I Built A $1B Company... Here's 7 Business Ideas I Would Start This Year

Immad Akhund

I hear about these companies that go from zero to 100 million in two or three years, or whatever. The equation is **very simple** from a company perspective.

Shaan Puri

What's up? We got my buddy **Imad** here. He's the founder of **Mercury**—an awesome company that I use and love. And you've created a billion-dollar company, dude. You did it. You did the thing that all of us, when we were living in San Francisco in our twenties trying to build these companies, you finally did it. How does that feel?

Immad Akhund

It's funny you say that. We raised our Series B in 2021, which was actually about four months after we did our last podcast together on *My First Million* in January 2021. Now we're just announcing our Series C, which is a $3,500,000,000 valuation, with Sequoia leading. In 2021, when I raised my round, that was at a $1,600,000,000 valuation. It was so weirdly *anticlimactic* because I started as an entrepreneur in 2006—so 2021 is about fifteen years. I've been working toward this objective of building a unicorn company. I moved out when I was 22, and for a long time that's all it was about for me: getting this really successful company. Then you hit this objective—becoming a unicorn—and nothing changes. I have this helmet behind me; it's from Joe Montana. He's an investor and a football player, and he sent me the helmet. That's literally the only thing I got when I became a unicorn. I mean, obviously I had a successful company, but I keep it there to show that these objectives don't really matter. You have to **enjoy the journey**, because you hit the objective and then you're like, "Okay—now what?" </FormattedResponse>

Shaan Puri

Dude, **Joe Montana** should make that his thing. He's... he should send a helmet to every founder who becomes a unicorn at

Immad Akhund

This will mean, at least, he sends it to the ones he funds because it says **"Billion Dollar Club"** on air. It's kind of cool.

Shaan Puri

Alright. I told you—I said I *love* to brainstorm ideas with you. I thought you would come with a bunch of **fintech** ideas because you're the bank guy, but you sent me a list of bullet points, none of which sound like **fintech**. I'm excited. Give me some of these ideas: things you think founders today could be doing, or that you would be interested in doing if you weren't doing **Mercury**.

Immad Akhund

Yeah, so I had three trends that I think are interesting. Trend number one is *space and space tech*. There are a few reasons I like the trend. Number one — I'm just a sci-fi nerd; I love space. I think it's really exciting, and it's finally hit a point where there are all of these enablement factors. A few years ago, getting anything to space was a $100 million endeavor. Literally, $100 million — because there was no SpaceX and all of the satellite stuff was these hardened, radiation-proof satellites. Now, SpaceX did something like eight flights in one week; it's insane. There's a repeatable channel to get things in space. It's much, much cheaper, and it's getting cheaper every month. Nowadays we have microsatellite platforms where it's basically a computer in space. It's not going to cost you $1,000,000 just to build the satellite — you can do it for about $100,000. This is a huge enablement. I've invested in a bunch of companies. I did a company called **Albedo** that's doing very low Earth orbit imagery, so you can get 10-centimeter resolution pictures from space. Everyone's kind of excited about *AI*, which I'm also excited about, but that's heavily — you know — everyone's doing it.

Shaan Puri

Competitive, yeah.

Immad Akhund

Whereas I think that, actually, space is still super open. There aren't that many people attempting things in space. The crazy idea I sent to you — which I don't love, and we need to come up with why we should do this — is a **nuclear power plant in space**. That would be really cool. Potentially, we put this thing on the Moon and then mine water from the Moon. One of the cool things about water is if you split it, you get *hydrogen* and *oxygen*. Hydrogen and oxygen together make rocket fuel, and you also get breathable oxygen, which is useful. Having power in space is pretty useful, and there are lots of interesting resources on the Moon. If you have a nuclear power plant there, you could probably... there's this other cool company that does a thing — I don't know if you've seen it; I can't remember the name — but it spins things super fast and then basically fires them out. I think it's very hard to get something from Earth's surface into space by spinning it around and flicking it, but on the Moon it's much easier: gravity is lower and there's no air resistance.

Shaan Puri

Right.

Immad Akhund

So yeah, if you had a nuclear power plant, you could power that. I thought that was cool. I'll talk about my second theme because it ties into space as well. The second theme is **defense tech** in general. Obviously we've seen Anduril, but there are kind of three trends in defense tech that I think are interesting. First, there's a lot of geopolitical disturbance. Everyone is basically—yeah—we're moving into a multipolar world. That means there’s India, there’s China, there’s America, and America is no longer the global policeman. Since the 1990s, America could act as a policeman because there wasn't another power to really challenge it. Now it's not just the U.S. that needs to rearm—it's all of Europe, India, and others. Everyone is arming, so you have many new buyers in the market. Second, autonomous vehicles have changed the game. Drones and AI in general can be applied to all of this stuff. Third, you're going to get a ton of drones. It’s not just billion-dollar fighter planes anymore. You'll have thousands or millions of tiny drones that are easy to manufacture. All of these trends make it possible for startups to compete again in ways that weren't possible before. I think there are two types of ideas that are interesting. One: how do you make more things autonomous? People are doing autonomous drones, but you could do autonomous subs, autonomous boats, autonomous tanks. When you make them autonomous, you can also change the form factor—you can make them smaller and produce thousands of them. Two: a lot of these autonomous things favor defense. I was reading about the Ukraine–Russia war, and it's actually like trench warfare, which is kind of crazy—we haven't seen that since World War I. They're literally digging trenches, and that's why the front lines are hardly changing: neither side has air superiority. Drones complicate air superiority because a tiny drone can take out a million-dollar plane. In that world, one side has autonomous offensive systems. The other side is focused on building more defensive autonomous systems. For example, you could have machines that automatically build trenches or autonomous mines that are smarter. I know that sounds bad to people, but you can design them so that if it's a civilian they don't detonate, and if it's a tank they do. You can create safer mines that move around. To tie back to space: in the next major conflict, space warfare will be a big part of it. If you can knock out satellites, you can cut down on Starlink and other communications, remove the visual reconnaissance layer, or knock out GPS. Some of that reportedly happened during the Ukraine war—there are rumors about satellites being disrupted. I think that's going to be a major area. So the counter is: how do you create defenses for satellites? Maybe you'll have stealth equivalents of satellites: move them only when the sun is obscuring the viewpoint, then harden them so they're not reflective. There are a ton of ideas around defensive capabilities in space as well.

