#2 - Tobi Lütke & Kaz Nejatian on Shopify's Country-Sized Economy

Balaji Srinivasan 00:00

This is the network state podcast. And I'm here today with Toby Lutke. And Kaz Nejatian, the CEO and CEO of Shopify, one of Canada's largest companies, Shopify is internal economy. Secondly, on the scale of a small country, on par with Greece or New Zealand, depending on how you calculate, they have solved many different kinds of management and technical issues in order to build a company that has such a scale. And we also get into what kind of small country they would build if they had the opportunity. Along the way, you'll learn a bunch of things, including how Shopify is building something that's like a new kind of Hanseatic League, as well as what the Hanseatic League actually was that kind of thing interests you. Let's get started. Thanks. Thanks for coming on. This is the second episode of the network state podcast, this new podcast on everything from, you know, obviously starting new countries, but all the people like yourselves who have built startups that are on the scale of small countries and protect policymakers and everything that's related to kind of building the future. And my first guest was metallic, a fellow Canadian founder, and you guys are my second. So you're basically the kind of people who've literally built giant economies. Okay. And, you know, the goal here is something where, you know, there's, there's some good tech podcasts, but actually, those tech podcasts are often, you know, they're, they're focused on the business of tech and so on. And that's fine. And that's good, you know, prices are important or what have you. But price has always been secondary to why I've been in this space since being secondary thing to you guys, like, you know, we're doing it for, for building something of meaning. And so that's kind of what I wanted to talk about today. So that is sort of preamble. Of course, I can ask you guys about your background and all that type of stuff. We can kind of like, hook it to that vector of, you know, we've built giant, you know, companies and giant communities. And what's next is the end of the internet just coins? Or do we have one more level? Can we get to cities or even new countries and so on? Right? And for, you know, everybody, you know, is probably heard of Shopify at this point. Toby, you're the CEO Kaz, you're the relatively new COO, right. You've been with the company for a while, but you're just recently became CEO, I might be remembering that right?

Kaz Nejatian 02:08

Yeah, I'm five months into the job. Well, but you've been

Balaji Srinivasan 02:13

with Shopify for a while. Yeah. But for years, one thing I just kind of wanted to do for folks, people have heard of Shopify now it was Canada's largest company, it is still one of Canada's largest companies, you've built an economy that's on the scale of a small country, and I want to talk about today is what it means to manage something like that. And could we go one step further to actually maybe even starting a small country or a small city? And what would the next step be? And what have you learned from what you've built? And so just like, you know, the scale of the Shopify economy, there's millions of merchants billions of dollars built from scratch. Their stats in your SEC filings are stats on your website, millions of merchants, you have your merchants have millions of employees? Do you want to just some stats that you want to rattle off?

Tobi Lütke 02:54

Yeah, I mean, I mean, it's funny to think about Shopify as, like a small economy, it's, this is definitely sort of a mindset you should take and, like be, should be, should explore that angle, obviously, like, it's, it's funny to think about this, because at some point, I, you know, it's now like 19 years ago, I wrote my first line of code on the Ruby project. And then at Lago way, it's like, collected, you know, people who joined the team, and then eventually it got a ticker symbol. And now it has obvious, like, various institutions along along the way for millions of merchants, partaking in the global network of commerce and global economics and says, there's also internal economics around, you know, if there's a theme store, there's an app store inside of Shopify, there's like a developer community building against that. And increasingly, I think it's a really, really good way to look at everything, because it's a big, it's a large exercise in, you know, creating incentives for people to you know, like, harness people's enlightened self interest to for the betterment of the entire community. And so, so there's definitely echoes of economics and like, in the company, but yeah, like it's, it's hundreds of millions of people buy from Shopify stores every year. It has, over view of estimating around 5 million people are employed by the app developers by the stores themselves are good jobs. One thing that almost everyone forgets in the context of economics and is that the vast majority of people in the world work for small and medium businesses. It's depends on the country, but it's usually it's always above 50%. It's usually as high as 80 90%. And that totally depends on the liquidity by which new businesses are created because new small businesses tend to go out of business for often for no fault of our own, like, like, like circumstances with the people who run the businesses for family changes might be, you know, a tragic death, which prevents any kind of continuity of a business but was created. Because it's so much so much we small businesses are built around their founders in most ways. It's a leaky bucket. And I think vibrancy of almost every economy depends on how quickly this leaky bucket is being refilled and standard of living tends to go up if, if a bucket is filled faster than you know, it leaks out. This is what we care a great deal about, right? Like we think that the world especially with word of entrepreneurship, is this sort of meta emergent phenomenon that just exists partly because some people just cannot work for other people, there's less probably some kind of gene that prevents people like that gives people 40 problems, but that bet necessitate them reaching for independent themselves. This has been true for all of human history. Sometimes it becomes a little loop, it becomes illegal in places in both places don't last. So it's like the fate of entrepreneurs is the it's a bit of a Where's Waldo story throughout human history, it tends to be not recognized or not talked about a lot. But like it's, it always plays a very important role. And so, you know, if I think of a chip, if I was trying to do like, a sort of somewhat HIePRO way, is the belief that we the entrepreneurship and economy is actually relates a lot more to the friction of starting businesses than the policies. Because there's the demand for doing it and wanting to do it is constant opportunity is variable. But we're living in a time of enormous opportunity around the internet. And the thing that's governing supply and demand, they're like, Oh, the thing that's preventing business from succeeding is often just visited Byzantine, almost Kafka esque complexity of starting these businesses, or the set of requirements that the medium dictates like, for instance, I started my snowboard store, which which worked well for me, because I could program and I like 20 years ago, and then also the internet and so on, and therefore I could overcome the hurdles in my way. But I don't think we want an internet where the only people who can start new businesses in retail space, other people also happen to be programmers. And so Shopify is trying to like less learning, like the learning curve, just just cause more definitive proof is,

