Monzo: A Peek inside the Playbook of the World's First Truly Digital Bank
When most banks are bribing people to join with £100 or a Rihanna CD, Monzo is doing the opposite: building a bank customers love, involving them in its development, and being so amazing the network blows up like Facebook. Monzo, the bank that broke Crowdcube, now has more than 100,000 customers. Next milestone: 1 million. Here’s how they’ll do it. (subscribe to FinTech Insider to hear the whole story)
The next goal, really, is a million customers. I think below that, we’re subscale, we don’t matter. At a million customers, the banks, the big banks, will sort of sit up and start to take notice.
What we’re trying to do is build the first bank with real network effect, so that you are encouraged to bring, you know, your social circles in, because it makes the product better for you, fundamentally. Skype is a great example of this network effect. The more people you bring onto Skype, the more people you can call. And I don’t think any bank has really ever thought that way. They’ve always thought your savings account and your mortgage are very personal, private things. Actually, money is very often a social thing.One of the things that you’ve been doing while you’ve been building the bank is the application process for a license. Tell us a little bit more, how has that gone? It’s the hardest thing I’ve ever done. It’s been robust and challenging, but I think it’s put us, as a business, in a better place as a result. You know, actually having to think hard about things like governance and risk. I think those are good concepts. They can be done very poorly, and really slow you down, but I think many startups don’t have that pressure to have a structured board, or minutes, or an understanding of what their main risks are and how they’re mitigating them. I think being forced to think about that early, in many ways is very helpful. This feels like, sort of, a golden year of challenger banks coming through N26 in Germany, and Nubank in Brazil, I think, are a year or so ahead, and I think they’re both pretty good indicators of where the UK will go, as well. But I think we need to get to multiple challenger banks at a million plus customers to really say this has now happened. I think we’re all just getting to the starting line, and the race hasn’t quite yet started. I see this week there’s been really interesting blogs coming out about risk and security. It’s been our Financial Crime Week, so we had a series of blog posts which are really, really interesting, and it sort of is the sexy area of Monzo. They’re actually doing some cool stuff. So using a lot of data analysis and machine learning to combat both third-party fraud and first-party fraud, and automate a lot of the processes that take up the time. So we’re really, really proud that we’ve managed to reduce our fraud exposure, too. We were losing a fair amount of cash in the early days, £30,000 or £40,000 in a week, we lost at one stage, and we’re down to often single digits now, of fraud loss. And that’s rules engine, plus then a machine learning engine, that’s really effective. I mean, it’s something we’re going to have to continue working on, to remain ahead, but yes, it’s a real center of excellence, if you will, inside the business. How are you guys set up? We’re at 75 now. There are a few principles we use, which is we hire very smart people and give them a lot of responsibility, and really try to push decision making down as far as we can, to the people who are actually doing the work day to day. Having cross-functional, decoupled teams is something that’s very important, as well. So not having an engineering department and a design department and a compliance department, but having a financial crime and security team, for example, which is made up of a lady who used to work at HSBC, and some fraud analysts, and some engineers. And they sit together, and work together, and do their daily send-up together, and sort of attack the same goals together, using their different specialties and toolsets. So they can operate very, very quickly to react to new threats and push out code, or update the product, or you know, change the terms and conditions, even, sometimes. They can do that without having to go to other teams to, you know, get blocked on their six-month roadmap. They have the capability inside that team to ship code. So small, autonomous teams.