Sam Parr

Alright, so a lot of people watch and listen to this show because they want to hear us just tell them exactly what to do when it comes to starting or growing a business. A lot of people listening have a full-time job and want to start something on the side—a side hustle. Many people message Sean and me and ask, "I want to start something on the side. Is this a good idea?" What they're really saying is, "Just give me the ideas." Well, friends, you're in luck. My old company, **The Hustle**, put together 100 different side-hustle ideas. They appropriately called it the **"Side Hustle Idea Database."** It's a list of 100 pretty good ideas, frankly. I went through them—they're awesome. The database shows you how to start them, how to grow them, and gives you a little bit of inspiration. Check it out. It's called the **Side Hustle Idea Database**—it's in the description below. You'll see the link; click it and check it out. Let me know in the comments what you think.

Shaan Puri

So, what I like about what you just said is that there are probably three core things I think about when it comes to great ideas. I will differentiate a *great idea* from a *good idea*. So, on this podcast a lot...

Sam Parr

Of times, we'll talk about.

Shaan Puri

A good idea means a business where there is demand. There might even be competition; it might already be established, but that gives you a bit of a playbook. It's not going to be winner-take-all — there's room for everybody to eat. What a good idea typically profiles out to is that you make millions of dollars, maybe tens of millions, maybe $100 million, maybe a couple hundred million dollars, with a higher likelihood of success. But you'll pretty much never get to the billion-dollar, or truly groundbreaking, new category-creation type of companies by doing that. What you're talking about is a different criteria for great ideas. A great idea, I would say, has three tests. The first is the **Peter Thiel test**. So, famously Thiel went to Y Combinator [YC] and was asked to give a talk. He drew two circles on the whiteboard, like a Venn diagram. He labeled one circle "sounds like a bad idea" and the other "is actually a good idea," and he colored in the middle. The point was: > "The best ideas sound initially like a bad idea but actually are a good idea." That's hard to do because you don't know what's actually a good idea — you have to think for yourself. What it is good at is it reminds you not to run away if your initial idea sounds a little crazy or out there, or if eight out of ten people who first hear it say, "Nah, no way." So that's the first thing: a great idea typically comes disguised as a bad idea. The second thing is some inflection.

Shaan Puri

So you talked about how for space — suddenly, if **SpaceX** is doing eight trips in a week, right — it’s like going to space went from somewhat impossible to possible, now a regular thing. It’s like, “Oh, I could just hitch a ride on one of these trips.” The trips are getting cheaper every year, or somebody’s already taking a payload up and I can just add something to it. I can build on top of their satellite platform that’s already going to be up there. I think that’s a huge inflection. The other one you talked about with defense is **autonomy** — just the ability to make things self-driving, self-flying, self-swimming... basically give a thing eyes and, now, because of **AI**, it has a brain. Suddenly you can take humans out of the equation for that job. When it comes to warfare or defense, that’s great because that means lives saved when you do that. So you have these inflections. Thing number two is autonomy. Thing number three, I would say, is that it **scares everybody off**. Like you were saying about AI in general: if I said I’m building an AI agent startup, that’s probably the biggest bloodbath of competition right now. But you’re kind of right that smart, capable teams doing interesting things in space still have way more opportunity and way less competition relative to the number of teams pursuing them. Same thing in defense — the average smart, capable team is not going after that right now. So there’s just a lot of room, and these are huge TAMs (total addressable markets). Space and satellites is a huge market, and defense is obviously huge, especially as the whole world is re-arming. So it kinda hits all three. Even though my first reaction when you say “nuclear power plant in space” is like, “All right, that breaks my brain a little bit — I don’t even know what that means, is that even possible?” — I kind of like the underlying thought process you have going in, even though I don’t fully get the end-product idea because I just don’t know anything about space.

Immad Akhund

Yeah, I mean, I have some simple ideas that are maybe good but not great, but more achievable. Here's a fun one I thought about last night: I thought it would be fun to have a **billboard in space**.

Shaan Puri

Yeah, the first task in space.

Immad Akhund

So, I had this a few years ago. I was going camping and I didn't have internet. There was a *Starlink* satellite launch — I don't know if you've ever seen it. I think they fixed it so it doesn't do this anymore, but they basically launched the satellites and were positioning them so they were all visible. There were about twenty in a string in space, and you could see them with the *naked eye*.

Shaan Puri

"With the telescope or just the *naked eye*?"

Immad Akhund

No, just the *naked eye*. You just look up and there's a string. Initially I was like, "Are those stars? Are they UFOs?" They're very bright—**brighter than a star**—completely in a line, and they're moving quite fast because they're in LEO [low Earth orbit], right? I was like, "Oh, that's kinda sick. You can literally, as a human, launch this thing that changes everyone's view from Earth."

Shaan Puri

Right.

Immad Akhund

I thought that was kinda insane. But yeah, I don't think it'll be that hard to put a **billboard in space** and make it so you could **view it from a telescope**. I don't think you could do it—like, I mean, you could probably do the naked-eye thing, but obviously people will complain. Making it so you could view it with a telescope would actually be kinda cool, and then people would buy that. This one's more a good idea. I mean, I don't think that's a $1 billion idea, but... well, do you remember?

Shaan Puri

Yeah, there was a whole service around *buying and naming stars*. Do you remember this? This was kind of a big idea in the nineties, maybe the early 2000s.

Immad Akhund

*Free-body problem* — that's like one of the... one of the ideas.

