All episodes

57. Immad Akhund - The Art of Irrational Effort

·Founder
Read transcript ↓Browse chapters ↓
Listen to the conversation
Watch the video

Related conversations

Description

Immad Akhund (X, LinkedIn) is the co-founder and CEO of Mercury. He is also a prolific angel investor and co-hosts the Founders in Arms podcast.

Mercury is a fintech banking platform used by one in three startups. It is Immad's fourth company in 20 years of entrepreneurship. Mercury is in the process of becoming an FDIC-insured bank and recently launched Mercury Books, AI-powered accounting software.

Immad grew up in Pakistan, moved to London at nine, and left for San Francisco at 23 to become an entrepreneur "for real." Mercury is built on something rare in finance: a willingness to spend a disproportionate amount of energy on details, the kind of care Immad admits "seems irrational.”

We start with a simple question (what is a bank?) and discuss why banking stayed bad for so long, what stays hard as AI makes software easy to build, and why Mercury's motto is "be extremely helpful." Entrepreneurship has shaped Immad's life, and he hopes Mercury can empower a generation of entrepreneurs.

I hope you come away a little more annoyed by the mediocre things we've all learned to tolerate and inspired to put an irrational amount of care into making them better.

Dialectic is presented by Notion. Notion is an AI-powered connected workspace where teams think together and create their best work. It works with all of your agents and keeps context across them and your team. Check out the new Notion Skills API. You can learn more at notion.com/dialectic.

Timestamps

Links & References

Show 8 more referencesShow fewer references

Transcript

4:16 Start: Banks and Why Immad Was Suited to Build One

Jackson: (4:16) Immad Akhund, thank you for joining me. Thank you for having me, I should say, in your space.

Immad: (4:20) I'm excited for this chat. How are you doing?

Jackson: (4:22) I'm doing great. It's nice to be in San Francisco. It's been a little while. My first question is: what is a bank?

Immad: (4:29) That's a good question. When I first started thinking about really doing Mercury, I had a lot of these kinds of questions. It's actually an amalgam. It's a bundle of features. Number one, it stores your money. Number two, there's a set of payments, like ways of sending the money in and out or using a card to spend it, et cetera. And number three, it's lending. I would say those three constitute what normally people think of as a bank. Lending is like getting money. It's not obvious that those three should be completely bundled. You can imagine storing something but not transacting on it that much. There's lots of places and people that do a lot of payments without doing a lot of storage, and then there are separate businesses that do lending. I think the storage and sending money kind of makes sense together. Lending you could consider separate. But historically we've had fractional banking, so you have money stored somewhere and you can also lend it out. That's how banks make money, but it also kind of facilitates society.

Jackson: (5:33) Right.

Immad: (5:34) Lending is the fabric of how people can buy houses and how businesses can grow. I think relatively early when I started Mercury, I read this book called The House of Morgan, which is all about J.P. Morgan and Morgan Stanley—it all comes from him. For a long time there's this whole robber baron era of America. All of that was the banks backing all of these things. They were not just passive kind of things sitting around waiting for someone to ask for a loan. They were like, "How do we construct new companies and bankroll them?" It's interesting to really think about what is a historic bank, what's a current bank, and what does it mean to be a bank. It's a fun question. I really like it. It goes back to an even more complex question: what is money? Money by itself doesn't have much value, but money is a promise to get something, either some labor or some capital from some person. So it's this really interesting concept. I don't know if I'm going somewhere where you were kind of going with this question, but—

Jackson: (6:45) No, I think this is exactly why I wanted to start here.

Immad: (6:50) Yeah. So to some extent, you could think of banking as the facilitation of money. It used to be kind of painful. There was no fixed value of money. Then we had gold, but you could cheat. Then you had governments that tried to impose it. There's something kind of pure about where money is today, where it's like, does it even exist? It's these digital bits going in and out. At the end of the day, if you think about money as not money, but as how well can you facilitate commerce and how quickly can you transact, digital money is actually pretty good. I know some people are kind of down on it, but if I'm running a business, I want to get a bank account really quickly and I want to start transacting quickly. These things have nothing to do with me running my business. I want to get on with the thing that I'm building, and that's the thing that has value to the world. The money aspect should almost be in the background. And it kind of, I guess, ties all the way back to why I made Mercury, because in America, it's just surprisingly difficult to do basic things in banking. Even now with Mercury, you can go online and get a bank account in 10 minutes, but at most of the major banks, you have to walk down to a bank branch. When we started Mercury, we signed up for every bank — nine or 10 bank accounts — just to see what the process was like.

Jackson: (8:18) It sounds miserable.

Immad: (8:19) Yeah. You'd go there, everyone would ask for slightly different information, and it would take three hours. I remember—I guess I won't say which bank it was—but we went there, did this whole thing, and it took three hours. We were like, "Okay, finally we got a bank account." We went home and it was like, "Oh, wires don't work on this bank account. Sorry." We had to go back to the bank branch and go, "Okay, can we get wires enabled on this thing?" It was completely ridiculous. But at the end of the day, none of that really matters. What really matters is people want to use Mercury because they're trying to build a business, and we just want to enable that and get out of the way.

Jackson: (8:57) I think it's interesting. Part of the implication of where you were going is that money is at its best—especially when the tools around money are excellent. It sort of gets to be a little bit invisible, and it's only there when you need it. I'm curious. Obviously, you've been an entrepreneur for, I think, 20 years now, maybe even longer.

Immad: (9:14) Twenty.

Jackson: (9:15) This is 20? Yeah.

Immad: (9:15) 2006.

Jackson: (9:16) Amazing. Amazing. Why was banking "you-shaped," and when did that become obvious? Had you always been interested in the J.P. Morgans of the world and things like that? Clearly there's an entrepreneurial thread, but I don't know that it's super obvious on the surface that building a bank was the most "you" thing. I'm curious how you related to that.

Immad: (9:40) This is my fourth company. I've been doing startups since 2006. I would say the bit that is very me-shaped is that I love the entrepreneurial journey. I started my first company in 2006, and I really didn't know what I was doing, but I knew I really hated my previous job. It was so depressing going into the office and feeling like I didn't have any agency. Then I started this company. It was basically like a Yelp for London. We didn't raise any money and it failed within seven months — honestly, it failed before it started. But I just loved waking up in the morning and going, "I'm building something." I could make all the decisions to do this, and I was just like, "If I can just do this for the rest of my life, that'll be amazing."

Jackson: (10:30) What about that specifically? Just the lack of the freedom?

Immad: (10:33) I think it's really cool to make things and wake up in the morning thinking, "I will make something." There's a very direct tie when you're an entrepreneur between making something and making it for people. You have the feedback loop, you have an idea, and you can just go do it. There's some level of irrationality where I just don't like people telling me what to do. I don't know why, really. I remember my—

Jackson: (10:54) You said something about your dad, right?

Immad: (10:55) Yeah, yeah. My dad, when I was a kid, he had a garage in Pakistan. He ran it, and he still actually has it. Whenever he'd come to London, he'd have to take some menial job because we'd moved from Pakistan to London, and he would hate it. He would never keep these jobs. At some point, I was like, "Why don't you keep these jobs?" I was probably a teenager being annoying. He was just like, "I can't work for people." It really stuck at the back of my mind: "Oh, yeah, maybe I can't work for people either." Then when I had to work for people, I was like, "Yeah, he's right. I can't work for people." It was just a random thing that stuck in my head. So I really like being an entrepreneur. As soon as I did it, I was like, "Okay, I need to move to San Francisco if I'm going to be an entrepreneur for real," because in London, I felt like I did all this networking and I already knew all the young entrepreneurs.

Jackson: (11:50) Was it obvious that it was going to be tech entrepreneurship?

Immad: (11:57) Something I realized quite quickly in London in 2006 and 2007, back in the early Web 2.0 days, was that being an engineer and being entrepreneurial was a relatively unique skill set, especially in London. It's kind of a given now, but this was 2006 where there were a lot of people wanting to build apps and wanting to do things, but they couldn't do it themselves.

Jackson: (12:26) The non-technical 20-year-old today can't imagine.

Immad: (12:29) Exactly. Well, now you can just vibe code it or something. I realized that this was early Web 2.0, but things were starting to happen, and it was obvious to me this was a big place to do business. I was uniquely suitable as an entrepreneurial engineer since I did computer science. So I quickly decided tech was the place to do it and San Francisco was the place to do it. To come back to your overall question about why this is me-shaped: as I went through this journey, I built developer tools. My previous two companies eventually became mostly developer tools companies. To some extent, developer tools companies are also serving entrepreneurs. Over time, I realized I really like serving entrepreneurs and the entrepreneurial journey. I had this mental map that started probably in 2011 where I looked at all the tools I used as an entrepreneur, and they all sucked. In 2006, everything was bad. There was no Stripe or Gusto or Slack or—I don't even know, was there AWS? I don't think there was AWS. I remember we used to do bare metal. Everything was just so painful. You could see everything was getting better as an entrepreneur, and it was very distinctive to me because I was like, "Okay, I want to build a business," and I'd always look at these ideas. Banking was in the back of my head even in 2009 or something. I was like, "Banking kind of sucks." All these things suck, but banking also sucks. It wasn't that I thought I would eventually build Mercury; it was just a painful part of being an entrepreneur. It wasn't until 2013 where I thought, "Actually, maybe I could do this." It actually came from a Y Combinator company — I always forget the name — [?Starling]? I should remember. But it was a debit card for seniors. It was just a couple of young kids that had a sponsor bank deal with Sutton Bank. I talked to them and realized, "Oh, you can do that?" Before that point, banking was this foreign concept, and I had no idea how you'd ever do it. But they made it real for me because it was literally two people, and they went and did the sponsor bank deal. At that point, I thought maybe I could do it, but I was busy with my startup. It wasn't until 2017, when I had my list of ideas and this was one of them, that I really dug into it. Even then, I thought, "Probably not for me," but no one else had done it.

Jackson: (15:21) Well, I was gonna say a bunch of those things you mentioned about the suite of tools for a startup between 2009 and 2017—a lot of them had actually gotten way better.

Immad: (15:29) Yeah, exactly.

Jackson: (15:30) And then banking's just banking—exactly the same.

Immad: (15:33) And I wish I could remember exactly who said this, but it's really borne out to be true: the hard part of fintech is making something that people want. It's not that different from another product. I think people get hung up a little bit with fintech and probably other things where you're like, "Oh, I have to solve this compliance thing or some Rails thing," or it's the equivalent of when you invent a technology and you think that's the hard part. The hard part is always building something that people actually want and care about. In that sense, I was great for it because I knew exactly what to build. I was very stubborn about the spec. So one of the tricky things was—I think Mercury was probably the first, I’m sure someone else can say they were the first—but we were the first popular thing that had domestic wires in a neobank in the US. Now it's not that big a deal, but in 2017 when we started Mercury, no one had done it. It was pretty much...