Kaz Nejatian 07:41

is it to the interesting perils of countries biology, which is like Toby and I are both immigrants. And the way I think of Shopify entrepreneurs is that they're immigrants from big companies, to people who left big companies saying, I can't work here, let me start something on my own. Which is like an interesting story of like, my family, like, can't stay here anymore. Gotta go somewhere else. And if you add the collective revenue of Shopify merchants, that's $444 billion last year, as the second largest company in the world, their collective revenue. And it's interesting thing about it becomes like our job is to make that integration pattern easier. Like if you think of like people who left on the Mayflower come to Mayflower like it was a really long journey. Some came, but not a lot. And a pattern of immigration got faster and faster, so people could emigrate out of things they didn't like, into things they did like, because just travel got easier. And our job at Shopify is to make that journey from a job you don't like, to a job you love may be difficult, but you'd love it easier. Well,

Balaji Srinivasan 08:46

it's really interesting, because, um, something I've thought about and you are maybe the experts in is a question I've always wanted to ask is, you know, there's this concept of a so called lifestyle business, right? Where, you know, it's not really swinging for the fences, you're just kind of having fun running your coffee shop, but anybody who's actually run a small business finds, it's not a lifestyle business, it's your whole life, you know, right. It's as hard as running like a big company, but it's harder, because you often can't get venture capital, you don't have the support of some, you know, franchise, like all the Starbucks franchisees, they don't have to deal with branding, they don't deal with marketing, their products are set for them. They have to negotiate with vendors. Basically, they have some buffer, because you've got this giant chain that is backing them, right. What I saw, especially during COVID, with lock downs, and so on is many of these small businesses just went under, for example, in in San Francisco, even even in normal times, like there's some ice cream shop that spent like $200,000 trying to just get through the regulations and then just stopped and other made people have 200k in disposable capital. And to put that into just regulations on before you've sold your first ice cream cone. He's actually kind of bananas. I thought about that. Because on the one hand, as you say, you know, Tobia and Cazal is that you know people are wanting to do this, and we depend upon On your businesses and small businesses, not just everything being scaled on their hand, it's really hard to do that, that pushes people I see this in medicine also, you know, the independent practitioner, there's so much regulatory complexity that they're pushed into being an employee of a large hospital chain, right, or maybe they go and work in pharma or something like that. So between that thesis of like, I want to start a small business or run a small business, and the antithesis of all this regulation is becoming such a, you know, frictional cost, as you said, maybe Shopify is the synthesis, where you are like a, like a centralized hub of this hub and spoke network. And you provide, you know, currently, right, obviously, inventory and logistics and pricing, more recent, you've done banking and taxes, maybe eventually you do compliance as a service. And you file all these forms for people in different jurisdictions and keep them compliant. So they can just focus on customer service. And so now, they're not wholly, you know, just on their own, but they're also not like fully dependent, you know, if they, if they really want you to get unbuckle from you guys and go somewhere else, you wouldn't want that. But you know, they could. And the other thing is that like the Shopify community, just like founders tend to, you know, tech companies tend to have similar things and stick together, it seems like a lot of your merchants would have a lot in common, and like organizing them would be really pretty, pretty interesting. So like, citizens of Shopify, or what have you, let me pause there, get your thoughts on that.

Tobi Lütke 11:16

I think that you touch on, like many things here, but maybe one thing to like, just like underline, partly because it's sort of one of my pet topics is just just a concept of lifestyle business. Like the term lifestyle business is like a construct of Silicon Valley. Right? Yeah. Like, it's actually like, pretty mean, because it's, it really is, like, I don't know how we renamed businesses to lifestyle businesses, just so we could call ventures businesses, and how this succeeded.