Shaan Puri

I haven't actually read it. No — it's on my kind of to-do list. Back in the day we had this business called *Birthday Alarm*. Birthday Alarm was a tool to remind you when it was your friends' and family's birthdays, and then you could send an e-card — just a greeting, you know, an animated e-card. There was this company that approached us and said, "Hey, if you add the gift of naming a star after someone, you get a certificate," blah blah blah. It was like $30 to buy a star in someone's name. That company was doing, I remember, about $20,000,000 a year. It was just a couple of guys. They approached us; we said no. But when we got to know them, I was stunned that they were doing that much revenue — at least at that time. I do think it was a bit of a fad, but still, that was very impressive to me.

Immad Akhund

I thought the other one that would be kind of cool is—when I die, it would be cool to send some ash to space. I actually don't want to be cremated, but if I was, having some ash sent to space as a piece of memorabilia seems interesting and would be easier and more practical. I do think *intelligence as a service* is an interesting idea. You mentioned "AI agents"—there are tons of companies going after these big categories, right? Like AI SDRs, AI customer support agents. I haven't seen too many companies go after more niche ideas where people are willing to pay. So, an AI SAT tutor, right? Imagine that. AI is great at education because: - (a) the content is extremely structured and easy for AI to generate, - (b) it can be completely personalized, and - (c) it's just extremely patient. I mean, when I try to teach my 13-year-old, I'm pulling my hair out after a few minutes. There's a whole class of tests where you could probably make an AI agent that's great at both the test and teaching it—like the bar exam or the MCAT. I mean, that's such a no-brainer. </FormattedResponse>

Shaan Puri

Is there somebody doing that? Somebody has to be doing that, right? Because...

Immad Akhund

I haven't heard of anyone doing a great job of it.

Shaan Puri

Nobody's won yet, which is amazing. But you look at Kaplan and the big test-prep companies — they were built, like, many decades ago. I used to buy those giant books with the practice tests in them from Princeton — I think Princeton Review or whoever — when I was studying for the SATs. Then it moved to video and online after that. But now you're right: if you just used **AI** — a personal, one-on-one tutor — I think that's obviously the best. I don't know if you've seen the studies or read about the "two sigma" thing. When it comes to tutoring, basically nothing has shown as measurable an effect as **one-on-one tutoring** in education yet. For a long time, the claim was that you could get about two standard deviations above the mean. By the way, this is kind of debated; some people say it was measured wrong and it's not true.

Immad Akhund

I think, *intuitively*, it feels exactly like when I'm learning something. If I talk to someone—like, if I had a gun to my head...

Shaan Puri

"Get the best grade possible on this test," right? If my life depended on it, would I choose just showing up cold? No. Would I choose just a book where I self-paced and self-studied? No. Would I choose a library of videos? No. If it was a tutor who comes to my house and sits with me for six hours a day and that's how I study for this, I'd say that's my best shot. And if it was the smartest, most patient tutor—with infinite knowledge—who knew exactly where my level was and was ready to work with me the moment I was ready, at all times of the day... right? That's an **AI tutor**. So I think that's kind of great because it's so scoped to, you know, one chat, one specific test, which people are not doing because they voluntarily want to, you know, educate themselves with, like, "I need to pay $X to get Y grade to get into Z college." Yeah, you're willing to pay.

Immad Akhund

Yeah, I think the good ideas right now in these AI-agent–type products are more *niche ideas* that you can expand from later, rather than trying to do *all* tutoring. I think that's actually quite a hard problem, and AI is not quite there yet. If you pick just one thing and completely nail it, then you can always expand from there as well.</FormattedResponse>

Shaan Puri

Right, right — I like that. What are some of the most interesting companies you've invested in recently? Is there any company that's *taking off* or *doing well* that you feel more people should know about but don't yet? </FormattedResponse>

Immad Akhund

You know, this is like what I put as *intelligence-as-a-service* company. I'm an investor in 11x, which does AI SDRs, and I'm also an investor in Decagon, which does AI customer-support agents. They just have these insane numbers. I don't think they're public, but you hear about companies that go from 0 to 100,000,000 (one hundred million) in two or three years. All of these hits follow a very simple equation from a company perspective: "Hey, I'm paying a ton of money for customer support or SDRs. If you can do the same task with some sort of fallback to humans at half the price, sure — let's do it." There are definitely a bunch of spaces where this applies. The big ones are sales, customer support, legal, accounting — a bunch of these. I'm an investor in many of these companies, and there's definitely a moment where they're really winning. Obviously it's still early, but it seems obvious to me that a bunch of these things will be delivered by AI on the other side. Every company that has sales or CS teams, the CEO and the board are asking, "How do you make yourself more efficient through AI?" Buyers are really looking for solutions, and there aren't that many actual [enterprise-ready] players in each of these spaces. There are a lot of companies, but normally it's really hard to do enterprise sales — it's a nine- to twelve-month cycle. When you have buyers that are saying, "Hey, I need to deploy something to show that I'm doing stuff in AI," it creates a moment where you can actually have explosive enterprise-sales growth. That kind of explosive growth is rare; it only happens in these kinds of trends.

Shaan Puri

"Would you — let's say you weren't doing Mercury right now. You had all your free time, and you had everything you know today. You wanted to find an *AI* idea, a *winning AI idea*. What would be your process? Explain to me how **Imad** finds—how he kind of comes up with ideas, vets them, and figures out which idea to pursue versus which idea not to pursue."

