly think you know what your community wants. We’re a really transparent company as well, so we’ll just go to our forum and be like, ‘Here’s something we’re thinking about,’ and then people will come back with a dozen ways you’re wrong, or how you’ve got your ‘Shoreditch bias’ in there and it won’t work for people outside. It’s a really powerful thing to have this huge community of people say what it is they want.

‘Listen to your users’ and ‘write code’ are the two most important things.

“When you’re a tiny bank and you don’t have any users, building a community as a proxy for not actually having any customers is incredibly useful.”

What about beyond the smartphone? Will we see banking increasingly done through smart speakers, cars, and other connected devices?

“We will go wherever the customer is, but how far are you from your smartphone right now? Most people are within a few inches of a high-powered computer that’s connected to the internet 24 hours a day. This is a strange world that we live in.

“I don’t see anything coming to change that, but I did do some work about 18 months ago to integrate Monzo with Alexa. You can ask Alexa what your bank balance is, and it’s a good novelty, but realistically you’re going to grab your phone and use that. So I don’t see any pressing need to go beyond that.

For interactive stuff your phone is going to win for a long time, but there are other ways to interact with your bank account, especially around automation, where the smartphone isn’t necessarily the best thing. That’s why we built IFTTT.

“One of my favourite examples is this idea of personal taxation – where people set up rules for themselves. For example, if I go to McDonald’s, I’m going to have Monzo automatically put £1 in my sin jar, and every time I record a kilometre of running on Strava I’ll have it take the £1 back out!

“If you do bad things you tax yourself, and if you do good things you pay yourself back. It turns out to be a powerful idea. Lots of people are building these systems that help keep their lives in balance. Even really simple things, like students being able to put their entire student loan in a pot and automate a daily budget, using IFTTT.

“Or if they support Chelsea, and Chelsea win, they can pay themselves an extra £20 to go to the pub. This sort of automation is another way of interacting with your bank that goes beyond just having a smartphone.”

You can hear more from Simon Vans-Colina on the future of financial services platforms, and the effects of Open Banking, at the Internet of Banking and Payments, which takes place on 20-21st November 2018, London.