How a digital bank with great growth metrics failed

In May 2023 Lucy, a digital bank for underbanked women entrepreneurs closed its doors after 3.5 years of operation. Below are some of the lessons we learnt from it.

First of all it didn’t feel good. Someone once said I had success bias because of successfully growing and selling Tigerspike. i.e. I had a skewed view that starting growing and selling companies is not difficult. Nothing was easy about what we did at Tigerspike but I sort of know what he meant; I haven’t experienced failure like this before and didn’t expect it. It really sucks, you have lost a ton of money but more importantly you have lost investors money. Investors who believed in the mission. This is the worst. Especially when for many of them its the first time they invest in risky startups

Lesson: if you want to invest in startups spread the risk by investing in a handful or in a small VC fund. This way if one of the companies goes bust you haven’t lost everything. Data shows first time funds out-perform established funds.

The short version of our journey is: with 2 other co-founders we set up Lucy three and a half years ago in Bali. We built really great technology, raised $1m (at a valuation of $10m) and launched a year later in Singapore. From there we grew mostly organically over the next 2 and a half years. As with all digital banks, profitability was far in the future, but that was ok, there were many VCs wanting to invest in fintechs like Lucy. And hey presto in summer 2022 we got a term sheet to raise a further $5m at a valuation of $20m. Happy days. Lots of momentum, and our metrics were good.

What did we do really well?

Our brand was really great. Lucy was named after my daughter Lucia and so I was really close to this company. The branding was some of Rebecca Yik’s best work and she is one of the best at this out there; we heard from everyone that they loved it. The UX of the app was really good. people used Lucy an average of 9 times per month, we had really great feedback. Tech was stable. Not many complaints.

Lesson: mission and brand and UX are really important they really did carry us a long way. And this is much harder than people think. It is how people feel when they interact with your company.

Our founding team was solid. Our CEO Debbie knew more than anyone about financially marginalised groups of women. She helped launch bKash (now Bangladesh’s biggest fintech company). Hal Bosher was the ex CEO of Yoma Bank that he built to Myanmar’s 4th largest. I founded and then exited Tigerspike, a tech company that among other things built digital banks.

I knew all about building digital banks. Hal knew how to run them. Debbie knew all about our customers. What could go wrong?

Lesson: Even with a rockstar founding team things can go wrong.

What went wrong

Summer 2022 was when values for fintech companies plunged. BNPL started it (which is good because BNPL sucks!, Profit without purpose. I wrote about it here), and then the digital bank valuations followed. Companies were worth a tenth or less of what they were before and the LPs behind our term sheet started to tighten their purse strings. So funding dried up. Our burn was over $50k a month. This isn’t that much relative to many but it is too much when you are running low on cash.

I was in this situation many times with Tigerspike and we did slash salaries, starting with mine. But I was 27–32. No kids, no mortgage, no wife. Our CEO Debbie wasn’t in a position to take a big pay cut, like me she is old (i.e over 45!). This is a penalty on older founders. When to start a company is a good question, there are arguments for when you are young and when you are old. I like a combination of youth and experience which I have at one of my investee companies Dungbeetle, whose CEO is 18.

I wasn’t rich at the time I took a pay cut at Tigerspike but I was young and could live off cereal and Sydney sun. It does help to have rich founders; after exiting Tigerspike I was able not only to invest in Lucy but to work for free which helped the company. This is unfair to non rich people but its life, you play the hand you have. The other side of this coin is if you are a founder with a previous exit you now have money. You are able to put more in which means you can lose more. My good friend warned me about this. I lost a lot of money when Lucy went bust. It hurt, but not as much as losing other people’s money.

So in our case, with an experienced older CEO pulling the ‘cut salaries!’ lever was much harder. Should we have found a younger cheaper CEO? maybe but then you lose experience. One of our competitors Jipay went bust 6 months before us and their CEO was in her early 20s. This is a tricky one but in hindsight we should have cut sooner and deeper. Things are easy with hindsight.