Balaji Srinivasan 11:52

It didn't come out of like rails or whatever. They were like, Oh, your business doesn't have to, like, like, I thought, wasn't it wasn't something out of Basecamp? Did they come up with the term? I don't think it actually originated from Silicon Valley. Am I wrong about

Tobi Lütke 12:02

that? It just saying, like, my friends, offensive secrets might have gotten done it to us. And okay, so maybe without laying blame at anyone's doorstep, it is a pretty funny thing, because like, you know, like, businesses were around for a long time. And they didn't call them lifestyle businesses, if anything, like if you're looking at sort of a techie path of businesses, which is like, you, you start a business by an application to Y Combinator, you get obviously teaching, and then afterwards, people give you money. But that sounds a lot more like silver spoon businesses, to me. And like, what, what entrepreneurs and SMEs do, because that's a lot more hardcore. So, I mean, this is the greatest thing about, like, I think you have a greater privilege we have Shopify specifically is that we, you know, like, I talked to hundreds of our customers through instant message and so on. And our customers are entrepreneurs, right? Like this. Everyone who knows founders of companies who build things knows how like, like, very exciting people, like hitter hitter dogs fingers, who just don't take no for an answer and like, like, make things happen. And it's really, really great. You know, I think everyone ends up becoming sort of the average of the five people they spent the most time with, and like, if you can Lord that number up to as many entrepreneurs and founders as you can, then that goes really, really well for you. And I think it inspires everyone. And so it's interesting, because when Shopify started, it was extreme like, like, I actually like first version of after launch had in the admin interface, because back then, like, this is 2006, right? Like, this is like a stone age. So free internet. Or at least first off as a service would, like no one knew what is what you put into an admin interface and what you don't. So if you actually had to, like forms, right, built into the admin, like this was one of the main tabs and Shopify, and no one used it. And in fact, everyone was like, why is this fair? Clearly, I don't want to talk with my competitors. So you know, sort of at the beginning, there was a very clear everyone was thinking about food of retail as a zero sum game, and as a rival grocery are all competing for the same dollars. I think one thing that has happened is that I mean, the physical world is permeated with zero sum thinking because it sort of makes sense and in a lot of ways, like, like only one business can be in this particular place in this particular street, right. Like, there is a lot more rivalry for finite resources. I think, like the online world has attracted positive some thinkers people build Internet businesses in In bedroom, and, you know, like, you know, the sneaker market just wasn't what it was, like 20 years ago, we just was, you know, yes, people bought sneakers when they had to, like goes play squash. And, you know, some people use fashion statement, but that market crews so tremendously just because people like created a community of people who really care, the craftsmanship changed with the fervor of aesthetics changed. And this is now a very, very large market where it wasn't before. And I think this sort of sense of, hey, there's so much more blue ocean opportunity in the world, because there's so much, so many more people who would be really into things if a supply would be there. And if a narrative and a story is there, then like it just it's it's migrated there. And now people are perfectly happy to share like, it's actually like, across Reddit across like, Twitter, communities, across Twitter, across Facebook groups, it's like, it's people are really sharing, people are actually comparing notes and people are super happy to, like help each other in any which ways we can

Balaji Srinivasan 16:10

in the SMB community or in the, or in the tech founder community in the SMB community. Sorry, I should I should say this saying is there some cross pollination from sort of the nonzero sum mentality of tech? Well, ideally nonzero sum mentality of tech into let's call it rather than lifestyle? Maybe SMB, would you? Would you think it's more neutral term? Yeah. Okay.

Kaz Nejatian 16:28

It's, I was a, I was a tech founder, whenever I was a YC. Founder, I was the type of founder Toby was talking about a second ago. And my mom's a Shopify merchant. Oh, one of the things that happens, I don't think it's ever happened transparently, when I was in YC, is Shopify merchants build applications for each other? Imagine if a startup found a growth hack and just gave it away, that would never happen. Because a growth hack is your growth hack was shuffling, which is we'll build you, we have a storage run, you build an application to help you with some part of your store. And you will literally give it away to other merchants. Because the I think they've had all our best machines have intuitively realized that the goal is to maximize the size of the DTC space, maximize the size of like direct connections to customers, which by the way, they all like this is why they all love shop, pay that magical button on the internet that makes shopping easy. It's because they can all be on the same side of the table against the regulations to ban them from having one click Checkout otherwise, right.

Balaji Srinivasan 17:40

Well, so I understand. There's two really interesting things about that. First, there's so can you explain the regulatory barrier to one click? I actually didn't know about that?