Jackson: (16:30) Why?

Immad: (16:30) So everything started with cards. The whole sponsor bank ecosystem exists to some extent because Target wants a card or some other thing wants a card.

Jackson: (16:40) Got it, got it, got it.

Immad: (16:41) And so it started with card, and then the next thing you do is, "Oh, I need to fund this card." So then they do ACH, and that's kind of most of the sponsor bank ecosystem: card and ACH and normally a prepaid card or something like that. But obviously, if you want to serve businesses and any real business, you need domestic wires. That’s a core part of it. And I really wanted international wires as well. But I was like, "I can't launch this thing without domestic wires." So we ended up actually doing two partner bank deals because we did one and they promised us domestic wires. Every month I was like, "When's it coming? When's it coming?" Six or nine months later I was like, "I need to run a startup. I can't wait for you forever." So then we did another one and eventually, even with them, we were probably the first people to ever do it because I remember it only vaguely worked, but eventually did very smoothly. So having a really clear thing that was like, "Here's the set of things we need," and there's a lot of these kinds of things that made Mercury instantly successful when we launched, but it took a year and a half to get to launch. And yeah, going web-first, that was again I think somewhat obvious. Most entrepreneurs are sitting in front of a laptop to run their business and they need to—

Jackson: (17:54) When you say web-first, as opposed to an app or mobile, you meant instead of in person?

Immad: (17:58) No, no, I meant instead of an app. There were a few SMB neobanks that had launched that were mobile-first because everything was like, "Oh, we need to be mobile-first," and that's the differentiator. I was like, "Yes, you need a mobile app," but the web app had to be pretty good. And it needed to be quite complicated. We had multi-user functionality and all this kind of stuff from day zero. Because as an entrepreneur I knew exactly what we needed, it really resonated when we launched because it wasn't a half-assed attempt at it. It was like, "Hey, this is the solution that every entrepreneur needs."

Jackson: (18:30) Totally.

Immad: (18:30) Yeah.

Jackson: (18:31) The meta theme is actually a deep care and intuition and understanding of entrepreneurship and what entrepreneurs need, more so than the bank. But the bank was a really great container for that in some sense.

Immad: (18:43) 100%. Actually, we just launched Mercury Books, which is kind of an extension on that, but it's a similar idea. I know as an entrepreneur that accounting systems suck. Maybe they work for bookkeepers to some extent—I think they also don't like it—but as an entrepreneur, it's completely useless. You get these spreadsheets and you don't know what's happening with your money. I think there's an underlying element to this: for so many years of my entrepreneurship, I was struggling for money. First startup, we never raised any money. Second startup, we raised like 300k and then we spent six months trying to raise more money. And the third startup, we went for eight years and we mostly survived on 10 million raised or something, and we eventually became profitable there. Because I struggled so much, I know money is such a source of stress for entrepreneurs. So all these things tie up to that, which is: how do you remove the stress for an entrepreneur? Yeah, I think accounting is a perfect thing, which is such a stress. You're always like, "What's happening with my money? Where am I spending it? Am I getting paid in time?" All these kind of questions that are kind of hard to answer as an entrepreneur. It's easy when your business is very simple, but as soon as you have 10 employees and a few hundred customers and all this stuff, it gets to be a complete mess. And it gets easier eventually because eventually you have a finance team and they're answering these questions.

Jackson: (20:12) They have to suffer using the bad tool.

Immad: (20:14) So actually, for me now, it's not that bad because I have a finance team. I'm like, "Oh yeah, get this invoice paid," or "What's happening with this?" Someone will go dig up the answer. But we were like 200 people before we had a good CFO. So it takes a while to get to that level.

20:30 What Traditional Banks Do Well & What Mercury Got Right

Jackson: (20:30) I want to come back to both accounting and to maybe that last theme, which I think ties into AI and products becoming services and so on. But before we leave banking, a couple of questions. Given the initial setup starting in 2017 with Mercury, what—maybe this is a dumb question, too—but what, if anything, are traditional banks, or maybe you could even say archaic banks, really good at? Everyone talks about all the things they suck at, and clearly they don't really build product in the way that software companies think about building product. But what have you learned in the years since about the parts of traditional banking that are actually really hard?

Immad: (21:10) Yeah, I think there's this funny thing: "If you don't know what the product is, then you're the product." But it's worth understanding: what is the product that the bank really makes money on? And it's mostly lending. Banks exist and make money on lending—that's like 80%, 90% of the revenue most banks make. And it's kind of like a mid-market enterprise sale. If you put it in a software startup context, it's like, "Hey, I give you $10 million to buy this commercial real estate." If you're making 5% or so net on that, it's millions of dollars. That's like a 500k ACV, but it's kind of guaranteed over 10 years. I mean, that's a serious mid-market/enterprise sales deal. And like any organization, as a bank you become catered towards the thing where you make money.

Jackson: (22:11) Yes.

Immad: (22:12) So these are really mid-market enterprise sales companies selling loans as a main product. So if you put it in that context, then I think the rest of why banks kind of make a lot more sense.

Jackson: (22:21) Yes.

Immad: (22:22) Because the deposit product is—I mean, they need deposits in order to serve those loans, but it's never been a place where they're trying to differentiate or be competitive. It's like the minimum possible.

Jackson: (22:32) An alignment issue in some way.

Immad: (22:34) Yeah, it's an alignment issue. And it's like: who do they care about and what do they care about? So what they're good at is—and we don't do it at all really—if you want to get a $10 million commercial deal, you go to a few banks, they'll give you really nice service, and they'll take you to some golf game or whatever it is. But they have some great salespeople that really focus on doing those big deals, and there's a risk-aversion that's kind of bolted into the system that's tended towards, "Let's do the safer, bigger deals where we can underwrite them easier." But I think they're good at that. Hopefully we'll get good at it too since we applied for a bank charter and we'll probably be doing bigger loans over time, but definitely something they're good at. And fintechs have actually not really made any headway in that space. There's a little bit of fintech lending that happens at the bottom end of things. But if you think about it, what's 80% of revenue in all financial services and fintechs haven't really made any headway into it is an interesting kind of thing.

Jackson: (23:39) It's a DNA thing. Yeah. Maybe on the flip side, what did you guys compromise on? You maybe gave the inverse of this question, which was like, "We really wanted domestic wires as an example." You built the product for a year and a half as I understand, maybe even longer, before launching. But what did you willingly say, "We're not going to have this, we're not going to have that, so that we can do this set of things really right"? Because I imagine in some sense this business has been in large part like a sequencing problem of: what is the core thing? And then gradually, and obviously now we're adding books and accounting and now we're going to become a real bank and so on. But especially early on, I'm curious what, if anything, you remember that you were like, "We're just going to—we're not going to have it, it's going to be fine."

Immad: (24:24) Yeah, that's a good question. I think when we first launched we didn't have checks. I really thought that we wouldn't need checks.

Jackson: (24:33) Checks super recently, right?

Immad: (24:35) Well, so we just added the checkbooks recently, like even check deposits and check sending. Some purest part of me was like, "Checks are stupid, we should get rid of them," but a lot of America still runs on checks.

Jackson: (24:50) Yeah, very San Francisco mentality.

Immad: (24:52) It's literally one of these—so we launch checkbooks on business but not on personal yet, and it's like one of the biggest requested features on personal now. There's little things like that. There's definitely an infinite scope of attachments. The thing about banking—and this is actually kind of the final thing that got me going like, "Oh, I need to build Mercury"—is that banking is like this initial thing that everything springs from. You need your bank account. Every single business has a bank account. Actually, there's not that many products every single business has. It has all your money, it has your finance team. There's so many pieces. So it lends itself to go, "Here's all the other financial operations and here's all your vendors." It lends itself to a lot of interconnectivity and a lot of features. But we were really focused, especially at launch, going, "What is the minimum we need?" The minimum was big because we needed cards, we needed wires, we needed multi-user, but it was really the minimum. And we tried to make that minimum really polished and nice. But we didn't launch saying we do this thing that no other bank does. It was much more like we did the thing that you might expect your bank to do, but we did it much nicer. Which is actually an unusual approach for startups. Most startups are very focused on: what's the new thing?

Jackson: (26:15) Totally. It's a really interesting word.

Immad: (26:16) I mean, to some extent it was new and fancy, and there was online-focused banking which still was relatively new then, but it wasn't trying to do some new and wacky thing. I really thought—you know, I've launched a lot of products over time—really thought we'd launch this thing and no one would care.

Jackson: (26:34) Really?

Immad: (26:35) Yeah, that's a weird thing.

Jackson: (26:37) Why did you take so long?

Immad: (26:40) Because I thought that's when the work starts. We literally launched the first possible moment we could have launched. It just took a year and a half to get to that minimum spec. That's what I was saying. The wires barely worked. We literally were like, "This was the last thing we needed to work," and we tested it and we're like, "I think it's working, let's launch." So our perspective was really like, you launch and that's when the work starts. That's when you have to collect user feedback. You can't even really collect user feedback until you launch. To collect user feedback, that's when you set up the distribution engine and all this stuff. So when we launched, we thought no one would care. But it was literally we launched and then we just grew from there.

Jackson: (27:25) Was it a show-don't-tell thing? Like maybe to go back to that point of not being a, quote-unquote, novel hook feature. Was it just like—and maybe did you guys have the demo.mercury at launch?

Immad: (27:37) No.

Jackson: (27:37) Okay. So it was more just like a few people tried it and the word of mouth was just that delightful.

Immad: (27:42) Yeah, I think there's a—so actually I had this slightly disheartening process six months before launch. I was like, "Okay, let me go talk to 100 companies." And literally 100, I have the list somewhere probably. And I was like, "Okay, let me see. What do they want from this? Would they be excited to try it out?" And I was trying to set up customers for our eventual launch. So I talked to 100 companies and I remember two of them were really like, "Yeah, let's go." And I was like, "Oh my God, we're working on this thing for a year and a half." And that's partly why maybe I was like no one will care. But I think the interesting thing I learned there is, actually, if 2% of people are like, "Fuck yeah," that's actually a lot of people if the market's big enough. So I think that's what really happened. When you first start, you just don't need that many people to go, "Oh my God, it's a snowball." But it has to be the 2% that really want it, have to really, really want it. And then you kind of work your way up to the people that are a little bit less interested. But that only works if the market's big enough. So I think what we really experienced is in that first six months or something, it was the 2% of people that were so sick of their banking and they were like, "Okay, I'm just going to try this. It's got to be better than what exists." And I think another part that's hard for everyone to achieve is the product was audacious enough. There weren't that many neobanks back then, so going out there and saying, "Hey, we are building a business neobank," was an interesting enough message that spread a lot. So you always have to have—I think when someone does something audacious, like building humanoid robots or whatever it is today, the audacious messages tend to be more interesting than the non-audacious ones. So the fact that it was a big and audacious message helped it spread to find that 2%.