Immad Akhund

**I approach things both top-down and bottom-up.** So, **top-down**—I actually did this as part of your podcast prep. I went to ChatGPT and said, "Hey, tell me every human worker in the U.S. that does knowledge work; give me a list of all of them and tell me how much labor costs across each of them." That's very top-down. You can then say, okay, there’s education, medical, lawyers, accounting, and all these knowledge-worker types like design, coding, etc. From there you can do a matrix: here are the most interesting ones, here’s the TAM [total addressable market], here are all the companies in those spaces, here are the bits that aren’t done. Maybe you want to avoid regulated spaces, or maybe you want to go all in on regulated spaces. Maybe you think, "No one’s touching doctors because everyone’s scared of doctors—why don’t we just make an amazing doctor for, I don’t know, Africa?" Because there’s less regulation, you can prove it out, etc. Then I also go **bottom-up**. I ask, what gets me excited? I love serving entrepreneurs because I’m an entrepreneur—my friends are entrepreneurs. I love the idea of people making things. So I’d say, okay, what are things that entrepreneurs struggle with? One thing I still don’t have a great solution for is: as a CEO there’s so much knowledge across the company. I’d love something—there are a few companies doing this—but I’d love to have an AI twin of Ahmad that is continuously absorbing all the knowledge across the company and maybe across the internet, and telling me stuff like, "I think MR should really pay attention to this." Even funnier, you could make it so on Slack there’s an MR twin and someone could ask it questions, like: "What would MR think of this idea?"

Shaan Puri

Right, yeah.

Immad Akhund

Actually, a funny thing came up the other day and I thought, *I'm pretty sure my AI could answer this.* Someone said our boardroom is kind of small and we just added two people to the board. They asked, "Hey, should we do the board meeting at our law firm?" I was like, *No—fuck no.* I'm never going to do a board meeting at a law firm. That sounds awful. I'd rather stuff, like, 16 people in a tiny room than do that. I feel like my *AI twin* could have answered that question. That would be my *bottoms-up* approach. I do think you should work on things that really excite you. For example, an *AI SDR*—I'm happy for those guys, but personally I would not get excited about that. I hate receiving sales emails and calls. Do things that would make you wake up in the morning and say, "I'm excited to solve this problem," because these things take such a long time. I've been doing Mercury now for about eight years. If I were doing something boring, I'd probably wake up and wonder, "Why do I have to do this again?" So the bottoms-up approach is to think about what you care about and what you want to see in the future, then try to come up with that. Then merge those two concepts together. Often you do something and find out it's a bad idea. Mercury actually had a very smooth ramp: I liked the idea, executed it, and we never had to pivot. But often, as you learn, you realize, "Oh, actually this doesn't make sense, but that does." Part of ideation is to actually go build things.

Shaan Puri

Do you... So let's take an idea. We'll role-play it real quick. Pick one of the ideas — plausibly you'd be like, "Oh yeah, that kind of checks the boxes." It kind of works top‑down and makes sense; bottom‑up, it resonates with me. What would be an example of that? Is it the *digital twin* thing, or is it...?

Immad Akhund

Let's do the—yeah, let's do the **digital twin**. I think that's kind of the most cool.

Shaan Puri

So, you have a — your **digital twin CEO** who's basically reading all the Slack messages, the documents, and Google Drive. He knows everything about the contracts that are all there. He looked at this; he has access to the CRM, so he can see anything in the dashboards, etc. Alright, so now: the first few weeks, what are you doing? Is it like, "I trust these five people; I'm going to bounce the idea off them"? Do you make a deck because that clarifies your thinking? Do you just go build a prototype? Do you go look at the competition? That's a...

Immad Akhund

**"Great question."**

Shaan Puri

"Where—where do you go?"</FormattedResponse>

Immad Akhund

With Mercury, I had this moment where we had raised the money and were doing the thing. I thought, "Okay, let me go talk to a bunch of potential companies that could use this, create a big pipeline, and also just learn what it's about." I talked to basically 100 companies. I remember getting kind of depressed after talking to them because most of them were lukewarm. It was only two of the 100 companies that were like, "Oh yeah, I would fucking use this—give it to me right now." That was it. It had taken us a year and a half to build this. This was about one year in, so we were kind of running out of momentum anyway because it was taking so long to build something. Then you go talk to 100 people and everyone's lukewarm. I was like, "What am I doing?" But I thought, I would really want this, so I just powered through it. It turned out that if *2%* of people want something and the market's big enough, that's probably big enough for an early adopter base.

Shaan Puri

Right.

Immad Akhund

But if you think about it, talking to 100 people and having only two people get excited about something is actually kind of tricky. Maybe I wouldn't have talked to those two people because, if I had talked to 50 people, I might have missed out on it completely. So I think it's surprisingly hard to vet an idea against humans. People just don't know, and it's hard to imagine an idea well. What I would do is give that talk, then go talk to some VCs and a bunch of users to get a sense of the reception. After that, I'd try to spend a lot of time thinking. One of the things that gets me excited about **Mercury** is that last year we launched a bunch of new things: invoicing, bill pay, reimbursements, and personal banking. Some ideas hit a dead end over time — you do the idea, and five years later you're just adding small features and selling it more. Other ideas fork a thousand ways, and Mercury is one of those. There are two reasons for that. First, there are so few people in banking, so a lot of these spaces are unexplored. Second, banking is this kind of platform: it has all your money and your finance team, so you can build a lot of things on top of it. So I'd want to think about, if I did the **"digital twin"** idea, what are the forks? Why does this become more interesting the more users you get on it? Maybe it's not just the CEO twin — everyone at the company has a twin. Maybe some of those people don't even exist at the company; maybe you have a comms person at the company that's literally just AI. I'm just riffing here, but maybe your AI twin shows up at Zoom calls. Maybe it becomes your coach and helps you through difficult times. I think that's an idea that could fork a lot of ways because it learns a lot about you: it's in your communication channels and accumulates knowledge over time. So yeah, it sounds like a pretty good idea that "we're repping on."