Kaz Nejatian 17:49

Well, it was a there was a patent that was patent pending one click checkout for a very long time.

Balaji Srinivasan 17:53

Right. Right. So that part I knew. So you mean patent barrier, but not basically, let's say, artificial barriers?

Kaz Nejatian 17:59

No, there's also regulations that have the card networks have regulations that require you to get audited every X month for your fault. There's like, there's banking regulations, like there's private and public regulations that make the world of entrepreneurship difficult.

Tobi Lütke 18:14

Germany actually has laws about check for checkout, like actually regulatory, like every button has to say these words in the process. This is how you have to represent like, luckily, they have a reasonably like, not a reasonably well chosen and confirmed like this in sort of a zip code of best practices.

Balaji Srinivasan 18:35

And the thing about that is it often codifies in model that becomes obsolete, like, for example, with a crypto payment, you don't need to take their address and so on. And probably it over specifies it to a point that a crypto payment would would maybe not fit into those regs, I'd have to read them. That's interesting. Yeah, I did know about the patents and the overhead of audits. And again, this is a thing where, you know, shop, pay your advertising all of that across the Shopify network. So it's affordable. It's basically like, it's actually part of the surcharge of kind of like using the platform that pays for all that collective stuff. But still, it gives them independence.

Kaz Nejatian 19:06

Like, if you think of like Shopify, we are merchants, I think for every $38 our merchants make we make one. And that's like, literally the goal is to accrue as much accrue far more benefit to the community than we take. If you think of like if you think of the community as being like, one thing as being a thing that exists independently of like, the corporate structure, where you try to do is maximize the benefit to the community. That's kind of how we think about it. Well,

Balaji Srinivasan 19:38

I mean, it is, I mean, if they're making x without Shopify, let's say you're making $20. And probably that's maybe they may not even exist without Shopify, but they're making $20 before and now they're making $39 You're happy to give you one and so that I mean, that's obviously like a good deal given everything that you guys give to them. You know what I thought I had, you know, what that sounded like, by the way is like open source for merchants. You guys have built like an open source app store for merchants, and they kind of trade away the things and it's sort of like, it's like branding, and it's kind of being known within the merchant community. Is that Is that how you should think about it? Like, what is the? Is that part of it as well as the DTC thing? Or am I off base on that?

Tobi Lütke 20:17

Yeah, no, I think, I mean, it's a spirit of sharing.

Balaji Srinivasan 20:21

I didn't I didn't mean to sound cynical it was in the spirit of sharing. I mean, like, it's kind of like, you know, you do you sort of have, you're giving back to the community, you are, you're showing, you know, how capable your firm is, in a sense, kind of, like, for all the reasons people publish open source also for those just a love of doing it and whatnot. So that seems to be an analogy. You're right, that maybe a startup might not give away their growth hack, but startups do open source and SMBs share Shopify applications. Would that be fair? Yeah. Okay, so Well, there's, there's when I think by the on the lifestyle thing, I agree, the by the way, the lifestyle term is patronizing. You know, here's kind of a cut that I've seen before on like startup versus SMB. Well, so SMB is a proven business model. It's, you know, you eating breakfast, and it's often debt financing, and the goal is not hyper growth, and it often does not involve new tech, whereas the startup is an unproven business model. It's like, tweet your breakfast, right? Haha, you know, rather than eat your breakfast, and it's more equity financing, it's high growth and it's New Tech. Go ahead.

Kaz Nejatian 21:23

Look, I think it's actually a very weird state is that and startups actually a bad assumptions of the financial system that favors risky over like the money if you actually go to a bank and you're like a startup has a massive risk and have some had some equity checks some VCs actually relatively easy to get that financing from, I mean, walk into a bunch of banks they'll give you venture debt with if you my mom, we're a Shopify merchants like the ability to understand like production capital is a heart is actually a tilt is one of the other things that makes starting a business difficult, right? Like my mom can't get a bank loan to start a small business just can't and if she did, she couldn't even grow it because banks don't understand like how those businesses grow the economics of it are weird. Whereas the established patterns for equity are there to establish patterns for for non equity financing, or these businesses have become very difficult. Which by the way was shot by capital exist i Chapa. Capital exists because we want to fund merchants not because in fact, this is this was Toby's been tokenized for smack this was how he This was his first interview question for me. He asked cuz I've read the bank. I've read the bank act of a bank charters, why don't banks let's delete lent money to small businesses to do remember this competition, Toby? Yep. Yeah. And what was your answer? I think you have to start in the 1600s. Because there's a long history of this going wrong for for small businesses, like starting from, you know, the idea of incorporation,