29:33 Why Mercury Wants a Bank Charter

Jackson: (29:33) Speaking of audacious, in some dimension you are now becoming a "real bank," as they say. I suppose I don't even know—candidly, I'm not totally even sure what that means. I'd love to talk a little bit about it, maybe to start. What's going on with this whole sponsor bank structure? What does that actually mean and what are the implications of that? And maybe as a follow-up, why even become a bank at all—an official bank—in America?

Immad: (30:01) This is actually not true everywhere, but in America, in order to issue a card on Visa or Mastercard, you have to have a bank issue the card. That's one thing. And then number two, if you're going to hold deposits, they need to be in an FDIC-insured bank. And I guess number three, wires and ACH, again, are also done by a bank. So you need a sponsor if you want to do any of those three activities. You need a sponsor bank or you need to be a bank. When you're a tiny fintech—like we launched with eight people or whatever—it's just not realistic in America to be a full bank. The expectation that the bank regulators have is very high. When you're small, I think the sponsor bank ecosystem has emerged to help support that.

Jackson: (30:45) And is that a relatively new phenomenon, with fintechs in the last 15 years?

Immad: (30:49) I think, again, it goes back to these shop cards, like the store cards. So I think some level of this phenomena existed since 1980 or something—I don't know, whenever the first store card came out, maybe even earlier. But a long time. There was definitely a period between 2017 and 2021ish where there was this massive surge of everyone—well, not everyone. When I first started Mercury, I was like, "Let me list out every single sponsor bank that's worked with anyone." I had a whole spreadsheet because that was one of the hard things about Mercury. There were maybe 20 or 30 on that spreadsheet. It wasn't that many. I think probably a lot more came online and did a lot more partnerships, and it probably peaked at 120 and that's actually somewhat declined since then. But that's the sponsor bank ecosystem. Other countries like the UK and Brazil, and even Nigeria and I think the EU, have lightweight charters where maybe you can't do lending, but you can get access to depository or whatever restrictions they have where you can be a smaller entity and get access to the charter. But the US doesn't have that method. So when I first started Mercury, I was like, "Okay, maybe let's go get a charter," or you can also buy a bank—I guess that's a third route. And it just seemed too hard. You probably need a $30 million minimum to start off with.

Jackson: (32:14) Of deposits or of capital?

Immad: (32:16) Sorry, of money. You probably need to buy the bank, and then you need some capital. So it wasn't accessible to me to go buy $30 million. It wasn't even obvious that—to go back to the previous thing—you need to build a great product that people love. And it's not obvious that you have to do all this fixed cost work of getting a bank charter before you can even figure out whether the product is anything people want.

Jackson: (32:38) Totally.

Immad: (32:39) So we went the bank sponsor route, and it's only more recently—a year and a half ago—where we really were like, "You know, let's go kick off this bank charter process."

Jackson: (32:50) Why so long? I mean, to go back to your earlier point, I don't know if you were instantly successful, but you certainly had a lot of traction. I don't know when the first billion-dollar valuation was, but it was not super recently.

Immad: (33:01) 2021, yeah.

Jackson: (33:02) Why was this the right time? Why not five years ago or five years in the future?

Immad: (33:05) There's a few elements. I mean, honestly, the sponsor bank stuff worked better than expected. The bank sponsors—there were more of them, they improved over time, and I know when we initially launched I was like, "You know, we won't be able to do X or Y," and there's still some restrictions on things we can't do. We don't have Zelle, for example. But A, we grew big enough that we could get our sponsor banks to do things. I mean, checkbooks—maybe some other sponsor fintechs have them, but it's very few, and it's a relatively new feature for getting people to go build that for us. But I mean, it's leverage. And also, the sponsor banks have better technology and better... in 2017 when I was looking at it, it was barely ACH and debit card. So the fact that we managed to get to a level where we can support a broader set of features has been useful. And we grew really fast. So it was always like, "Hey, are we going to jump this to our top of priority?" and it was always like, if the features are good enough and we're still growing, let's just focus on growing and adding features and supporting customers.

Jackson: (34:20) What changes now? What does this actually unlock?

Immad: (34:23) I think there's three things. Number one, we're just at the scale where the sponsor bank ecosystem is not really made for the scale that Mercury operates—just the amount of deposits we have and customers, and it's kind of tricky. I think it would be much better for us to work directly with regulators. And I would say regulators have been really open to working with us. But when you're really small, you're kind of relying on the sponsor banks to have a compliance team and a risk team that really guides you. At this stage, our compliance and risk teams are bigger than our sponsor banks' compliance and risk teams, and the customers come to us first. So we really have to be the first line that really holds the bar—the regulatory compliance bar. So that's one really big thing for us. Secondly, we would like to do more stuff for our customers. We talked about Zelle, we talked about lending, and these things are just very hard to do as a fintech. So I think we'll deliver a much better product to customers. And then thirdly, it will end up making us more money. There's a middleman, essentially.

Jackson: (35:27) Primarily because of lending or just a broad suite of—

Immad: (35:30) No, just like for every single thing we do right now, we pay our sponsor banks, and if we did that in-house we would have more money. I mean, there's more costs associated to it as well, but especially as you scale, the cost kind of gets cheaper over time. That's not the primary motivator. That's like third on the list, but it's still a factor.

Jackson: (35:51) Yes. And it sounds like, really more than anything, it's just a product of an extreme amount of scale.

35:57 What's Scarce When Software Is Easy

Jackson: (35:57) During the Books launch where there was a line I really liked from Sonya Huang at Sequoia. She said as the cost to build software asymptotes to zero, we should expect to see more bundling and consolidation of previously independent categories of software around the things that are actually scarce post-AGI, like a bank charter. I'm curious how you've thought about—and as I mentioned earlier, I want to get more into the weeds on the product side—but how you've thought about the fact that in some ways you do things in a very Silicon Valley startup-y way, which is moving really fast and being super nimble. Obviously, you guys have done a ton with AI tools and CLIs and things. Then there's another thing that seems true to me about Mercury beyond just the bank charter, which is just this kind of slow, deep compounding that seems very clear. You have a long-term vision and you are focused and steady. If you wanted to, I'm sure you could be launching new products every two days. The frenetic nature when you land in San Francisco, the mania of everything that feels possible in software now, is just so different to 2017, let alone when you started your career. I'm curious how you're thinking about the parts of Mercury that are the durable core and then the parts that are like, "We want to be really experimental, we want to give agents credit cards," which you guys have done. You're clearly not throttled, but you are sensible.

Immad: (37:20) Sure. One thing I really care about—actually, this is something else I thought about when I looked at my previous bank accounts—is you click around on the interface and everything looks completely different. I don't know if you've noticed that about your bank. You click on international wires, and it seems like you're on a different website. Things don't work between these systems, and it's always been this kind of frustration for me. But I think it speaks to the fact that it's very easy maybe to build a lot of things. Greenfield kind of stuff is easy. But things that work really well together and are really thoughtful—unfortunately, AI is not that thoughtful. If you really want to build something great that works really well within an ecosystem, and most of what we do at Mercury, we do the next thing because it works really well with the existing things, if that makes sense. You want to have a user permission system that works across all of your different products, and you want the UI to work seamlessly, and you want the data to flow between them all, and all these kind of pieces. I think it's much easier—actually, a lot of things that Amazon does, they launch a new domain and it's a true two-pizza team and they don't have to talk to anyone else. That's not that hard. Maybe we could do more of that, but the real value, every time we think about something new, I'm like, "Why? Why would this product work better at Mercury than it would work anywhere else?" It has to be a combination of we have the money, we have the data, we have the users—kind of a trifecta. We want to make it so if you use this product within Mercury, it's just a no-brainer because it works so well with everything else and it gets better the more products we have. There's just a more thoughtful— It just takes more time to build all of that stuff and make it really cohesive and high value. Another point that I wanted to make is software. If you think about when I started in 2006, software became easier and easier to build over time anyway because these tools got better. It got easier to deploy. We got Rails and then Node. It became easier to build software. To some extent, AI is a continuation of that, where the cost of actually deploying something small gets smaller, but the hard bits are still hard. It's still hard to build trust and brand and network effects and moats and the bank charter stuff. It's hard. Actually, that's where more and more of the value goes. Maybe when Mercury started it was 50% software. I don't know what the percentage was, but maybe 50/50. And now, if the building of the software gets 10x easier, then the other part becomes 95% of the hard part. And actually, that's where most of the value goes. The software part is easier, but caring about great product design is even more valuable because everyone's building a lot of software. I see a lot of companies where they just throw a lot of stuff out there. Over time, maybe the first product everyone was like, "Wow, this is good product design." But when you throw five things out there and they're all kind of hodgepodge and you're half-supporting some of this stuff, over time—I think about this a lot, actually. Why do companies get bad over time? It seems like an inevitable march. Very few companies have stopped that march of degradation. I think this is one of those things. You throw a lot of products out there and you don't really maintain them over time. They don't work well together and decay. There's a bunch of other reasons. Obviously, culture degrades over time and people move away from customer value to customer extraction or whatever. But it's really important for me to think about how I build a company that gets better over time, ideally not worse over time. I think Apple's really good at this. Whatever you launch, you should have a high bar for it. I guess this is a low bar to say you shouldn't be ashamed of it, but you should feel like you're giving delight with everything you launch. We try to hold that really high bar to it.

Jackson: (41:36) Yes, yes. I was thinking, as you were speaking, about—Stewart Brand has this, he's on part one of two into this book about maintenance. It's a similar thread around how it takes a great deal of intention and effort to resist decay. Decay is actually the default.

Immad: (41:53) Yeah, that's the default. It's like entropy is sucking it out of you all the time and you got to fight it.

Jackson: (41:59) Yes. And it could be so. It's funny too, just in this tension, because you've also talked about how banking has infinite TAM and there's infinite ideas. It's very clear you're not someone who is short on possible things to do, but you have found the right sort of balance to say, "Actually, cohesion is what matters above all." The cohesion of this product working really elegantly is way more important than any exciting new feature.

Immad: (42:25) Yeah, I mean, there's also only so much new stuff you can invest in. We go through phases where I'm like, "Oh, we should do these five things." Then the next year I'm like, "Okay, we've got these five things underway; everyone will get mad at me if I get a sixth thing in there." We launched the books thing; we worked on that for about two years. At any one point, we have two or three things that are cooking. I have kind of an investment model in my head where I'm like, "Okay, this is books before launch." It's like a pre-product-market-fit startup. How much do we invest in that based on the idea? As these things do well, we invest more in them, but there's only so many new things you can have going at a time. There's obviously smaller features; some team will go build something small. The agent card is like our normal cards team kind of built that as a smallish feature. But for the big product stuff that's much more meaty, there's only so many of those you can attempt at a time.