Shaan Puri

I was getting excited. Yeah. Well, with **Mercury**, it seems like—this may be too simple—but you'd been a founder for ten-plus years, and you realized you didn't personally have a banking product you loved. Every startup needs a banking product, so you wanted to build one for startups. There wasn't anything out there that people truly loved. Maybe they don't hate their product, or they don't spend much time thinking about it, but they don't have one they love. You basically built the one you would want. Your bet was that there would be a lot of other founders who think like you and would like what you like. Is that fair, or is that oversimplifying it? </FormattedResponse>

Immad Akhund

No, no — that's 100% fair. I think, generally speaking, I'm a little unusual. If I have to call someone to do something, I get so annoyed about it. Yes, same. I want to press a button. I'll delay for weeks for sure. But if I have to call, uhm... that's just the way banks are — except for **Mercury**. With most banks, you have to call to get a wire done. I don't want to... I guess they're kind of gone now, so I can probably bitch about them a little. But with **First Republic**, if you wanted to send a wire, you'd first try to do it and you'd hit a limit. So you call them, raise the limit, and then you do it. Then they'd go to the next phase and they'd call you and you'd have to authorize it. Then someone else would have to call afterwards — it was like a multi‑week process just to figure out how to send wires. So I was like, "This is what..."

Shaan Puri

So, you guys started sponsoring me, and I was like, "Alright, cool. I have an idea for the ad read."

Immad Akhund

Because I...

Shaan Puri

Got six Mercury codes myself, so I was like, "Trust me — I already know. You don't need to give me the talking points; I know why I use it." I was like, "I know you have all these fancy features, but for me, as a founder, I move at a certain pace." As a founder, that's one of the only advantages you have. You don't have the biggest team, you don't have the most money, you don't have a lot of things. You don't have the most experience, and you don't have a brand in the market that everybody knows. But your pace is the one thing you have that all the big companies don't. The problem is, if you start hitting walls because you can't move fast, that becomes very problematic. I like to have an idea and buy a domain on GoDaddy in two seconds. If I have a trademark, I like to be able to file it online. I want to talk to a lawyer and schedule a call. If I want to start a bank account, those little things — they sound like one-offs, but they add up and they create momentum. Momentum begets momentum. So I will always default to using any tool that can operate at *founder speed*. With Mercury, it was like, "Oh cool — I can just type. I don't have to talk to any human being, I don't have to call anybody, I don't have to drive and park somewhere." I can get my bank account open, get money in, and wire money. I used to wire for my e-commerce business at Wells Fargo, and every single month — just to pay our suppliers — I'd have to drive to a Wells Fargo bank.

Immad Akhund

Oh man.

Shaan Puri

Make an appointment, which I always forget to do. I go inside, and they gave me this *freaking*—*freaking*...

Immad Akhund

Hour to do it, yeah.

Shaan Puri

They gave me this thing like it's the *"nuclear codes"*, and I had to have that. Then I had to sit there an hour while she asked me questions, and I'd have to make sure the numbers were right. I was like, "dude, I wish I could just copy-paste, push a button, and be done with this." Then I had to get my limits raised. They had to call another person to raise the limits because they would only let me do $50k or $100k at a time. It was super annoying. It's amazing that literally just feeling like, "okay, these people get me and they'll build a product that moves at my *speed*," was so important. I also then took that to hiring. My best hiring heuristic became: did this person speed up the pace, keep pace, or slow down pace? Anybody who slows down pace for me, even if they have other strengths, rarely works out. If we have a discussion or an idea and you're like, "cool, I'll get to you next week," it's like, "wait, wait, wait — we said this was important."

Immad Akhund

Yeah, let's just get it up.

Shaan Puri

There's still eight hours left in the day. What are you talking about? How long could that possibly take you? It's not that they won't do it — it's just that if their default expectation, their default *clock speed*, is slow, it's not going to work. I want my software *fast* and my employees *fast*.

Immad Akhund

Yeah. So actually, once you describe that process, it sounds awful. I think people just get used to being abused by the current products, right? I think, in some ways, because I came from the UK, I had a slightly better expectation of banking. In the UK you can move money instantly, and the software was better for various reasons. So I think there's some element of the fact that I had this *outsider perspective* as an immigrant. I do think that's part of it — that people in America were okay with, once a month, having to drive to Wells Fargo to do something basic and having this... </FormattedResponse>

Shaan Puri

Like... beg them to let me do something with my own money. Yeah, exactly—convince them, and sometimes they say no. Yeah, dude. The term **"abuse"** is so good. It's like that's going to be a new filter for me on ideas: *"Can I describe the current abuse that the product is doing?"* I was in this conversation the other day and it was super... I went to that Dialogue conference—don't know if you've ever been, but, like...

Immad Akhund

Oh, yeah. Look at this. This is the *dialogue couple*.

Shaan Puri

Oh, perfect, perfect. So I went to [an event] like two days ago, and in the room was the CEO of Renaissance Technologies—like Renaissance, one of the best investors in the world—the CEO of Bridgewater. It was basically the who's who. In the "money conversation" there was a group of ten people. The topic of crypto came up and somebody who was a dean of a very famous university was basically arguing, "Crypto... I'm not convinced." I just sort of— it was the *abuse* thing. So I was like, "So you—you want to park your money in something like the dollar, right? Cool. Got it. I understand a bunch of reasons why." But the abuse that was normalized was: the dollar has a whole arm, the Fed, whose job is to have a *2–3%* target inflation. "Oh, oops, sometimes we go way above that, but we'll try to get back to 2–3%." I was like, "So your default expectation is that in thirty years your purchasing power is cut in half? That's like the accepted financial abuse of saving all your money in U.S. dollars—every thirty years your purchasing power gets cut in half."

Immad Akhund

It's a lot worse than that because of *asset appreciation*, right? People don't think about it, but houses and many other things have appreciated well above inflation.

Shaan Puri

Inflation — even measured in a way that doesn't include your house or your rent — the kind of screwed-up version of inflation... even that, the adjusted EBITDA of inflation, right, where it's like this specific basket is supposed to be 3%. So it's a system where **savers get penalized**. Anybody who saves gets penalized in the **dollar system**. It's a blind spot. People really don't — they've accepted that abuse. They're so used to it that it's just taken as fine. It's not even a problem; it's not even a known problem for them.