Tobi Lütke 23:01

cause actual answer was, so you have to go back in the 1600s. And I wrote a book on this very topic, which was my way the shortest job interview I've ever had, because we've had we've done like, like anyone was anyone who reaches back and answers next expert on esoteric topics, but it's highly relevant to the job. It's my it's part of a team. So

Balaji Srinivasan 23:20

well, that's awesome. What was what was the what's what's the proceeds of the book, it's,

Kaz Nejatian 23:23

it's honestly, it actually becomes like how incorporation started this will be very boring for everyone. But the idea of a corporation starting with the British East India Company, it was a very weird thing you buy shares in this thing. And those shares would, would give you the ability to have an annuity right, those annuities could then be bundled up and sold as equity. So that so they were much more easily modifiable in like, for people then annuities coming out of a small flower shop, for example. So banks start like building the risk models around that rather than this. So if you're a small business, the type of loan you usually get requires a personal guarantee to not actually lending you money to the business, they're lending money to Balaji and saying, Hey, you, you have a house, I'll give you money for business, which the terrible deal for small businesses right under the underwriting engine to underwrite small businesses doesn't exist, if you can fake it, but doesn't exist. And one of

Tobi Lütke 24:23

the consequences of this is like it's a super undesirable secondary order effect here, which is that like, like, I'm thinking of several of our largest missions, like like Bucha are really like 100 million revenue, billion dollar plus revenue merchants were have started on Shopify, and they have a six Shopify store but the founder started, right like, if you have to reach like to get to capital, you have to reach through the corporate barrier, and you'll have to personally guarantee the number of ones Door, which was for successful discovery of something that didn't work that taught you a lot also disables your ability from partaking into in the economy again, even though you will have no bid for skills. So I think this is this, this is really regrettable I read, you know, obviously the bank is chatter isn't like that clear on these kind of things. But like, one of the reasons why the banks have the protections that they do and the guarantees and the Federal Reserve's is because it's recognized that real lending to small businesses is a economy improving and society improving institution, yet, they don't do it. So so so we have to kind of replicate these, like institutions inside of our own mini economy. Now we can do it because again, like well, wonderful, like, like, because you're so early part of the journey of our emotions, very, very corporate, often after signing up for Shopify, so VPC, the traffic buildup is, you know, if we have inventory visibility in these kind of things, so they can underwrite people massively better than what banks could do. But this is only just business path dependent consequence of us being just more motivated to do it. Because we are doing it as a compliment to the mission rather than as a I don't know why pink? I mean, because banks have to enforce the Charter

Kaz Nejatian 26:21

rights.

Balaji Srinivasan 26:22

Mission. Not I mean,

Kaz Nejatian 26:25

look, I had three planes sales startups have failed spectacularly before the fourth one did, okay. Like, like, if I was a, if I was, if I was a founder of a business that Silicon Valley didn't understand, the first one would have bankrupted me. I wouldn't have told number two, three or four, like, and that's it. And it's a very simple Toby talks about there's a very weird capture that has happened in our large institutions, where the words they say, are not the same as the thing say to Yes, like this is honestly like the very weird thing. And if you didn't shuffle capital, which by the way, is, you know, font, overwhelmingly, people who would not get access to any loans, because or, you know, minorities, women, people would launch like, low educational pedigree, like if you think those people, it's actually not Shopify is one kind of check, but it's other Shopify merchants that are funding it, right. Because what works is Shopify merchants have paid us we take that money and put it back into the economy.

Balaji Srinivasan 27:22

Yeah, well, it's what's interesting is it might also be something where eventually I don't know where the regulations are on this, but you could have the Shopify fund, so that those merchants on platform who are successful could put some percentage into invest, you know, kind of backing your fund and investing in the next expert Shopify merchant, which is a platform they understand. I'm not sure if that's something you've thought about.

Kaz Nejatian 27:43

It's almost like you work here.

Balaji Srinivasan 27:46

Well, the thing is, so essentially, as you know, you mentioned underwriting. And I think that's like, really important that you guys and other folks, you know, when you have enough scale, you can use machine learning to do better underwriting and just pull straight margin out of that, because the bank does not know every single action that they took on the Shopify platform since the beginning, you have seen their six, you know, their five failures before their six success, so you have a track record on them. And then you can, you know, you can basically scope and bound your risk appropriately with all the signals that you now have over 10 years of operational, all the data exhaust that you collected for all these other purposes can now be repurposed for underwriting. Am I wrong about that? Would you agree with that?

Kaz Nejatian 28:29

Yeah, that's actually what Shopify is. I mean, the very pluck, is that right? Like, if you want to start a business before Toby started Shopify, the amount of money you had to spend to get the equivalent Shopify was a lot. The fact that we give you Shopify for $29 is actually an underwriting decision. Like we are underwriting like, it's, it's, it's us, essentially taking coupon on a very, very long bet.