Jackson: (43:30) You have this line on this idea of AI making things easy and certain things being scarce around the work that requires pain. Can you give a few more examples of that? Obviously, there is a lot of the initial work for the company. I would imagine this bank charter process has a good amount of pain there. I don't know if "friction" or "pain" is the right word, but what are other examples of things that are maybe unexpectedly bumpy or frictionful?

Immad: (44:00) Maybe "hard" is the right word rather than "pain." I think talent is always hard. It's a lot less obvious to point to, but how do you have a strong culture, hire people against that culture, hold a high talent bar, make sure people have the right level of autonomy, and then do good performance management? What's the full life cycle of that? Honestly, if you ask any entrepreneur what's so hard about their job, it often comes down to trying to hire someone or having some issue with someone. That's actually quite a hard part of the job, but it compounds. If you can get some great person, everything feeds downhill from the people. Great people who are really thoughtful end up making great products and great customer experiences. So I think that's a hard one. I don't know if we're amazing at it, but we've tried to get better at it.

Jackson: (44:59) I think you mentioned books was a two-year process. Why does that take two years and not two months or longer? I don't really have a great sense of that.

Immad: (45:08) We often talk—especially as media people, entrepreneurs, and people on social media—as if this stuff has existed forever. But AI for product development really only got amazing in December last year. I think there's a chance books would have launched a year ago if we were already there when we first started. It comes back to what we talk about with Mercury, and it's almost worse now. The bar of something we launch has to be pretty good for a large number of people. Some things are just complicated. It's not something you can do quickly. It's like banking. Banking is also a hard product. Even if you were launching it from scratch now, it would be hard.

Jackson: (45:53) And it's hard primarily because of edge cases, because of regulatory, and because you can't screw things up? All of the above?

Immad: (46:00) There are just a lot of different types of businesses and they do a lot of different things with money. They all need to be accounted for in a way that's tax-ready and available. The two or three things that are generally hard: integrations are kind of hard. You have to pull in data from all sorts of places. It helps a little bit that we have a lot of first-party data in Mercury with your banking and cards, but it's still hard to pull all of that, pull from third parties, and put it in a data model that really works for people. So integration is historically trickier. You kind of have these two users: you have accountants and bookkeepers it needs to work well for, and it needs to work well for founders. So you're designing for two people. Accountants and bookkeepers spend all their day in these accounting systems, so it's a very high bar for productivity tools. There are tons of little details. How do we make it so they don't have to go to their mouse and the key bindings all work amazingly well? And then it has, probably—

Jackson: (47:03) It also, by the way, has to be really good for the founder who's going to look at it, I don't know, for an hour a day and needs less detail.

Immad: (47:08) And then, yeah, just a ton of edge cases. Each edge case has a way that it already exists. It's always a little harder when you're developing something where mature software is out there that people have high expectations for. We are asking them to stop using their existing mature software that has a bunch of checkboxes and say, "You should now use this new thing." You don't necessarily need 100% coverage on all the checkboxes they've developed over 30 years, but you need to be pretty high coverage for them to consider you. Even now, I wouldn't say we're fully there across every type of business. If you're a simpler business, it works, but if you have complex accruals or inventory management, we haven't built all of that out. There's still a ton to build.

Jackson: (47:55) It's interesting to go back to an earlier point around there not necessarily always being a shiny, new, super-exciting novelty thing. You're really building to solve pain points and make the experience less frictionful and painful. Depending on the customer, the thermometer on their pain has to get hot enough that they're like, "I need to stop using this crappy piece of software and switch to Mercury," versus, "I'm coming to Mercury for this feature I desperately want." It's a different disposition in some way, I think.

Immad: (48:29) One thing that has helped with that for Mercury is that people need new bank accounts for new businesses.

Jackson: (48:36) Got it.

Immad: (48:38) About 25% of people do switch from whatever bank they're using, but the rest are coming as their first company, first bank account. Something similar will work in books as well. I'm sure it'll be quite hard to get you to switch from QuickBooks or whatever you're using, but you need an accounting system, you're not using anything right now, and this thing is integrated with your bank account. You're good.

49:00 Becoming the Default for New Entrepreneurs

Jackson: (49:00) It's really like a bet on entrepreneurship. Maybe go back to where we started the conversation: this company, on some level, is a bet on new entrepreneurship.

Immad: (49:10) Yeah. If you think about Silicon Valley, there have been downturns, but every year there have been more entrepreneurs and more funding. I think it would be reasonable to say the next 10 years of entrepreneurship will be greater than the last 10 years, even though the last 10 years were pretty crazy. We're at the stage where more than 40% of pre-seed companies use Mercury. Normally we tell people one in three startups use us, but the earlier you are, the more likely you are to use us just because we didn't have the same market share with more mature companies. If you play that out for the next 10 years of companies, hopefully half of them are using Mercury, and that would be a really powerful position to be in. The future is defined by those companies, not the existing companies.

Jackson: (49:57) There's a stat I found: as you said, one in three startups banks with Mercury, yet in 2025, 73% of new customers came from outside the tech startup category, showing a broad appeal of what Mercury has built. Those are primarily new companies that aren't tech startups. Fascinating. Can you share anything about the rough distribution?

Immad: (50:21) It tends to be mostly digital entrepreneurs. You're still spending most of your time in front of a computer.

Jackson: (50:30) That's most new companies broadly, I would guess.

Immad: (50:32) Exactly. E-commerce is our second-biggest single category. Then professional services and consultants—there are a lot of those, and a lot of them use Mercury. Those two are our second and third. Real estate is another big one, about 5% of our customers. Then there's the VC and finance type that's almost in the startup ecosystem.

Jackson: (50:52) Was there a deliberate effort to become—at one point, maybe even still today—the startup bank? Obviously, for a little while, it was you guys and SVB, Silicon Valley Bank. We saw how that went. Was there a deliberate intention to move beyond startups, or did that just happen naturally?

Immad: (51:14) We've always been a horizontal product. It’s never been the case that we specifically said, "Don’t use us." But we started off with the tagline "Banking for startups." For a long time—I would say until 2024—we really focused on that. We were open to other businesses, but everything we did was startup-centric, and I really wanted to be focused. There was a big opportunity for us to win, and I would say we really did win it. When the whole SVB thing went down, we were pretty head-to-head in terms of market share in the early-stage space. We were about 20-ish percent and they were about 25%, and then maybe they were at 30%. By about 2024, the flippening had happened, where we were the dominant bank for early-stage startups. I should say all of this is as a neobank; we're not currently a chartered bank. At that point, we saw we had grown in these other categories. Even then, we were quite big in e-commerce and a few other things, but we never focused on it. It's hard to go 10x when you're already at 30% market share. We still want to do a good job on the startup side and keep dominating that, but to get the next 10x, we needed to go outside that space. That's when we really thought about how to do that. Even though we're a horizontal product, there's a lot to think through. Our acceptance rates outside the startup space were not that great because startups have LinkedIns with people who worked at Google and they have websites. The input data we used was very startup-centric. We made a bunch of changes with our cards product. Our cards product was very much geared towards the idea that you've raised a million dollars and now you get a corporate card.

Jackson: (53:15) I see.

Immad: (53:16) Whereas most small businesses don't have a million dollars straight away. There were a lot of small changes we made, and then that really took a year and a bit to start growing very fast outside the startup category. In 2025, we grew really fast outside the startup category.

53:34 Product: 100 Little Things Better

Jackson: (53:34) We've already talked a bit about product, but people really love your product, and anytime that's true, I think you should be proud. But in your category, I think that's especially special. There was a great tweet from Ashwin around the books launch. He said Mercury's unfair advantage is that they just do the obvious things that every bank should have done 15 years ago. That's cheeky in some ways. I'd love your reaction to that because, on some level, I'm sure that is true, and there are probably things that aren't obvious. Does that feel like a compliment, or does it feel like it's underrating aspects of the product?

Immad: (54:17) No, sometimes we do things and I'm like, "Why don't other bank interfaces do that?" When I first started Mercury, I really thought maybe there was a law that makes it so you can only see transactions within the last 90 days. Have you noticed that in your bank? You click back, back, back, and after 90 days, you can't. You have to go to bank statements.

Jackson: (54:37) There's no rule there.

Immad: (54:38) Yeah, it's just that they all use the same backend software, and that's been a limitation from 20 years ago. But there's no law about it for sure. I think there are two factors. Number one, we are very thoughtful, and we spend a lot of time thinking about it and trying not to be stuck on whatever seem like the limitations. But also, the bar was not that high.

Jackson: (55:01) Do you think the bar is higher now?

Immad: (55:07) The bar hasn't got higher with the incumbent banks. I don't think they've changed. There's definitely a slew of other Neobanks there that compete with us. We think whatever we're doing, we try to talk to customers and do it from that perspective rather than, "What are other people doing?" So at least for us, the bar is always high. There's definitely a massive premium to just being thoughtful and listening to customers and going like, "Okay, what are you frustrated about?" We try to be very close to customers. I'm very active on X and LinkedIn and I try to make myself available. Obviously I get busy sometimes, so I don't answer everyone's DMs. Sorry, everyone! But I try to look at them and I think there's always these things that you can learn from customer frustrations. Our designers, our engineers, our PMs are often talking to customers. We try to really be available and then obviously our customer support can talk to our engineers. There's just so much you can do. Actually, 80% of the fight is doing the little things. If you just keep doing the little things, that compounds over a long time as well.

Jackson: (56:18) Can you talk a little bit about the demo product and how that came to be?

Immad: (56:24) For people that don't know, you can go to demo.mercury.com and it's the full experience. When we first started, we made this design choice where we wanted to be able to develop the front end and the back end separately. We had mocked data to make it so when you're developing the front end, you don't have to go connect to the back end and wait for a response and get them to finish the work. You could just go do that. So it just so happened that we'd built up this set of mock data and a front end that work completely independent of the back end. At some point—I can't remember how this happened, it was probably like a hack week or something—we were like, "Okay, why don't we clean up that mock data and put it on a separate website?" We made this infrastructure choice which made it easy-ish to do, so we did the first version relatively quickly.

Jackson: (57:16) But it's—I mean, just to be super clear for people who haven't seen it—it's the full product just with fake money, basically fake data. It's super obvious in hindsight, but other products like this don't have this type of thing.

Immad: (57:34) I really hate talking to people if I don't need to.

Jackson: (57:38) I'm so sorry.

Immad: (57:39) No, not that way! I hate calling service providers. That's one of these things in the back of my head: how do we make Mercury in a way that no one ever has to talk to us? I remember I had some interaction with SVB and maybe I was interviewing some RM or something, but one of the metrics was: do they get to have a conversation with their customers, and how often? They were literally optimizing to talking with their customers. In the back of my mind, I've always thought of Mercury as the thing you come to if you don't want to talk to people, because it's so easy to use.

Jackson: (58:15) That's hilarious.

Immad: (58:16) So this is part of it.

Jackson: (58:18) Very Gen Z-proof product. You don't want to talk on the phone.