Immad Akhund

Yeah. I mean, I think part of the reason it's not a problem for richer people is because they get the **asset appreciation**, right? People aren't sitting on cash. Unfortunately, a lot of the abuse actually happens to poorer businesses and poorer people, who have to sit in cash because they don't have assets — they're living paycheck to paycheck and most of their money is not in assets. I think that part is particularly sad. I don't necessarily think **crypto** [cryptocurrencies] is the only answer there. You could buy gold; you could buy stocks. Sure, Bitcoin has obviously done particularly well in the last 10 years, but yeah — it's definitely something where there's something a bit weird about the system for sure.

Shaan Puri

Yeah, exactly. You can debate the solutions, but the problem seems... not debatable. It's just accepted — it's an **accepted abuse**. Tell me: you guys had this crazy situation where **Silicon Valley Bank** went through this crisis. Where were you when that happened? And then, what did the next — I don't know — 24–48 hours look like? What happened to **Mercury** during that time? Because it seems like you were a big beneficiary of that situation in terms of customers coming to you or adoption.

Immad Akhund

Yeah, I mean, to set the scene: **SVB** has been around one year longer than I was alive — it started in 1983. So I always remember that. When I was working on **Mercury**, **SVB** was part of the seed deck. It was presented as "the 800-pound gorilla," the bank everyone in Silicon Valley used at that time. They were worth $16,000,000,000, and I think at peak COVID they were worth $40,000,000,000. So this was a big company. When this started happening, we'd obviously been competing against each other, but it was kind of a weirdly indirect competition in the sense that **SVB** just offers something very different. It was the stolid Silicon Valley Bank that's been around forever, and **Mercury** was the upstart fintech. We used **SVB** in my previous company. I liked those guys — they were friendly and they cared about startups. Their software wasn't great, but none of the banks had great software. To be clear, I didn't feel anything bad about them, and I felt really bad for the people there. I thought they'd be fine. Then Thursday rolled along, and people started freaking out and panicking about things on the internet.

Shaan Puri

Did you find out about it at the same time as everyone else, or did you have some *inkling* or some *foresight* that it might happen? </FormattedResponse>

Immad Akhund

In December — so most of the bank run on **SVB** happened in March 2023. In December 2022 there was an article about the fact that **MBS** — mortgage-backed securities — were massively underwater. Some VCs had freaked out and pulled their portfolio companies out. I had seen some of those rumors in December 2022, but people were calling it a crisis and freaking out — like "the sky is falling." Our default is to ignore that kind of noise. Then on Wednesday it kind of kicked off. We started seeing people panic and asking, "How quickly can I get a Mercury bank account?" Normally, people don't want to switch banking services. A lot of Mercury's customers are *day one* startups: you just start a new business, you don't have a bank account, you need something, and we're there for you. It's much harder for us to get people to switch from an existing bank. So yeah — we were obviously reaching out to every startup in Silicon Valley with our sales team and other marketing efforts.

Shaan Puri

Did you have, like, a kind of emergency meeting — like, "Hey, here's the game plan" — or what? How did you organize that to sprint?

Immad Akhund

My initial position was: "Hey, you know, **we do not say a word of this publicly**." We didn't want to jump on their misfortune. But anyone that comes to Mercury — let's support them. We did a bunch of things. We changed the product. So, you know, we're like, "If you're a non‑SVB customer, we can always onboard you next week, right?"

Sam Parr

Right.

Immad Akhund

But if you're coming from **SVB**, let's make it... let's do it in half. So, and so we met — you know, we had this *flat connection thing*, and we put that right at the top. We said above it: "SVB customer? Connect your SVB and we'll prioritize you." SVB customers were really easy to prioritize because they tended to have more money. They're already vetted by a bank; they're not new customers. We basically tried to think of all the ways we could speed up the process for them, and we really focused on providing that process.

Shaan Puri

Was it like 2 × 5 × 10 × normal? Like... what—what—what did you see?

Immad Akhund

It's so—right now, **Mercury** has about 200,000 customers, and we had maybe 100,000 back then. **SVB** had a lot of money, but not an insane number of customers. I'd say they had around 30,000 or 40,000 total customers. So the volume was big in terms of dollars, but not necessarily in terms of numbers; it was significant. I think what we published is that around 8,000 customers moved over to us in a two-week period. That's big, but not crazy compared to the normal monthly growth we get. The other thing we did: everyone was saying to us, "Hey, SVB failed — why won't Mercury fail?" I would tell people verbally, "Hey, we're not a bank ourselves; we work with partner banks, and your money actually has **FDIC** insurance." At that time we had $1,000,000 in FDIC insurance. Then I would say, "Okay, if you have more than $1,000,000, put the rest of it in U.S. government T‑bill mutual funds," which we have as part of the platform. I would repeat this over and over to people. But obviously, if Ahmad says, "Hey, your money's safe," that's not very fulfilling. **SVB** was also saying, "Your money's safe." Yeah — what we did...

Shaan Puri

"Over that bank been freed."</FormattedResponse>

Immad Akhund

Yeah, exactly. So I was like, that's the one thing I don't want to say online. So what we did over that weekend is we built this product. It was called **Mercury Vault**, and it would basically visualize where your money is. We'd say, "Okay, this much of it is in FDIC-insured accounts up to this much." We also actually worked with our partner banks to extend the FDIC insurance. We went from $1,000,000 to $5,000,000 over that week. So if you had $6,000,000 in your Mercury checking account, we'd say $1,000,000 of that is not FDIC-insured; $5,000,000 is. Then we'd help you set up your treasury system so you'd put in U.S. government T-bills. So we built this product, and that really worked. It's so powerful to... so you

Shaan Puri

"Used the *product* to build *trust* instead of words."