Balaji Srinivasan 28:52

It's funny, it's kind of like, have you seen the movie, the founder, but the devils in the movie, I'm not sure if actually, it's true in real life. But supposedly, you know, the key insight for McDonald's was that they weren't actually in the retail business or the restaurant business, they were in the real estate business, because they've made the presence of McDonald's, at least at one point in history, boosted real estate values are rounded enough, that actually they made the maximum money from buying the real estate on and then having the McDonald's there. And hamburgers were actually a way of increasing the value of that real estate. Right? And so it's, it strikes me that that's actually in some ways similar to this, where — I'm not sure I can make it alliterate — but it's not simply the small companies, it is the sort of capitalization that could be, you know, the next big dog leg up, right? Because, you know, that's something that helps them helps you, you know, make something happen, but you wouldn't be able to do that without obviously the platform itself and being, you know, building everything you did over the last 1515 ish years. Like maybe that's maybe that was totally obvious. And

Tobi Lütke 29:53

you know, it's not it's just like total agreement.

Balaji Srinivasan 29:57

Essentially, what I wanted to do is move the center of opinion to within, like our community, and actually also the policymaker community, to realizing that actually, we can go from just starting new companies and communities and currencies to new cities, and maybe even new countries. And how to do that. Well, in a sense, obviously, you guys have heard the term tech policy, right. And that's like a very, that's in a sense. So it's hugely important. It's like a small bowl way of talking about network state or startup city. The reason being, because tech policy is like the blocking and tackling of, you know, for example, where the laws around the German checkout thing that you just mentioned, right? But tech equals network and policy equals state. So the ultimate level of like, tech policy is a new state where you can determine that policy from scratch. And, you know, it's the same like, you know, attraction of wow, I can do a new business from scratch, and I have to build everything, but I have root access over it. That's amazing. So that's kind of that's kind of the goal of this. And have you guys seen me read chapter one, the network seat? Have you seen the book? Should I describe it at all? Yes, no,

Kaz Nejatian 31:06

no, we've both read it. Or at least I've, I've read the first three chapters, and my wife read the whole thing. was the conversation a couple days ago.

Balaji Srinivasan 31:16

Okay, awesome. So, you know, I very much considered a v1, a work in progress, you know, just like you guys, like any, you know, entrepreneur or founder, I'm like, excuse my dust, right? Like my dead? No, there's a dent in the side of the door there. Okay, I'll fix it. But, but it's a v1, I'm working on a V two and actually like the MX eight, movie, and so on and so forth. And actually huge additions to the book where everybody's like, Oh, you dumb engineers, how are you going to build the roads? And I'm like, you know, who's going to build the roads, but engineers, it's civil engineering. Right? You know, like wastewater treatment, electrical engineering three phase, like, that's, that's what we all learned in school, like, I know how to do some of that. And, you know, a lot of the folks in our community have, you know, hard engineering backgrounds, right. And so the, the concept is, in the V to the book to, to not just flesh out and answer every single fact. And so in bit kind of have it also translate into multiple languages. It's gone viral in like Japan, and Catalonia all these places where I never expected it was pretty cool. I was like a video in Japan with like, a million views on nursing. So the bring it back to, to this podcast, basically, you know, we talked about the scale of Shopify economy, we made the point that Shopify is on par with a small country, you know, stats you mentioned, you have your merchants have 5 million employees, Shopify, a 650k. Developers, if you have even 2 million merchants and shoppers, merchant population alone is somewhere between Latvia and Slovenia, your GMV for q3, if I'm not mistaken, I saw that report is like 47 billion. So that's like, annualized at $184 billion GMV, like gross marketplace value, total value of all the transactions. And that's like, you know, if you compare GMV to GDP, it's not exactly the same, but it's in the ballpark Greece's at $203 billion, New Zealand's at $204 billion. So Shopify is economies, like on the scale of like Greece, or New Zealand, which actually pretty legit, right? And so, you know, when when you're building something like that, it's obviously more than just setting up a storefront. There's, you know, you guys now have the balance product, which is like a bank account, you got your tax product, which is doing, you know, like a, like a, I guess, if I'm saying is like a better sales, tax computation, but eventually, maybe it's like a Intuit for small business. I don't want to give away a product roadmap, but am I am I in the ballpark there?

Kaz Nejatian 33:32

I really, really like our tax products. It's a very, it's very proud of it, and it's very loved.

Balaji Srinivasan 33:40

Well, you know, what's, what's interesting with that would be you guys could since you actually understand the tax code and could put that into code. You could give popups and tooltip saying, Hey, you should do X, Y, and Z. And, you know, you can optimize your thing here with a complicated code over time, as you get on base which is sales tax optimization, then you could actually give tips to people to like, do X and do Y and do Z, which hasn't actually happened yet, but I've often wanted to see if maybe you're the right guys to do it. Totally stay tuned. Okay, fine.