Immad: (58:21) Or at least engineer-proof. That's part of the reason Demo Mercury was exciting too. I was like, "Why are we talking to people, telling them about the product? They could just try it out." That was part of it.

Jackson: (58:35) I love that.

Immad: (58:37) But it's cool, right? You can try the whole product. Some people try to hide their product or make it hard to see. We're like, "I mean, if you want to copy it, you could—whatever." I could probably pull up some stats, but it really makes a difference when people go to that demo before they sign up. I think there's a 30 or 40% higher chance of them becoming active. It's really a big difference between someone not going to the demo and clicking on the demo. And it makes sense. Banking is a big decision. It's nice if you can try it out and see. It's actually one of the other things that I've always struggled with with Mercury. We'll just go around saying it's better, but it's very hard to go like, "Yes." I could talk about individual features that are better, but it's not—

Jackson: (59:19) This is show, don't tell.

Immad: (59:20) Yes, exactly. But it's also like a lot of other products have this one shiny feature that's better. They're like, "We're better because this X is so much better." Whereas for us, it is 100 different things are better. Our search is better, our virtual cards are better, and our sign-up is better. It's hard to really sell the "better" without experiencing it.

Jackson: (59:40) We talked a bit about this. There is a cohesiveness to the product that is felt, and it kind of has to be felt while doing multiple different things in the product. I was telling you when I walked into the office, there was just a sense of calmness and cohesion in this physical space, which isn't too surprising because the product itself—in some sense the product is a place. It's a place that I have to go do these things. And by the way, these things that I do, usually when I have to do them in other products, they cause my eyes to bleed. So it's cool. In some sense, the demo site feels like such an elegant display case for Mercury, which is actually all the things at once that have been considered to fit together.

Immad: (1:00:26) Yeah, 100%. I think one of the reasons the product is great is because we dogfood it internally. Mercury's always run on Mercury and I guess we're either the biggest or one of the biggest customers on Mercury. That's always been really important to me, to dogfood the hell out of it. Now that we have personal banking, we give that to all of our employees for free. I think the demo and the mock is part of that. I just want people in Mercury to be clicking around and getting annoyed about things all the time. It's one thing, customers complaining about something, but if an engineer has to go do something and they're like, "This didn't work," then they could just go fix it. You want to try to collapse down the steps it takes for a problem to be fixed to the smallest steps. The demo definitely helps in that as well. Because it's one thing if you're like, "Oh, the backend something something," but as you're designing the product, you're also checking the demo as you're doing it. There's all these weird things, like what happens when it's empty? What happens when—and you experience that as you're building it. So that also ends up being a better product.

Jackson: (1:01:37) When it comes to—you have this, I think, amazing engine where new companies want to use Mercury, whether it be on the personal side or what you could call hunting for larger customers. How do you think about tactically growing the business? It seems very likely that you are going to continue to be the default for startups and maybe broadly for new entrepreneurial companies started in the U.S. That’s an amazing position to be in. Did you do personal out of a sense of, "Wow, there’s this huge massive market that’s underexposed," or was it actually just getting so many requests from current customers? When it comes to how we go get more large-cap companies using Mercury, are either of those deliberate strategies that you want to pursue, or are you comfortable riding this new company formation wave?

Immad: (1:02:34) We definitely want to continue to be great at the new company formation wave, and I don't take that for granted. I think that itself requires work. Then we think about exploring new verticals. Obviously, personal banking is one of them, and you can sign up for personal banking without having a business bank account. Books is another one. Right now you need the business bank, but we will make it a standalone product as well. When you have one engine that works really well and it scales to 400,000 businesses, that engine is the powerhouse. But you do need to invest in the future engines. So, we're investing in both Books and personal and other stuff.

Jackson: (1:03:13) Maybe a different way to ask this question—sorry to interrupt—is that personal doesn't feel as obvious as Books to me as an offshoot. Maybe it is. Maybe I have the wrong read there.

Immad: (1:03:26) Your point is: why would we get into personal?

Jackson: (1:03:28) Yeah, without knowing—I'm sure there's so much nuance here, but you could make the case that personal is just a distraction in the same way that going after giant billion-dollar companies who already use QuickBooks or a different bank account might be a distraction.

Immad: (1:03:45) Personal was initially started as a five-person team. It really is the business product with some stuff stripped down to see if it works. It did well, and then we were like, "Okay, let's invest more in it." But the backbone is, again, the horizontal thing. Most of the features and elements of personal are the same product. When we make improvements in business—for instance, we launched Mercury Command, which is an inbuilt AI agent—that also works on personal. It wasn't like we needed a separate team to do this.

Jackson: (1:04:15) I see.

Immad: (1:04:15) So, the product is actually fairly similar. Sixty percent of our customers are LLCs. With an LLC, you need to move your money back and forth between personal and business. This distinction of, "This is a startup and it's a separate money source," doesn't quite exist for a small business where— Basically, I would say it's a similar product without a ton of extra EPD and really strong use cases for why you'd want business and personal together. We try to construct these things so it is a separate team that's not going to distract everyone else. It's not like the business invoicing team is thinking about personal all the time, and they shouldn't have to think about each other. You can think about them as autonomous units that can go after a separate vision from the rest of what we're doing. In terms of why invest in Books or personal or invoicing, it's a combination. I really like things where they have a big TAM by themselves, our customers really want them, and we have some "right to win." Those are the three dimensions we look at, and I think both of those fit that really well. I want to come back to this other thing which you were hinting at, which is worth talking about: how do we get bigger companies to switch to Mercury? We actually have a ton of big companies using Mercury, but most of them started when they were small and they've just grown. We have several companies with more than $500 million on Mercury accounts. Historically, it was actually quite a hard thing to get bigger companies to switch because banking is entrenched and you don't want to think about switching. Even if we show a great demo, people feel it's just too much hassle. What's changed more recently is we've got this much broader product suite. We can go to a company and say, "We're not just selling you banking. You're also going to use corporate cards, spend management, invoicing, and bill pay." Often, when people switch to us, instead of just switching the bank, they're switching three or four separate products they're using into one. It's a consolidation. It's a much better system, and it tends to save them money and hassle. Now, that switching function is working a lot better. In terms of volume, it's not as big, but those companies are bigger when they switch. It's only in the last year where we've had a compelling reason to switch beyond just banking.

1:06:51 The Right to Win

Jackson: (1:06:51) The "right to win" framing is interesting, both on that last thread and, I guess, for something like Books, how do you think about the right to win?

Immad: (1:07:04) First, can we make that product better being Mercury than a standalone company trying to build an accounting solution? To me, we've got that data. If you're literally using Mercury for banking, cards, invoicing, and bill pay, there's not that much more to do with your books. That's a lot of your core data already in the system. It's already reconciled, categorized, and set up really well. It's just a much better product together than separate. Then, is there a distribution advantage? For us, it's like, "Hey, most people set up banking before they set up an accounting solution." So we can be there when you need it. We can even help. You don't need the full solution initially; we'll start categorizing your transactions from day zero. Since every transaction is already categorized, you can end up getting a shape of a P&L without it being your full accounting solution. It's not always data; sometimes it's payment flows or some sort of unique thing already at Mercury. And the third one, which is a little more abstract, is brand and trust. Do we have the right to win there? In the case of Mercury, it's always been a painful thing to build up enough brand and trust to win your banking, because most other products don't have "Step one: Send me all your money." It's like, "Now you're using this, give us all your money so you can actually use it." Most other products are just a SaaS thing.

Jackson: (1:08:47) Give me your email.

Immad: (1:08:48) Yeah, exactly. So the bar is quite high for banking, but because of that, we win that relatively early on in our relationship. And things like Books are a high-trust situation, but we already have that trust. I mean, obviously I don't take it for granted. We have to keep delivering on that to have people trust us. But it does mean that a user's already in a trusted relationship, so they're more likely to be responsive to products that also require trust.

1:09:19 Brand as Care

Jackson: (1:09:19) Yeah. On the brand side, which obviously is fundamentally intertwined with trust—with a product like this, what are the other things that make a brand? It's clear to me that for you guys, maybe the brand could just be, "Hey, this product's amazing," but you guys, even from an aesthetic standpoint, have lots of things around little details and animations and the corporate card that's a partnership with an artist and is so elegantly done. I'm curious how you think about brand, particularly coming out of that lens of, "We need to be a brand that people trust at a very foundational level."

Immad: (1:09:54) I think brand and culture are kind of similar. And actually, it's hard to have a brand that doesn't align with your internal culture as well. There are a few elements we have to our culture that I think seep through. One is we try to be honestly really helpful and low ego, and I think that shows up in the brand. I think there's a lot of brands that don't talk in very authentic, human ways. They talk in these kind of weird ways where they're trying to be someone. We try to be authentic and human and treat people like they are sophisticated people, even though this is obviously not a space that they're experts in. We try not to talk down to people, and there's a lot of ways that shows up. Then care is a big part of it. It's like, how much do you care? There's little attention-to-detail things that make a big difference. I posted this cool particle animation we did as part of the onboarding process for Books. But I think there's these little things where you spend a disproportionate amount of energy that seems irrational. And that's kind of what art is in some ways. It's like, "Why did you even do this?"

Jackson: (1:11:12) Great definition. Especially today more than ever before.

Immad: (1:11:17) But yeah, it shows that we really care and we're willing to put in an effort beyond what seems completely rational at times. You can't do that everywhere. But in core parts, in onboarding experiences and things that we deliver to users, there are places where you can do that and people really appreciate it. I think brand and product also tie in together. One thing that I think really hurts the brand of banks is they'll charge you all these fees and it's always confusing why you're getting charged something. I remember when—I won't say who—but every month we'd get some weird charge in my previous company from our bank and we'd be like, "What the hell is this?" And then I would literally go, "What is this?" and they'll be like, "Oh, sorry, we shouldn't have charged you. That'll go away." But it happened consistently. I'm like, this is not a high-trust situation. But it's those types of things. There's lots of ways you can make money out of customers and we want to build a long-term sustainable business that makes money. We can't be giving away everything for free. But I think you can do that without feeling like you're confusing people and taking advantage.

Jackson: (1:12:28) You make more money in the long run too, if they trust you.

Immad: (1:12:30) Yeah, exactly. It really is a combination of things, which makes it actually tricky. And as an engineer, it's always kind of tricky. I'm like, "What's the formula for this thing?" But there isn't really a formula. You have to kind of live it.

Jackson: (1:12:43) You made a comment somewhere that it was really important for you to be visible as the CEO and the founder as part of that brand and trust exercise. I don't know if you were that in earlier companies—maybe it was less relevant. You have a podcast and it seems like over time you've really leaned into that, and you're good at it. I'm curious where that came from.