We’re taking the best bits of what we see from the Spotify engineering culture.Everyone in the company goes through, effectively, an A-level in banking now, that’s one of the things you do in your first month or two at Monzo, and then a bunch of specialist training on treating customers fairly and financial crime, and all sorts, so that these people have the authority and the know-how, and the specialism, to be able to make the decision. And I guess, sort of, not punishing failure. That sounds weird, but allowing people to take fast, small decisions, and if one or two go wrong, if they can be corrected and rectified, actually that’s a better outcome than spending six months making a really big decision where you don’t know the outcome. So moving pretty quickly, taking small steps, so a couple of stumbles aren’t fatal. We use these microservices where you can update separate bits of code, you can deploy certain features to only a subset of your customers. So, you know, yesterday, for example, we pushed out a spending report. It was, sort of, a new trial. We pushed it out to 1,000 people first, and it was broken, right? It said February, not January. But no one-, a couple of people noticed because they were in that 1,000, and we turned it off and we went back, rewrote the code, redeployed it, tested it again. It was like, “Okay, this works now, now let’s roll it out to 100,000.” So you can do these kind of stage releases, so that the risk of it going wrong is very low, and the first 1,000 people are kind of opted in to be guinea pigs. Is this a pace you can keep? We’ve slowed down because we’re going through launching the current account, and there’s just a lot of work you have to do upfront with a big regulatory barrier, getting our restrictions lifted, We’re trialling internally faster payments and direct debits and all that great stuff, and seeing how we can add the Monzo-like delightful polish to a direct debit. And it is possible, but we can’t roll that out yet, because of the regulations. So by growing the team a bit, by decoupling them and, sort of, getting that regulatory barrier lifted, I’d hope we’ll actually be able to move faster again. Another thing that we have done quite well, but I’d like to improve a lot, is giving visibility into our design process. So we have a community forum which is incredibly heavily trafficked by about 5% of our user base. And Hugo, our Head of Design, is on there a lot, and Tristan, who writes all our great copy, spend a lot of their time there. And Hugo posts sneak peaks of new features which are incredible. You can actually, sort of, explore these designs and these prototypes and comment and feed back, and say, “I don’t like that, or that,” and it really does change how the product is built. I’d like to bring those sneak peaks more fully into the app, so that if you’re not a hardcore power user who goes to the forum, you still see that process. We’ve got three or four new banks coming into the market. Who’s going to win, and where are we going to go? I think the two can co-exist. So I’ll take Atom and us. I think actually, there are very real synergies. I think Atom, as you say, are taking the, kind of, balance sheet heavy, get the cost structure low and, sort of, lend-, take deposits in and lend out, and make sure you can get your NIM, and just do it at a much, much lower cost of servicing than the big banks. I think we’re going the total opposite approach, which is very balance sheet light, which is all about data identity, and those two plug in really, really nicely. It’s hard to see where-, or it’s hard to say now where Tandem, or someone like Starling, are going to end up, without having seen a product, so I’m looking forward to having a play, but I think that kind of marketplace model supports a range of challenges. The winds are well and truly in your sails to make this a fantastic year, right? A lot of the stuff the government has done up to now have been very mechanical. The current account switching service is a great example. It’s mechanically easier to switch now. There hasn’t yet been a reason to switch, but yes, they provide a lot of those building blocks and those enablers so that someone, hopefully Monzo, or Tandem, can come in and really ride those winds, as you say. But it feels like, yes, a pretty interesting year or two ahead. And PSD2 and open banking in 2018, as well, I think, will just give that another kick. From a marketing perspective, how are you going to break out of London? I think it’s product and word of mouth. I don’t think it’s marketing, or at least not in the traditional sense. It’s not above the line Tube ads, although we did try some Tube ads. They don’t work. It’s product.
How did Instagram or Snapchat or Facebook grow? It wasn’t above the line advertising. It wasn’t really even marketing, certainly not for their first 50 million users. It was an amazing product with true network effect, that’s very, very easy to sign up to and really gets you hooked in.And so we’re just trying to learn all of those lessons from consumer internet about funnel optimization, about onboarding, about making a seamless, slick experience. About virality in referrals. Like, I love Dave McClure’s pirate metrics for startups. It’s, sort of, acquisition, or even above that, awareness, acquisition, activation, retention, referral, and revenue. Just focusing on that funnel, basically, again and again, and just hyper-optimizing it…Something like 85 or 90% of our new signups come through referral. Nubank have done it. They’re at, a million, or two million customers in Brazil, and have used virality and network effect as their primary growth mechanism. You measure K=factor, for each user, how many additional users do they bring along? I think for Nubank it’s above one, which is the holy grail. Basically, until you hit market saturation, you just grow exponentially. Every user brings another user who brings another user. So that’s a really interesting, powerful mechanic that I think certainly the UK banks have touched, and we’re just doubling down on it. It’s a different approach to bribing customers. Yes, for sure, and basically every other bank seems to have this price that they’re willing to pay. It’s between £100 and £150, and some just say, “We’ll give you £150 if you switch to us,” and they’re very upfront about it. Others say, you know, Tesco Bank, 3% interest up to £3,000 a year. It’s like, okay, that’s a £90 bribe, but you’re packaging it a different way. To close out, what’s the golden rule? What’s the golden rule you live by? I’m going to rip off Y Combinator: “Make something people want.” Enjoyed this interview? Share with your friends and subscribe to FinTech Insider for more fantastic stories. Want to build a next generation bank like Monzo? Come talk to us. Jason Bates led proposition and product development for Monzo and Starling banks before co-founding 11:FS. Drop us a line at firstname.lastname@example.org.