🎧🍌 Lessons From Building Mercury with Immad Akhund (Co-founder and CEO, Mercury)

Lessons from selling a company for $45 million, how to get free legal advice, cultural mistakes most startups make, how to fundraise, and the embarrassing pitch that raised a Seed from a16z

👉 Stream on Apple, Spotify, and YouTube

Immad Akhund is the Co-founder and CEO of Mercury, a digitally native bank for startups. Mercury is supported by investors like a16z, Coatue, CRV, Chapter One, Ryan Peterson, Scott Belsky, Serena Williams, Terrence Rohan, Zach Coelius, Todd Goldberg, and more. Before Mercury, Immad was a part-time partner at YC, co-founded Heyzap which sold for $45 million, and co-founded and scaled Clickpass to millions of users.

Find Immad on Twitter and LinkedIn. Check out his new project, The Curiosity Project, where he interviews the founders of companies like Replit, Jasper, and Substack about AI, nuclear energy, and the future of media.

---

This episode is brought to you by Secureframe, the automated compliance platform built by compliance experts. Book a demo here.

To inquire on sponsorship opportunities on future episodes, click here.

---

In this episode, we discuss:

• The future of banking

• Why Mercury’s beautiful onboarding was an accident

• Cultural mistakes most startups make

• Two lessons from selling his first company for $45 million

• How to get free legal advice

• Why it’s better to build a network of entrepreneurs than investors

• How to fundraise

• Why its contrarian to start a non-AI company right now

• How Mercury is thinking about AI

• Why each operational team has its own engineering team

• The embarrassing pitch meeting that raised his Series A from a16z

• How crypto used up all its goodwill

• The hiring advice he gives every founder

👉 Find on Apple and Spotify

Follow The Peel on Twitter, YouTube, Instagram, and TikTok.

Thanks to Zac and Xavier at Supermix for the help with production and distribution!

---

Transcript

Find transcripts of all prior episodes here.

Turner Novak: Immad, thanks for joining.

Immad Akhund: Yeah! Thanks for having me Turner.

Turner Novak: So first question, I'm a Mercury customer – big fan. I think of it as my bank, but it's not technically a bank. Can you kind of explain what that means and how the product works?

Immad Akhund: Yeah. From a regulatory and licensing perspective, we are just software providers to our partner banks. For most of the banks you use, whether it's SVB or whatever, the actual mobile app and the website are made by other software providers. Banks don't, generally speaking, have engineering teams that build this stuff.

From a licensing perspective, it's fairly similar. We don't touch our customers’ money, like there's no point where we take custody over your funds. If you raise from LPs, they send money from their bank, and it goes straight to Choice Financial or Evolve Bank & Trust, which are our two partner banks.

And then when you make an investment, you basically instruct us to say, “Hey, I want to send this much money to this other bank account.” And we then work with the bank’s backend systems to send it.

I had this idea in 2013 and I started in 2017. It wasn't obvious whether you could deliver a great experience by just being software that sat in front of banks. A few people had done it, but I thought by this point – wherever we are in 2023, it's been six years – I thought we would need to get our own bank charter to really deliver a great experience.

But it turned out we can deliver a pretty good experience, working with these kinds of sponsor banks. And that whole ecosystem has really evolved in the last six years to enable that.

Turner Novak: Would you ever get a bank charter?

Immad Akhund: Every year we look into it – at some point it's more of an ROI calculation. It's actually not that expensive.

One of the reasons we don't do it – and I think you as an early-stage investor will appreciate this – if you were to think, what are the two characteristics of an early-stage startup or high growth startup? Number one, it's unprofitable. And number two, it grows fast, right? Those two characteristics are basically antithetical to what bank regulators would want of a bank.

A bank that's unprofitable and growing really fast is very scary. And actually, if you look at SVB’s failure, we could talk about why did they do this, etc., but a lot of it stems from the fact that they grew really fast.

Don't quote me on this but I think they grew from like $90 billion in deposits to $200 billion deposits in two years because 2021 was crazy. And when that happens, it creates this pressure and the CFO at the bank is like, “Hey, we grew deposits. What should we do with all this money?”

Obviously, they decided to put in 10-year bonds, which turned out to be a bad decision. But it is that high growth pressure that creates bad decisions at times.

Having said all of that, we have been profitable for the last four quarters, so at least that part of it, we've ticked. We still want to grow fast, so it's something we do explore.

I really think about it from a customer perspective, like, “Okay if Turner is using Mercury, would he have a better experience if Mercury had its own bank charter?” At least right now, I don't think that's the case.

Turner Novak: Yeah, it's a really good product. I can tell it's all you guys think about, the user. That's why I signed up, because I looked at a bunch of banks and I think Todd Goldberg, who I think is an investor, was like, “Hey, check out Mercury.” And I was like, “Oh man, this is great.” I got an account in less than a day or whatever the time was and logged in and it was super easy to use.

Immad Akhund: You know, one thing that's interesting which is maybe a good lesson for other people is, it was almost accidental that Mercury's onboarding experience was quite so good.

What happened is, we started this in like 2017 – August 2017 is when we incorporated. We were basically ready to launch in November 2018. But our partner bank at that time – I guess I can say who it was, it was BBVA – they had made these promises to us.

They were like, “Oh, we're going to have wire support. We're going to have immigrant founder support.” They'd make these promises and every month we were like, “Okay are we getting them?” And then a year later they had not delivered on them, so we ended up switching partner bank at the last hour.