Immad: (1:13:08) Yeah, I think around 2017, when I was like, "Okay, I'm going to do this," there were two elements. A, I think with social media especially, the brand is also the person. It's much easier for someone to go viral than a company to go viral, if that makes sense. But there's also, if you want to trust Mercury and it's a company, it's harder to know who to trust. Whereas if you trust Mercury and Immad is the CEO, that's a much more direct relationship. So I started—and I have a hard time just saying things on social media, I don't know how other people do it—but I really focused on, "What are entrepreneurial lessons I can give people?" I've always been an active investor as well, so I'd often have a conversation with an entrepreneur and say, "Oh, that's a generalized lesson that everyone should have." So a lot of my content historically has been trying to be helpful to entrepreneurs, which I think aligns with Mercury's brand anyway. Over time that's become more and more true since we started. I think the human behind the thing matters a lot more than just the brand does. Actually, if you think about J.P. Morgan or whatever, historically that's been true in banking. The person is the thing that you're trusting and that relationship is very important to people.

Jackson: (1:14:36) I think it even—you used the word "care" earlier. You can express care through a brand and through product, obviously, but it's also like, "I know that person really cares."

Immad: (1:14:47) Yeah, yeah. And actually it ties also to, if someone finds my email or whatever—I get WhatsApp messages all the time—I will respond. Often fintechs and banks are faceless organizations where you get screwed and some decision's made and you can never reach someone. You can never get support for anything. Whereas I'm fairly accessible. People can get to me and I want them to have good experiences, so I try to be responsive to that.

1:15:19 Culture and Being Extremely Helpful

Jackson: (1:15:19) You mentioned culture. A great quote from you: "I think it's an ill-defined word. And I'm an engineer, so I hate ill-defined words."

Immad: (1:15:26) Oh, did I say that?

Jackson: (1:15:28) On a podcast somewhere? Yeah. You've talked a lot about culture elsewhere, but it seems that you have been really deliberate about it. I think you were really deliberate about it early on. Was that a reaction to past entrepreneurial endeavors? Was it something that you've been honing? In some sense, maybe a strange way to ask this question might be: how does Mercury's culture compare even to your last company, and how has that evolution happened?

Immad: (1:15:58) My last company, again, I didn't know what culture meant really. I think for a long time it was like, do people stay around and have drinks together? That kind of stuff. So we were very not deliberate about it. And over time, it tended towards whoever was the loudest voice in the room. There was definitely a disconnect because there was no culture, so it was kind of all over the place. And it annoyed me. So at Mercury, I was like, okay, what does... Again, I think culture is badly defined. Actually, Ben Horowitz has a great book on it, Who You Are Is What You Do, I think is what it's called. That's probably the best culture book I have read. But it really came down to that for me. It's like, what are the personalities that I want at Mercury? And what are those personality types and things that we encourage for people to do? That's a much simpler question to me than this nebulous, "What is culture?" That's where we focused. We wanted humble, curious people that are low ego, that really care about customers, really care about product. And we wrote like six of these things down, which has changed over time, but the nucleus has been fairly similar. Then you’ve got to write them down, you’ve got to hire against them, you have to encourage them. If someone's very anti those cultures, you have to performance manage them. But step one is writing them down. We did that really early on, just when we were four people. And it doesn't have to be some huge exercise. I think every company should do it. Just spend a few hours and go, "What do you really care about?" It doesn't have to be what we care about. Mercury is not an overly "crush your competitors" kind of culture. My sense of Uber at the start was like, "Do whatever it takes to win, crush your competitor," and obviously it worked for them. It's probably a different set of people that were attracted to that. But I think you have to pick your culture and, ideally—and this bit's the tricky one—it has to bleed through into your product and brand as well. Your culture should be aligned with what your customers care about as well. I think on the whole they care about us being helpful and curious and making products like that.

Jackson: (1:18:18) You mentioned the word helpful a few times.

Immad: (1:18:20) Yeah.

Jackson: (1:18:21) What does that word mean to you?

Immad: (1:18:26) I think at the end of the day, actually my sister said this once to me because I was a bratty teenager or something.

Jackson: (1:18:34) Older or younger?

Immad: (1:18:35) My older sister. She was just like, "If you're not going to be helpful, what's your purpose?" That just kind of stuck with me. I don't know the exact phrasing of it, but I was just not being very helpful around the house or something like that. But I was like, "Yeah, you're right." I think a lot of value that I think about, what I want to do for people, is be helpful to them because otherwise, what's the impact I'm having on the world? I don't want to sound cliched here, but I really genuinely want to be helpful to people. As a base thing of a purpose for life, I think trying to be helpful to people is a good way to think about things. If you approach things like that, it's not just altruistic. Obviously, you want to build a long-term company that also sustains and makes money and all of that, but I wouldn't want Mercury to succeed in a position where we weren't being helpful to our customers. I want to succeed because we're helpful, not despite being unhelpful. But yeah, it flows through a bunch of things. It's how much we care about customer support or how much we care about small customers that maybe don't make us that much money and how much thought we put into the product. Helpfulness is this core thing, but it's also internal. If I'm working with someone and I know they're going to be helpful, it's a very different relationship. I really don't want any part of Mercury to be in competition with each other or fighting each other. A lot of politics stems from people not being helpful to each other. They're thinking about their interests over other people's interests. So I think it's a good motto for life. I guess I haven't thought about it as that until this conversation. But yeah, why not be helpful?

Jackson: (1:20:18) It's a nice word. Maybe the reason it stood out to me is it's a common word, but not a word I thought about very much.

Immad: (1:20:27) I think the actual cultural motto we have is "be extremely helpful." Just being slightly helpful is not good enough.

Jackson: (1:20:36) You said somewhere you have a product interview that you do for almost everyone, including non-product roles.

Immad: (1:20:41) Yeah.

Jackson: (1:20:42) What does that mean?

Immad: (1:20:45) I can't remember the current construct of the interview, but it's like, "If you were developing this feature..." and I think the normal one is something to do with an Uber feature. So we just have someone walk through, if they were developing this feature, how should they develop it, and see how they explore that space. I learned a lot of lessons in my last company when I was trying to hire salespeople and how to hire salespeople and what kind of salespeople to care about. I ended up managing the whole sales team. I was the CEO, but I was managing the whole sales team and it was just so much harder to think about than engineers. With engineers, you're just like, "Can you code? Yes." Whereas for salespeople, I didn't know how you figure out whether they're good or not. So from that, I was like, okay, one thing I learned with salespeople is actually if they're product-minded salespeople, they think about—it actually goes back to helpfulness and curiosity to some extent. What is the type of personality that ends up building good products? It's someone who can be curious about things, who sees problems in things and says, "Oh, we should do it like this," who wants to build helpful solutions. So it's like systems thinking plus curiosity plus helpfulness leads to good product thinking. But you kind of want that everywhere at a company. When we think about the product, it's not like how do they think about building an actual mobile app or web app. It's more like, how do they think about the systems that they're developing within Mercury? There's almost no position in Mercury where good product thinking won't be helpful. And it helps that everyone's on that same page and they can talk to each other about it.

Jackson: (1:22:24) There's an empathy inside of it that I also think is worth noting. The defining distinction between Mercury and a traditional bank is that you are first and foremost a product company, I would argue, and most banks are not. So maybe this is telling—the atomic unit of the person who would be here is somebody who's empathetic to and attuned to what makes the product really helpful.

1:22:51 Command, Agents, and AI

Jackson: (1:22:51) We briefly talked about it, but obviously over the last year, and really in the last few months, you guys have started to aggressively launch AI features, from the MCP and API stuff, CLI, to, probably most importantly, Command, which is, for people who don't use Mercury, basically a ChatGPT-style interface for using the product that has some great guardrails and so on that we can talk about. But the experience of using it, you start to squint and you see, "Oh wow, this is the product going from a place, an interface that I do things in, to something that looks more like an agent or a service." I think everyone in Silicon Valley making software is kind of reckoning with this. Is all of software just going to collapse into a text box?

Immad: (1:23:37) Yeah, it took us a while before we launched any kind of user-facing AI products. Again, we're in a pretty high-trust environment where you don't want to hallucinate. I think we wanted the models to be good enough to really believe we could build a compelling product, which happened sometime end of last year. I kind of think of it as tying back to our overall vision: how do we help entrepreneurs? How do we make their life easier? How do we make finances not stressful to them? I think there's a world where it's not even a text box. I think the text box is a stepping stone. What the end solution should be is: what's the problem someone's trying to solve? Let's say they're dealing with a customer and they'll just CC Mercury or Command, and Mercury will deal with that customer for you. We'll get that SaaS agreement signed, maybe we'll negotiate it for you, we'll make sure that the customer pays on time, and we'll help you track how much they're supposed to pay you. All these things—at Mercury we have a 20-person finance team or something—maybe not that big, 15-person— and they do a bunch of stuff for me. I don't have to think about being paid, or paying a vendor, or negotiating contracts, or dealing with job offers. These are all things that entrepreneurs have to deal with daily. It's such a stressful thing and it's hard to outsource really, because you can't afford people who are spending all their time doing this stuff. But Mercury has a lot of this context where we can help you make that payment; we can store those details for you. I'm not saying we're there yet, but I think about that as where we want to get to, where people don't come to Mercury, maybe at all. Instead of coming to us to do a specific action—"I need to send a wire to my landlord" or "I need to invoice my customer"—people just turn to Mercury wherever they are. It could be in Slack or text or email or maybe ChatGPT, whatever interface they're in, and Mercury solves the problem for them. That's where we want to get to. It definitely wasn't possible two years ago, and even a year ago it wasn't. I don't think 100% of what I just outlined is possible even today, but there's an obvious line of sight that AI is getting good enough that you could have these agents that take care of things that finance teams or HR ops would be doing for you. We can really automate a ton of it.

Jackson: (1:26:26) Do you worry at all about durability or moats in this world where the product is collapsed into a single agent, especially an agent that I'm interacting with via another service?

Immad: (1:26:42) My take on it is the future is the future, and the worst thing you can do is fight the future. I want to accelerate the future. The future is: these things are easy and someone's taking care of you. And we want to help entrepreneurs. I think Mercury is uniquely situated to help with the future. Other people will be thinking about how to protect their systems, or whatever. One thing I've thought about is a lot of things that were problems have been helpful to Mercury. COVID was obviously a major problem—I'm not downplaying it as a major pandemic—but it accelerated Mercury because everyone was online and you literally couldn't go to a bank branch to get stuff done. Mercury was already the future.

Jackson: (1:27:37) It was just pulling the future forward faster.

Immad: (1:27:40) So I just think of AI as another accelerant. The bar of how much better our product is than an incumbent bank is already pretty big. We will embrace AI and do it with the right controls in the right way, but in a way that banks probably won't. They can't even build mobile apps, so it'll take a long time for them to build an agent. At the end of the day, what is aligned with customers? What is going to happen in the future anyway, and how do we help facilitate that? I think if we do that in the most aggressive way we can—and the future is not well defined either. Is it going to be that everyone is in Claude Cowork or ChatGPT, and that's the main interface where most people run their business? Or is it still Slack or text? I don't know, but we'll try to explore all of them.