Immad Akhund

Exactly, yes. And it also spoke to the moment. Instead of hiding away from it, we weren't saying, "Let's not talk about whether your money's safe or not." We were like, "Okay—let's show you." You don't have to trust Emile's word or some marketing; here's a product that shows you. It wasn't something completely new. This is basically my general view on things: if you're saying something over and over, if customer support is doing something over and over, how do you make that part of the product so people don't have to ask? It's just so much more powerful—it feels tangible to people. I also felt people reacted like, "Okay, instead of us going to another bank that could also get a run, we're going to something where you get this extra **FDIC insurance**." Mercury is a fintech; it's not a bank. We are not holding your money. The money never touches Mercury's bank account—it goes to our partner banks, and they have an **FDIC sweep network**. That was really important to the people we were speaking with. Product-wise, I think entrepreneurs appreciate that. The fact that we spent the weekend building a product for them that actually answered their biggest question made people go, "Oh shit—Mercury gets it." I think that was a really powerful moment for us.

Shaan Puri

Yeah, that's dope. Have...

Immad Akhund

This also—I actually felt good as an entrepreneur, because I think sometimes, when there's a crisis moment, you can feel a little bit... like, "Oh man, what am I supposed to do here?" Obviously, you can talk to customers and do other things, but **building product** is what we are about. As an entrepreneur I was like, "Okay, let's spend the weekend—everyone, let's go do this." It gave us something really tangible to do. Actually, the whole team thinks—even, I guess, two years later—the whole team is like, "Hey, we worked that whole weekend; we built this product together." Everyone's still excited about that moment because it did the thing we were good at: solving that problem.

Shaan Puri

Yeah—yeah, you got the adrenaline rush, but you stayed in character. You didn't go out of character and try to do something you don't normally do. You have these philosophies that you sent over. One of them—maybe related to what we just talked about—is *"go all in on asymmetric upside bets."* Can you unpack that?

Immad Akhund

I feel like a lot of people kind of try entrepreneurship or do things, but they don't go **all in** on them. When I did my first startup for seven months in the UK, it was a moment for me where I realized, *this is what life is about for me*. I did college — computer science — but I was never that excited about it. I just did it because I was good at it and it was kind of fun. Then I worked a job and I hated it. I worked at Bloomberg for a year and I was worried that that's what life was: having this grind. One of my friends said, "Hey, let's do this startup thing," and I said, "Okay, let's try it out." That was the first time I did something and thought, "Oh shit, this is so much fun." That startup went nowhere — it had no users — but just the idea of working 9:00 a.m. to midnight, building something myself and launching it... there was something about it that spoke to my core. I thought, if I want to do this, where can I do it best? At the time it was San Francisco — I would still say San Francisco — and I felt I had to be there to do it. I had basically no money. I packed up my stuff. I got into Y Combinator with my second startup and moved to San Francisco. Even in those seven months in London, I was all in: coding and building all day, then going to every single event possible. Honestly, half of them — maybe more — were a waste of time, but if you work really hard and do all the things, you eventually build a network. That's actually how I came to San Francisco: I met someone through extreme networking. They'd already gotten into Y Combinator and needed a technical cofounder, and I wanted to go to San Francisco, so I joined them. I feel like a lot of people half-do things. They're like, "Oh, it's my side gig," and they don't fully commit. I had no money when I quit my job to do my first startup, but that forceful grind — being all in — pays dividends in a way half-doing things doesn't. Over time, going all in applied to other parts of my life too. I went all in on family: I met my wife three weeks before I moved to the U.S., we had a long-distance relationship, then we married, and we had a kid two years later. I was still in the startup grind, but I was all in on building a family. Going all in on things has a big upside that half-doing doesn't. The asymmetric part is: the worst thing that could happen is my startup didn't work and I'd have to get another job. So I was always like, "Okay, that's fine."

Immad Akhund

I was getting paid like £40,000 a year or something, so I was like, okay. You know, I'm losing £40,000 a year in salary—maybe £50,000 if I get a raise or something. But the *asymmetric bet* is, you know, I could make a million or whatever. It's also an asymmetric life bet: my downside was that I'd have a shitty time and not enjoy my life, and my upside was that I'd have a great life—I'd be extremely motivated and excited about it. So that, to me, is extremely asymmetric: doing things where you're extremely excited and having fun is a rare gift. I think a lot of people don't do things they enjoy every day. I have some friends like that, and it's kind of depressing talking to them. Having the ability to do things you really enjoy in life is such an important thing. If you find something like that, you should go all in on it. </FormattedResponse>

Shaan Puri

I love what you just said because I wrote "going all in" right as a reminder. I have these note cards, which are usually just reminders. The best information is not brand‑new information — it's just true information that I needed to hear again, needed to bring to the front of mind, or needed to remember in my current context. I think the way you described it actually made it click for me in a new way: **going all in is a skill**. I did the same thing as you. I had a startup straight out of school. I had this idea and was having all this fun with my friends. I was either going to go to med school, as I had planned — I took the MCATs and all that — or I was going to do the startup. I decided I was going to do the startup thing. Then I felt this gravitational pull. I got a job offer: someone was going to pay me $120,000 and I had to move to Indonesia. It was like this guaranteed salary… a safe option. I remember my co‑founders — my two buddies — and we told ourselves this stupid story. We're always telling ourselves stories. I told them, "I'm going, dude. The lessons I'm going to learn there are going to be so useful for us." I didn't have the courage to just say, "Yeah, I'm going there because I really want this guaranteed, kind of safe salary." Instead we told ourselves that I needed to learn business over there before I could go do business here. Within two months when I was there I realized it was a mistake. I needed to be all in. I quit my job. I told the owner, "Hey, I'm out of here." He was like, "What? It's only been two months." I told him I needed to see this through — otherwise I'd always wonder. I wasn't lit up every day. I had more fun over there, and there were no guarantees, but I felt the guarantee of being lit up every day would be better for me in the long run. So I flew back and I was like, "Alright, I'm all in," and we moved. The first question was: is this the best place to build this company? If not, why are we building it here just because we're already here? That seemed like a terrible reason. So we immediately moved to the hub of where that type of company was built. It was a restaurant chain, and most of the top fast‑casual restaurant chains have been launched in Colorado — Chipotle, Smashburger, Quiznos, Noodles & Company. It's like the Silicon Valley of, you know, crappy fast food. So we went all in there. Nothing really came of that startup. I lost money and arguably a year of time working on something that didn't work. But I absolutely didn't lose. Not only did I set myself on a different trajectory, but more importantly I learned the muscle of going all in. Fast forward five years: you look back at your same smart friends from school who have been in a bit of a corporate life that whole time, and you see that muscle has atrophied. They don't have that muscle, and that muscle is really important — because whatever you're going to do, the winning idea, you're going to need to be able to go all in to make that idea work.