Tobi Lütke 34:10

Like it's like, but like, you're right in vitro for meta learning of community is like, like, there's nothing better than like, making it so bad. I've only a few people have to figure out the complicated stuff. And when we amortize it over the entire thing, or we absorb the complexities of it, everyone else doesn't have to, I think the and this is probably like one of those places where things get really close between, you know, networked state ideas, crypto and, and Shopify, like what we want, like we think a lot about incentive design, but we love is that we are sort of, like, as a business very much on the same side of the table of with our customers, but the way I think, for businesses to accomplish, so we see it as our biggest opportunity to grow, for success of our lack of people on our platform, because we are participating in the revenue too, as we're taking a small part of the increase, but we can also provide, and so yeah, like, the most wonderful thing that we were sort of hoping for when the company started, and that has proven out over and over again, and it's reconfirmed every single time we launch, something like Tex is that the capacity that entrepreneurs have finite attention, and how that is being invested is like matters. So if they have to file their taxes and have to run through a lot of complexities, or if they have to, you know, just like build a logistics system for couple of years, or, you know, like, all these kinds of things happen in versatile, like in the success journey of every merchant, traditionally, every single time we subtract one of those by just taking the complexity of it, and just putting it into, into into Shopify, and just kind of removing that, like, the time actually gets spent on improving the product, improve your marketing, finding product market fit, and they grow faster. This is why I'm saying like the world is much more shaped by friction than by policy in the end. Now, sometimes policy causes friction. But it's very hard for, in my mind for not for for a politician to say, to launch a entrepreneur Action Plan, which causes like, put some money into some kind of thing, I think that's more that it's very hard to do anything other than just distort the playing field, it actually often causes a negative effect, because now people will be distracted by filing a bunch of application forms rather than actually improve a product. And so I think we can act as a like a like a as a as a thing against that sometimes

Kaz Nejatian 37:08

does interesting thing happens, which is like, I want to hope the tone of Shopify, I think of Shopify is like, the core platform that everyone's on. And every other service we provide is voluntary. The merchant can take it or not, they can take our payment system or not our tax system or not, which is kind of like how the government's used to work, right? Like when the US first came up with the greenbacks, like California and Oregon opted out saying we don't want this like, this doesn't work. I was like, I think it's become like a very like a because an opt in system. It's a voluntary system for merchants, they can decide for themselves if it's good for them or not. So it's become a very, I think it's actually allows us to aggregate their wishes in a way that is totally voluntary and opt in, rather than like this, like weird thing that happens elsewhere, which is like, you must take this closed thing or not.

Balaji Srinivasan 37:56

It's a new Hanseatic League. You guys, if you're ancient.

Tobi Lütke 38:02

Just like I was just like, I mean, I'm also like a customer for Rio to comparison to like, random points of history. So it's, well,

Balaji Srinivasan 38:09

I will explain that to our viewers who don't know, actually cares, you're, you're a historian economics. You go, Okay? Well, the Hanseatic League is like an alliance of merchants in Northern Europe, that was actually almost completely it was it was for mutual self defense was almost completely non violent. And it's a really interesting model for sort of network cooperation between, you know, basically independent entities. And so thinking of you guys is kind of a modern network on the article. It's kind of kind of interesting. I have a few actually, remarks I want. And let me jump to the next point. First, on the attention point, it's interesting, you know, lots of folks have talked about this a lot is talked about context switching, ever seen that meme that has, like, four levels of brain at the end? It's like Galaxy brain first, right? So it's like, you know, what is the constraint on a business? You know, like, and I'm not saying obviously, this is sometimes true, but like, level one is money. Level two is time. Level three is risk, and then maybe level four is attention. Right? And so money and time are kind of obvious. Risk is something I think, and maybe you could invert three and four. But you know, for example, this is something that you know, Brian Armstrong and I talk a lot about like at big companies, the scarcity, but the scare versus budget is often a risk budget, like the budget take a risk at small companies, it's often a tension, like what they focus on. Right? So maybe those two are tied. And you know, being able to just automate something by clicking, you know, is a huge savings. The second kind of thing I was gonna say with respect to that was, here's a vision of the future, which I'm not saying is necessarily 100%. People will argue there's parts of it that aren't good. Okay, but let me let me bounce it off, you get your thoughts. In a sense, let's say you're running, not a Shopify store, but a retail outlet offline, what you're doing is essentially, arbitrage you are, I'm guessing that the water bottles that you buy wholesale from the manufacturer, you will be able to markup at five or 10% and sell them cold on the beach at this place at this time, and make, you know, a markup there, right, you are trying to literally buy low and sell high, this particular good. And you multiply that across not just the water bottles, but the apples and everything else that you're buying. And you have to take inventory risk while doing that right. Now, once you put that into a Shopify storefront or let's say an Uber Eats storefront or something like that, now, suddenly, you've got a lot more analytics, and you could maybe have multiple storefronts, and you could try different branding, it's way easier to change the signage. And so now the arbitrage becomes pure, you're just hitting keys on a keyboard, and you're like, Okay, apples here, bananas here. And maybe you don't even have to execute the order on the back end until the person orders on the front end, because your latency doesn't have to be that fast if they're ordering it online, right. So in a sense, where you're kind of doing is, and of course, it's far from this to the still a lot of physicality to it, and maybe I'm wrong, and maybe you'd counter argue, but in a sense, what you're doing is you're moving it closer to almost pure arbitrage like an intellectual game of hitting keys. And actually, you know, drop shipping reduces the physicality all those things, do it. Tell me if you agree with that disagree with that?