So the designer and the front-end engineering team, which was basically me and another engineer – we had a bunch of time on our hands because we were like, “The backend engineering team has to do this whole integration from scratch.”

So we re-did all of the onboarding. We did all the bits that when you do the initial MVP, you're usually like, “Oh, let's skip that bit.” At that time, I was like, “It's a one-off experience, why make it that good?” But we ended up making really good and it creates this initial impression that really lasts.

I think if you talk to a lot of customers, everyone's always like, “The onboarding was just so seamless and great,” even though you've been using it for years. You could talk about some of the other stuff, but it creates this first impression that's lasting.

It has made me re-evaluate how important the onboarding experience is, and now I'm like, it's actually really important. We are about to make the split, so we're going to have two engineering teams that are focused on the onboarding experience.

Turner Novak: I was looking probably two or three months ago – another bank I was looking at, my email chain with them was 76 emails over the course of a couple months. And I don't think I ever even got an account made. That's the other side of the experience, so I've definitely always really appreciated it.

Why is it bad for a fintech company or for a bank to grow fast? What are some of the downsides of that, like why don't regulators like that?

Immad Akhund: The reason banking is regulated in the first place more than a software company would be, is that it's based on this fractional reserve model. For normal banks – and this isn't true for Mercury customers, which I can also talk about – like SVB or whatever, if you put $10 million in, they will take $8 million and they will lend it out.

Turner Novak: And that's because they need to make money in order to pay you your interest.

How banks make money through the fractional reserve banking system. Source: Bankroll Zen

Immad Akhund: Yeah, to some extent it's also because they want to make money. And that provides a valuable service, right? Like the $8 million goes to other businesses and consumers. It helps the economy. Lending exists for a reason, it does something similar to what VCs do, right? You invest now for a future return. You are helping build something and that's what kind of what lending does.

Most people couldn’t afford a house, but if you give a mortgage it lets them get the house today, get the value of the house, and they pay it off and the bank makes money and they get they get an output. That's kind of the banking model that's existed for a thousand years or something, right?

Turner Novak: But it's changed a lot – or at, at least the software around it has changed quite a bit.

Immad Akhund: And yeah, you could say the fundamentals of the internet changed it a lot as well. But anyway, so coming back to the original question, if that's the model, it's obvious how depositors would get screwed often if you did not regulate it because the bank's incentive is to go take risky loans or just not be that sophisticated about it. And they would eventually screw it up.

This used to happen all the time, like even in the seventies there were way more bank failures. And 200 years ago, this was just the normal, like 10% of banks failed. The regulations over time were like, okay, we need to protect consumers. And that's why regulations exist and that's why banks have to be regulated.

It’s like any sort of regulation, like the FAA or whatever would like look at a plane crash and go, “Okay, the plane crashed. What were the reasons that that plane crashed?” And that's happened over a hundred years with banks.

If you look at banks and they fail, like what are the reasons they fail? One of the reasons is they grow fast and they do bad things, like they take riskier loans. Failure is just more likely to occur in this situation where you grow fast. If you grow slow or you've been doing the same thing for 30 years – there’s is a term for it –if you've done the same thing for a long time, you're more likely to continue to do it successfully.

Anyway, if you 2x your deposits, then you will end up doing 2x the loans normally. Most systems have a hard time absorbing 2x loans. The way Mercury works, which is relevant and interesting – Mercury is a software company and can't lend in those deposits.

The two banks we work with are much smaller than the amount of deposits we send to them. Evolve is about $2 billion and I think Choice is around $3 billion, and we have tens of billions of deposits. So those banks can't actually host those deposits on their balance sheet.

The way they function is they have these sweep networks. They work with a bunch of much bigger banks that are in the sweep networks. I was told by at least one of these banks – not to name them, but you can actually look them up online, which banks are in the sweep networks – there are trillions of deposits in the sweep network banks.

The money is not held by us, even by our partner banks. It's not held by them. It gets swept into a bunch of these banks and those banks as a whole can lend against the money. That’s partly why they want it, but because it's distributed, you end up getting more FDIC insurance.

We guarantee that if you have $5 million deposits, it'll get swept to at least 20 banks that give you $5 million in FDIC insurance, so that's a very different model to a normal bank model.

Turner Novak: Yeah, you're almost like lead gen or deposit acquisition, sort of a sexy software layer.

Immad Akhund: Yeah. There's definitely a piece of that. And there are payment volumes. Banks actually are like a weird combination of like three things. At its core, it's a deposit function, a lending function, and a payment function. And there's no actual strict reason why the same company needs to do all three of them.

It made sense when like you had to walk into a local branch and there was only one bank in your town, but in the modern internet parlance, you would expect that to get like unbundled, and to some extent we do that.

Turner Novak: Yeah, and I think we're going to stop talking about the banking system. If anyone is interested in walking through all this, he did a podcast with Logan Bartlett. If you just search Logan Bartlett, you can find it. We'll throw it in the show notes, it's like two-hour podcasts. He spends the majority of it really digging into all this stuff, so definitely recommend checking it out if you want to hear 10x more going way deeper on this stuff.

How do you think the industry is going to change over the next 10 years? Because that's definitely relevant to Mercury and what you're doing. What are you expecting as you think about positioning Mercury? What can we expect on the product side? Anything you can share?