Jackson: (1:28:30) On that last note, we briefly spoke about agent cards. My sense is we're probably on the precipice of things getting pretty wild. Maybe it will start in crypto or whatever, but agents are going to be—I already have an agent that has my credit card in a vault. These things are going to start ramping. I think you guys have been really thoughtful and relatively cautious, maybe is the right word, when it comes to this. There was a frame somewhere where it's like, our overall philosophy towards AI was to assume that LLMs will hallucinate. Rather than wonder if that could happen, we focused our efforts on ensuring that it'd be virtually impossible in those instances for the AI to do anything unintended or harmful to a user's account. But I think people are going to demand more and more, for lack of a better word, chaotic possibilities. We keep coming back to this theme, which is like you are balancing: we are a bank, we need to be reliable, we need to reduce harm, and we are actually pushing on what is possible. But this feels like a really, really hard domain to think through.

Immad: (1:29:35) Maybe. I don't think actually this is generally true. The thing that I think bank regulators or sensible people want is not that misaligned with what customers want. I think customers would not want an agent to go rogue and spend a million dollars or something like that, and we don't want that either. So the way we build Command is it can't do any action for you. It can present an action and you still have to click, and it actually goes to a separate non-AI kind of server to actually go send the money or whatever. So I kind of think about that across all of this space. Maybe you're fine with your agent having $50 a day or $500 or whatever, or maybe on a specific thing you give it $10,000 to go spend. But you'd want to have those controls. You can treat it similar to what would you give to an outsourced person that was buying on your behalf. There's only certain levels of controls that you would trust that person with, and you'd be pretty careful about it. Maybe over time this may happen with AI—over time you'll trust AI with your money more than you'll trust yourself or something. But for now, we're still in a world where people will want to have controls but you want to reduce the friction. So yeah, that's kind of how we think about it. It's like, oh yeah, the agent can spend money, but you have to approve it.

Jackson: (1:31:05) At what interval? Like approve every transaction, approve up to a certain...

Immad: (1:31:09) So I think it depends on each person.

Jackson: (1:31:11) You want to give the person the maximum flexibility.

Immad: (1:31:13) Yes, exactly. I think people will want different levels of flexibility and control. At least on bigger payments going out, I think we will require people to click on the approve, but it can be all set up for you. And that's kind of the annoying part. When it comes to card spends, I think it'll be much more like we have an agent card and you can set a lot of controls on it. But maybe the controls are a little more free because this is booking a hotel and the card is issued to be only able to book a hotel. Maybe it's a little bigger then because you've got this control on there. So you need a sensible set of controls that deliver the best experience to customers. I think everyone's trying to figure this out. Even this kind of idea of the agent sets up the transaction, but you have to click to approve it—I know we did that, and I know other people are exploring that. Where's the purely text thing, where's the interface, and where's the boundaries between that? I think text is not always great. Our Command agent does charts and things like that.

Jackson: (1:32:16) Right, right, right.

Immad: (1:32:18) But you can meld these things together. I think that's the future. It's not like the text box and the Command... the ChatGPT-type interface is purely text. I think over time it gets more complicated and delivers dense information in the right way.

Jackson: (1:32:35) This is a ridiculous question for regulatory reasons, if not other reasons, but do you think there will be a future—can you imagine a future—where agents can open Mercury accounts by themselves?

Immad: (1:32:51) I guess, at least right now, they're still making an account on behalf of someone else, right? I don't think we're there yet where there's neither a human nor a company behind that agent. Someone's got to pay for the compute and all that kind of stuff. There may be an eventual...

Jackson: (1:33:07) I might tell my agent to set up an LLC, though.

Immad: (1:33:10) Yeah, but then you're still behind that LLC. Who owns the LLC? I think at least in the US banking system, we would want to KYC and KYB—the LLC is a real LLC and someone owns it.

Jackson: (1:33:22) Well, and somebody needs to be liable too on some level.

Immad: (1:33:26) There's sanctions, there's all sorts of things. So there's definitely a world where agents are basically AGI or ASI, or whatever we want to call it, where an agent owns their own compute. Maybe you've seeded it with $1,000, but they're making their own money, they own their own compute, they own their own company.

Jackson: (1:33:46) I don't think that's some crazy distant world, by the way.

Immad: (1:33:50) It's not today. But yeah, in that world, either the agents—these AGI entities—will have to have alternative systems, like they'll have stablecoin wallets or whatever, or the US banking system needs to have agent identity, something like that. Maybe there's worlds where we're issuing driver's licenses or IDs to agents...

Jackson: (1:34:16) Maybe we'll all be entrepreneurs, agents or otherwise. It's going to be a strange one. I just have a few more things before we wrap up.

1:34:26 Ambition, Investing, and Leading

Jackson: (1:34:26) I mentioned earlier: unlimited TAM, unlimited ideas. When you think really ambitiously about Mercury—at least that you're willing to talk about—what types of things come to mind?

Immad: (1:34:38) I mean, there's still so much to do, right? We actually have this... we call it the Cosmic Cube. The X-axis is the types of customers we can do stuff for. The Y-axis is the types of features we can build for them or products we can launch. And the Z-axis is kind of compliance and quality and trust and support and all of that kind of stuff. So we've had this concept for, I don't know, five years or something like that. But if you approach the world like that, it's everything. It's like, what are the types of customers we do today, which is a very relatively small set of types of businesses. But we'll keep expanding that. We don't do anything internationally. We have a little bit in that personal space, but it's still very new—we only launched that in January. So there's this kind of user ICP expansion that has a lot of dimensions where you can do stuff. Then there's the types of products. You can kind of split it between financial products, where we don't have many lending products and there's a lot to build there, and then there's the software products. We just launched Books, but you can kind of think about what are the other types of products that deal with money and that people care about from a consumer business angle. Then I guess the third dimension of products is this kind of agents. What are the agents that will help people and what are the use cases they work against, and whether you have sub-agents... yeah, you can imagine an AR agent, an HR agent, an AP agent, a bookkeeping agent. There's just so many ways to help people with their finances and make their lives easier there. So a lot of those categories are huge TAM categories. I don't know if the Z-axis by itself has TAM associated with it, but there's so much to do. I mean, I think if you think about US banking, it's a $2 trillion market. It's actually kind of funny—it's as big as all of the software SaaS industry or something like that. But that's just the U.S., right? So it's a huge market and I really think we're just right at the start of it. It takes a long time to kind of compound in this system.

Jackson: (1:36:53) I understand that one of your cultural values is humility, but I'm curious if you have any specific ambitions around that very, very large market you just mentioned.

Immad: (1:37:07) How big do we want to be?

Jackson: (1:37:08) I don't know.

Immad: (1:37:09) Yeah, I mean, I think at some point I want to be a public company. I think that's the last thing as an entrepreneur—the last thing I haven't crossed the Rubicon on.

Jackson: (1:37:18) And is that a product of scale in a public company? Obviously, a $5 billion company at some point in time would have been public. It isn't today; it makes more sense.

Immad: (1:37:27) Yeah, I think because we set our path in this bank charter process, there's a lot of steps to get that and get that stable. I think one of the two things that public companies should be is scaled and stable. Ideally, the world wasn't like that and we could have subscale—still not unstable, but volatile—companies earlier on that would go public. But that delayed our thinking when it came to going public. So there's no imminent timeline, but at some point we'd like to do it. Outside that, the main thing is having more and more people that love Mercury and giving them more and more value, which will end up meaning that we make more money and we have a higher valuation. I don't personally see a cap to that. I think the markets are huge and we'll grow into that if we do a good job for our customers.

Jackson: (1:38:29) How has investing—I mean, you've been an incredibly prolific angel investor and now have a fund. You've talked a lot about investing, but I'm curious how that specifically influenced this seat as the CEO of Mercury, if at all.

Immad: (1:38:43) Yeah, a lot of our early customers were investments I'd made, and they were just the right size and obviously they had some trust in me. It wasn't deliberate necessarily, but I think 30 beta customers were almost all companies I'd invested in. So there was some help there. I know a lot about all sorts of things—AI and harnesses and even some lab stuff and LLM technology. I've done a bunch of space tech, so I know all sorts of things about that. Investing is this funny thing where if you want to invest in a space, you end up talking to five or six people before you invest in one company. Maybe you've had three hours of conversations on this space. Entrepreneurs tend to—part of being a good entrepreneur and being good at pitching is simplifying complex concepts into an understandable narrative. It's such a gift as an investor because you have really smart people at the frontier of what they're doing spending thirty minutes to an hour breaking it down for you. They're just answering all your questions. I really enjoy those; that's why I still invest, really. I want to be helpful to people, which is the second reason I invest, but I don't really do it for the money. I think it's just a fun conversation. You do end up learning all sorts of things and it keeps you abreast on whatever. I think I know a lot more about AI and it's definitely informed our strategy there. There's also an element where you look at a two-person company and they do all this stuff, and you're like, "Oh man, why does Mercury have 1,300 people and we aren't doing more?" It stokes this feeling that small teams can actually get a lot done. How do we construct Mercury in a way that we can get a lot done?

Jackson: (1:40:52) How do you think you've improved? You talked about the culture side, but how do you think you are different as a leader today in 2026 than you were, let's say, in 2021 and then in 2016 at the last company?

Immad: (1:41:06) I definitely made my most egregious mistakes in my previous companies. I think this is a benefit of being a fourth-time entrepreneur, that the worst of my mistakes I was better at. There's still a ton of learning that you do over time. The company changes, so you need to change the way you behave. One thing that I've learned to do, I guess somewhat recently—but maybe people would still say I'm not that great—is you have to be a little more careful about what you say when you're at a bigger company and people don't know you. Sometimes you can just be like, "I hate this." When you're three people together, people understand it. But when it's a bigger company and they don't know you, and you come into something and say you hate it, people can feel really disheartened. You have to spend a little bit more time giving more context and probably being a little bit less emotional about how you react to things.

Jackson: (1:42:09) It's amazing to imagine that you don't know everyone who works here. It would be impossible for you to know. It's crazy.

Immad: (1:42:15) Yeah, we actually have an all-hands next week. Nowadays when I go to those, I'm like, "Am I supposed to know you?" Until about 180 people, I interviewed every single person, which eventually just became too much. But at least you had that one connection point. Now I definitely don't know everyone.

Jackson: (1:42:41) You mentioned you were in this acting CTO role this year. What has that been like? Has that taught you anything more broadly in your real day job?