Lesson: read the landscape early and cut costs deep and soon. I am involved in many companies and without exception they don’t act quickly enough or cut deeply enough. I can not stress this enough.

We were over optimistic about at least some of our $5m term sheet investment coming in. I am over optimistic in general and this is not a good trait to have in some situations. We did get very close a number of times but just didn’t close it. In general we took too long getting to the decision maker and asking direct questions. Don’t waste too much time with non decision making people, this goes for VCs and for sales. Get to the person who can sign off the money and ask them directly: “what is the procedure for approving this investment? It is likely to happen and if so how quickly” Don’t be timid when it comes to asking this.

Lesson: Optimism. It can be great. You need it when you are in a startup but it can be your enemy too. Nothing is closed until money is in the bank. Be an optimist but have a pessimist nearby.

In the end it was the Singapore regulator who put the final bullet into Lucy. People were using Lucy to scam other people out of money. This was happening across the board in Singapore, but was happening to us more because Lucy accounts were the easiest and fastest to open.

“send money to this (Lucy) bank account or I shoot the puppy!”

People paid to save the non existent puppies, and then the scammer would transfer their ill gotten gains out of Singapore and withdraw the cash.

Whenever you have payments / bank accounts you have financial fraud.

Honestly it surprised me how much of it there is.

Lesson: in any regulated company you must be able to identify and react quickly to incidences that the regulator will shut you down for. This is hard. Because scammers are smart and won’t stop.

Those are the main lessons I took away from this. Other things I noticed / learnt are below.

Other learnings

• It helps not to rely on other people’s tech. We relied on Rapyd which we were generally happy with but we did have issues like all tech has issues. Tech goes wrong and if its not yours you can’t fix it. For example Rapyd stopped offering support for our debit cards so all of Lucy’s customers suddenly couldn’t use their cards. And there was no solution. We couldn’t fix it. Less lucky were people who built on top of Railsbank (we almost did), and Wirecard (my friend’s bank did and she had to close it). Both went bust.
• Debit cards is not a good business to be in. You lose money on all small transactions. This is not sustainable unless you are making more money elsewhere, which we were but only just. Most digital banks have this issue. Don’t go into retail digital banking. Go into SME banking. That was our original plan but again, the tech meant we went first into retail. This was a mistake. Tech constraints should not drive your roadmap (all very well to say but they do!). Retail is far less profitable then SME banking.
• Facebook and LinkedIn advertising were a waste of time. The Facebook algorithms always change so its just a dark art that you can’t monitor well, and LinkedIn is too expensive. Avoid them both if I were you. LinkedIn do yourself just by connecting and posting content, as you not the company. We grew as much or more when we were spending less.
• Tell staff (and investors too) what is happening often and early. We didn’t do this well enough, of course there is a tendency to not give bad news or not communicate if there is no new news. Treat staff like brothers and sisters not like children, they can handle bad news it means you are all in the trenches together.
• When someone says “we have raised a $50m fund to help women / some marginalised community!” it is probably not true. What they are really saying is “we are planning to raise a fund to help women and I am announcing it now so that everyone says “well done” and I look good. We talked to many funds that didn’t really exist yet, one of them even had our company on their LP deck. They don’t have money. Find this out early.
• Underbanked customers seem like a good segment. But they aren’t very profitable (which is why banks don’t want to bank them!). Many Singaporean VCs didn’t like that we tried to bank domestic helpers I assume for this reason (plus some other bias I suspect). This also sucks because our mission was to help the underserved populations. It can be done profitably but its hard. If anyone can do it Debbie could. But we didn’t.

Opportunity

Thanks for reading this far. I hope you are a bank or fintech company because here is an amazing opportunity: We built our tech for under US$2m, and that was an amazing achievement, credit to Hyunuk, Alvin and the team. At Tigerspike we did projects for 3 times that amount and produced something half as good as what we did at Lucy. If anyone wants a front end for a digital bank that has been proven to work contact me, there is a bargain to be had here.