Immad Akhund

Yeah, **100%**.

Shaan Puri

Alright — the last, last philosophy I want to hear from you, because I *dig* these: **"There are no rules — just construct your work and your life the way you want."** What do you mean by that? Why — what makes you say that?

Immad Akhund

You know, I feel like when you're in **Silicon Valley** everyone is trying to be influencers or VCs. Everyone's got ideas like, "Hey..." I remember in 2011 when I had my first kid: it was very unfashionable to have a kid and run a startup. People were literally like, "What are you doing? You're going to fail your startup," or whatever. Actually, one thing that's a complete contradiction is that the **only way you succeed is by doing things that are unusual**. If you're doing the normal things, how can you succeed? Everyone's doing those things. Weirdly, if you look at every single success, it's a contradiction—because it had to be. A really common maxim in Silicon Valley is: "Don't do consumer startups." But the biggest startups—Amazon, Facebook, Google—are consumer startups. Those are the $3,000,000,000,000 companies. So you have to understand where most advice comes from and why it exists, and whether you are the exception. The only way you succeed is by being an exception. You have to ignore the rules and just do the thing. That said, there is useful knowledge out there and reasons people give this advice. You should understand why you are different and what the problems with consumer startups are—there are lots of them. Generally, the best things I've done have been by ignoring the rules and doing things differently. Most of the time, if you deliberately pick a set of rules to ignore because you have a strong reason to ignore them, you're much more likely to be successful.

Shaan Puri

That's a great— I forgot what the phrase is, but it's like **"experts know all the rules and masters know when to break them."** So, you know, that's sort of the best signal of mastery: when you know the rules and intentionally choose which ones you're going to break. I'm learning the piano right now, and my teacher will be like, "Oh yeah, this is Bach," or "this is what Beethoven did," and he'll specifically call out that he did this thing that, you know, normally you wouldn't do. Right? This doesn't make sense—these are two different keys, or it's not part of the scale—but it works because of this. And it's like... and they make...

Immad Akhund

Up — the reason why it works.

Shaan Puri

Yeah — they make up the reason, but it's post-fact, right? Exactly. It becomes a new rule later. It's like, "Oh yeah, but if it resolves tension then it works." Okay, cool. But at the time, this sort of breaks that rule. And that's what all the best ones do. What I like about what you just said is it made me think it's worth writing down. I'm going to try this after this podcast. It's worth writing down not just how I want to win or how I want to live my life, but specifically what rules — what common best practices or generally accepted good practices — I am intentionally choosing to break. I'll give you an example. The other day I was having a conversation with a guy who's really smart and wise, and he said this great line. Everybody nodded and took notes. He said: > "Either our friendship is sacred or the money is sacred, but you can't have both." I said, "So does that mean you just never do business with friends?" He goes, "Never." And everyone's like, "Yeah, that makes sense — it could be really messy."

Immad Akhund

Yeah, and I...

Shaan Puri

Go... I don't know. I think that everything you said is true, and yet **I choose that mess** because I think there's a great upside in finding the people you love and doing life with them. Life is a lot more fun when I do projects with friends or invest in people's companies. I bet on people. Sometimes it doesn't work out, and sometimes it gets a little hairy, but I choose this mess — it's a rule I want to break, and I'm comfortable living with that. I can always go back and change my approach, but for now that's a rule I'd want to break.

Immad Akhund

But that's a good example where you want to understand how to *break the rule* as well, right? It's not just, "I'm going to break the rules; therefore I don't care about it." You want to go, "Okay, if I'm putting money in a startup, I'm going to make sure there's a contract and everyone understands how we work together." I do think when you break the rule, you have to go out of your way to mitigate the potential downsides of doing so. Breaking the rule is normally the harder thing, so you have to understand *why* and *how*... </FormattedResponse>

Shaan Puri

The rule you broke was: as a startup founder in SF [San Francisco], with all the options to choose from, you chose the **highly regulated, highly compliance-based** route — the hard path that requires building for a year and a half before you launch. That’s not the *quick-and-dirty prototype* vibe; you can't use that playbook there. So you decided to break that norm and go this way because that's what you wanted.

Immad Akhund

Yeah, yeah, but I spent a lot of time understanding what I was getting in for. I spent three months just talking to people and researching compliance, rules, and laws — all this stuff. *I wasn't doing it blindly.*

Shaan Puri

Yeah, alright. So, **Ahmad**—where should people find you, and who should reach out to you? Who do you want reaching out from things like this? When you're out there, you attract certain people. I like to say you want to have your **API**: tell people who should connect with you and on what basis. Give me a sense of that—where should people find you, and what should they reach you about?

Immad Akhund

The easiest way is on **X** (Twitter, as it used to be). I'm just Imad. Who should reach out? I'm always willing to try to be helpful to entrepreneurs. If you have an idea, you want to pitch me something, or whatever—reach out. I try to be quick and say "yes" or "no" if I'm interested or not. And, obviously, if you're interested in **Mercury**, go to **mercury.com** and sign.

Shaan Puri

Up. Alright, love it. Thanks for coming on, dude. </FormattedResponse>

Immad Akhund

Yeah, thanks for having me, Sean.