Immad: (1:42:53) It's been fun because I'm an engineer. When Mercury started, I was coding until we launched for the first year and a bit. I kind of missed it, so it's nice to go back to an engineering mindset and think about problems. I'm not coding that much, but still I'm fairly deep in some technical problems that I haven't been for a while. At the end of it, a lot of it is obviously still humans and organizations and products. It's not too different from the CEO job. In fact, being the acting CTO is just the more recent thing I've done. Before that, I was the acting risk operations person. Before that, I was the acting something else. Different CEOs approach things differently. I tend to be—whatever the biggest problem is, I'll dive in there. I'll try to figure out, make some progress, and then potentially hire someone, or sometimes just not have to do that. Every year as an entrepreneur, I end up doing different things. The CTO is just the newest thing. At least the good thing about this is I know what I'm doing. Whereas when I was risk, I was like, "I guess I'll figure it out." I can be a little deeper here. But it's also fun because AI is changing what an engineering team does and how it operates. I started as acting CTO, I guess January-ish, but the whole environment has changed since January. The number of PRs we do a week has doubled. Everything is changing. As an entrepreneur, I love that. Most people don't like continuous change, but that's what I live for.

Jackson: (1:44:51) Surf the wave. I may have mentioned earlier, but there's this clear attunement to aesthetics that is part design and brand and care and all these things. Is there any place you think that comes from? Is there an art thing? Where do you think that comes from?

Immad: (1:45:11) Actually, my cofounder Jason, who's our head of product and design, he is a very artsy person. I will give him a lot of credit for the aesthetics. For me, I really care about the customer experience. There's design to me and aesthetics—there's definitely the art side of it, but there's also an art to simplifying things and making them understandable and easier. To be a great product designer, you have to be a really good problem solver, and that bit really resonates with me. Whenever I look at any product, I try to remove my biases of, "Yes, I understand how Mercury works and how banking works," and go, "If I'd never seen any of this before, would this button make any sense? Would this flow make any sense?" How do we really be extremely empathetic? So that's one aspect. And then I also try to get really annoyed about things. I think just being annoyed about the world is quite a useful attribute as a CEO. You're just like, "Okay, why does it have to be like this?" So I try to harness that annoyance into feedback on product ideas or whatever it is.

Jackson: (1:46:33) You've got to contain it, I think, but it's probably very useful.

Immad: (1:46:36) Yeah, but there's probably a negative kind of annoyance where you're just frustrated and can't articulate it. But positive annoyance—where it's like, "Why does this stuff exist? It could be better. This is how it would be better, and why don't we do it like this?"—that kind of thing I think can be pretty positive.

Jackson: (1:46:53) Yeah.

1:46:54 Family, Pakistan, and America

Jackson: (1:46:54) You have three daughters, as I understand. Is there anything that you particularly want to instill in them?

Immad: (1:47:05) My two favorite personal attributes, which I think are real gifts that I would love to give them, are being optimistic and being curious. And maybe the third one is being resilient. But it is hard to teach these things. It's not obvious. Actually, teaching anything as a parent is quite hard. I think the main thing you have to do is be an exemplar to those things. If you live it and breathe it, you can teach it. I've always thought that I'm fairly optimistic and I'm very curious, and both of those things are such free gifts. If you're just optimistic, most of the time things are actually pretty good, and if they're not, you don't worry about it all the time until the bad thing happens. So optimism is great. And then curiosity—the world is such an interesting place. If you're learning and curious, a lot of my enjoyment comes from learning things and being curious about them. I also think optimistic, curious, and resilient people are the most successful at whatever they do. Those three things are the core of being able to do anything successfully.

Jackson: (1:48:24) Yeah. They're also, I think, an engine to keep going.

Immad: (1:48:29) As an entrepreneur, so much shit goes wrong early on that you have to go, "Oh, yeah, that's fine. The last three products I launched didn't work at all, but this fourth one is going to change the world."

Jackson: (1:48:44) You were born in Pakistan, you had a tour in the UK—I don't know that that was necessarily the highlight—and then you ended up in the US and you've been here for a long time. I'm curious, maybe particularly for Pakistan and the US, what gives you pride about each of those places?

Immad: (1:49:07) It took a while to get it, but moving to America is a tricky place. I went through four different visas—they don't make it easy—but I remember getting citizenship, which I think was in 2018 or something, and you feel a lot of pride. That's the first time where you really feel like you're welcomed and there's a nice ceremony. Actually, the step from green card to citizenship is relatively easy, so I remember feeling a lot of pride then. Now America is actually the place I've lived the most. I feel like there's something nice about America where, and I guess it helps I'm in San Francisco where there are a lot of immigrants, but I feel like you can become American in a way you can't become British. So I've always liked that about America, and obviously it's a very entrepreneurial place; I'm very aligned with it. Pakistan, to me, was where I was until the age of nine. That's still where my family comes from and my roots. I have a somewhat funny story which is probably a big source of pride for me. On my mom's side, my 12th grandfather—you have to go 12 generations back—was the ruler of the Durrani Empire. Actually, her family, the Barakzais, were basically in control until the Taliban.

Jackson: (1:50:33) Wow.

Immad: (1:50:35) The Durrani empire stretched from Afghanistan to a lot of Pakistan to a bit of India. But we were quite poor growing up, especially moving to London. So it was this weird thing that I grew up with, where I was like, I might be poor and I might not have anything, but I always felt like I could achieve anything because of this weird familial… I do think these stories that people tell themselves—like, what are your self-limiting beliefs? I didn't have a limiting belief because I was just like, "Well, I can do anything because this guy did this crazy thing."

Jackson: (1:51:11) [Overlapping] Great-great-grandson of an emperor.

Immad: (1:51:14) So I think as an entrepreneur, that's always been— And that's something I really like about San Francisco. You can go to an event and see some person who built a billion-dollar company. I did Y Combinator and you get suddenly exposed to a bunch of people who come to do these dinners. You sit next to them and you're going, they're smart, but they're not some superhuman person. I think it's hard to believe you can achieve something when you think it's someone else in some other situation. Whereas if you feel like they're just another human and they have their own struggles—that's always been a really empowering thing to me. I don't think I'm amazing or anything, but I also don't think everyone else is that amazing either. We're all trying to figure stuff out. So I guess I talked about Pakistan and America. I don't know if I have a great one for the UK. It was definitely a formative set of years for me, and I was there from 9 to 23. What gives me pride from there...

Jackson: (1:52:19) For what it's worth, I was actually asking primarily about Pakistan and the US. I know that it seems like the UK thing was important for your resilience, maybe.

Immad: (1:52:30) Yeah, sure, it was important for resilience. And now also, I think people probably talk shit about the social welfare system or whatever, but I'm definitely a product of it. I went to university for basically free and I went to an amazing education system. I definitely wouldn't have got that in Pakistan. So in that sense, I have a lot to be thankful for in the UK.

1:52:56 Money, Freedom, and Energy

Jackson: (1:52:56) We talked about it at the very top around your answer for banking. I think money is something that in some sense is like this fuel, and it can also be an area of obsession, and it can also be an area of frustration and pain. But I'm curious if you have any advice or thoughts for people on how to make it an empowering idea and substrate rather than this thing—again, either on one end of the spectrum, obsession, or this thing that rules over you and is a pain point.

Immad: (1:53:31) So there's this funny concept—which I think maybe this is too libertarian a concept—but money is like freedom, in a way that you have to imagine a world without money to see how important money is. A world without money, what does that mean? I wouldn't be able to sell my services in any way. I'd have to do my own thing, I'd have to go farm or something. Or maybe I can figure out how to build something and I have to barter for it. Or maybe you have a communist-type system where someone else is defining—maybe still industrial, but there's no money—and someone's defining who can buy what and who can work in which job and all these things. So both of those things are pretty messed up. Money is this massive unlock of freedom which is hard to really get because we're just so used to it that we don't realize we can swap this abstract thing for anything, and we can sell our services and we can build things. I think that's one view on it, which I think is interesting to think about. There's an alternative to this thing—this is actually freedom. And it's such an abstract freedom that I think we take it for granted. Not that I'm saying there aren't problems with society—I think problems with society actually mostly are government-based, in my opinion, but that's a whole separate conversation. I think the second thing that's really interesting about money is for certain activities, the upside of what you can get out of them is unbounded and the downside is bounded. That is what got me really comfortable with being an entrepreneur. I was like, hey, I am forsaking a job, right? But the maximum I'm going to get out of the job is whatever. I think I was making $50,000 or something when I first quit. So I'm losing $50,000, but I could make a billion-dollar company. I'm not saying it's easy, and that equation is obviously different for different people, but doing things that have unbounded upside and capped downside is great. And actually, I try to teach this to my kids: to do things that have unbounded upside and capped downside, and not the reverse. There's a few things like drugs and gambling—

Jackson: (1:55:58) We're under-calibrated on risk, or miscalibrated on risk, I should say.

Immad: (1:56:02) Yeah. Where the downside is unbounded, where you can really ruin your life. So that's actually a nice framing for money as well—to try to think about what are things that have this kind of unbounded potential to them? I feel like it's changed a little bit. If you're an AI researcher, you can get unbounded potential without being an entrepreneur as well. But maybe you can still have that framing.

Jackson: (1:56:27) There are benefits to entrepreneurial life that aren't just money too.

Immad: (1:56:31) Sure.

Jackson: (1:56:31) That is freedom.

Immad: (1:56:33) Your phrasing was money.

Jackson: (1:56:36) A quote from you: "I think something people don't understand about time. Time is about energy, not about time." Can you say more about that?

Immad: (1:56:48) So I think it's like, if you're doing something that gives you energy—like having this conversation or talking to entrepreneurs or playing sport, like tennis—you can do that for a really long time. Whereas if you're doing something that's kind of annoying, like fundraising or something, sometimes when I'm fundraising, I do four hours of pitches and at the end of it I'm like, I'm so done. I can't do anything else today. You can do one thing for 12 hours and the other thing for four hours. So I think that most people actually have a lot of time in the day. There are many hours in the day. I think the limiting factor for most people—at least for me, and maybe other people are slightly different to me—is: am I doing things that give me energy and I can just keep going at them because I'm gaining from this over time, versus things that are just pulling me down and I'm finding it hard to get through the day? This really struck me when I was in my first job. I'd show up—and I wasn't like a hard worker—I'd show up at 10 and I'd try to leave at 4. I was just so tired at the end of it because it was such a draining thing. And then when I quit and I was doing my own startup, I'd wake up at nine, go till midnight and I'd still be energized. Time-wise, I was clearly doing way more time, but somehow I was still feeling energized about it. I also think about this in the context of having kids. I think a lot of people forsake life for being an entrepreneur. Whereas I've always had this feeling that I want to be an entrepreneur forever. So I was like, I'm not going to give up on my life because I just have to find a way to do this alongside my life. In that context, it can definitely be like, how do I construct my life in a way that when you're at work, you're maximally productive because you're filling it with things that give you energy? But there's no hiding from the fact that as an entrepreneur you have to do things that are a grind that take away from that. You just have to kind of balance it. And then, I think of family life as something that gives me energy.

Jackson: (1:59:10) That's all I got. Thank you, sir. This was a joy and an honor.

Immad: (1:59:13) This was a great conversation.

← Back to all episodes