Tobi Lütke 41:18

I think I think it's I agree with it. But I think that ends up eventually, like getting stuck on a local maxima. Like I think for Scott. For a couple of years, there was a period of time where like, drop shipping became the specifically drop shipping directly from China became a very common tactic that employed and the reason why that ended up in problem is because it turns out that retail is not quite as I mean, obviously, like cold drinks on the beach, is there's a clear demand for this kind of thing in a moment. But what we see most of the businesses that actually grow other ones, which aren't purely utilitarian or like aren't like marketing directly to home or economic goods, but rather a building communities, right. So the, I mean, so it gives an extreme example, the Supreme, right, like, which, you know, has built an extremely close community around their product and including its own, you know, narrative stories. And, like, and so on, I think you see this generally like, like the best products end up being created. And then some kind of intersection of interests. Like for instance, I, like just just take a real example of something like someone who grew up on a platform is Allbirds. Sneaker company, now they make great visible world sneakers, but they also like, extremely sustainable, sustainably done. And, like, that's an intersection of multiple interests, people, people want great shoes, where they make fantastic ones. And then also people who are environmentally conscious. This is a wonderful thing about the internet, you get into this, like Kevin Kelly world was wonderful essay 2000 678 about finding your 1000 true fans. And, like, I think the exploration around like making something but like it's been meaningful to you beyond like the pure sort of economical, like, like Cost of Goods plus 10% markup or 50 plus markup or whatever. Like if you if you build something that's a story and that causes lifetime value, and optimize for that you actually do better than sort of a supermarket. Like liquid.

Kaz Nejatian 43:41

I'm known for saying I disagree. So I disagree actually with you a little. I think there's a misunderstanding of why people buy, like buy things. I think there's a basically a thing of like, the interesting history of trade, gone back to a guide who would see it was actually like we have fossils he died in in Europe, I think 1000 years ago. We have a grid like this fossil fuels, but the trade that happened that was local was essentially trade for things you need, like food knots, those things. Those are exactly what you were you didn't go very far looking for those things. You just hung out around your village and got what you wanted, but trace within you Want, like would go very far distances. You go people, people back before the wheel would travel hundreds and sometimes 1000s of miles for, for what is now called lipstick essentially

Balaji Srinivasan 44:39

that's that's the reverse of what you might think you might think for the things that you need, you'd go far into things that you want you buy them closer,

Tobi Lütke 44:45

this is a this is a really fascinating. So so this is, I mean, this is a real fossil, but I think you can call it I mean, it's sort of like preserved in the glaciers of Alps. So they analyzed everything that was on him, he was a trader, he was moving from one place to another and couldn't quite make it. Actually, I think I think they found error in his back so someone may not make it and but they analyzed the items on his body and like the goods he was traveling with for sort of arbitrage goods, the nice to haves on his body, very good jewelry, Verve of like tattoos, and so on. Then Bender charted out where they came from came from a region of about 3000 kilometers around the point they died, like, and that just tells you how much how vibrant trade actually was, like six to 8000 years ago, like our story is like, not quite right. Like, like trade probably caused language in order. I mean, nothing is cause and effect Exactly. But like, like we have been traders and merchants and butters for as long as we are recognisably home or anything. And I think that's it's tends to not be written into the story quite as much.

Balaji Srinivasan 46:02

Do you know Gobekli? Tapie? I'm probably pronouncing that badly. It's like this the set of ruins in Turkey that pushed back like the dawn of civilization like 1000s of years from before people thought it was

Tobi Lütke 46:15

the one with a very peculiar religion.

Balaji Srinivasan 46:18

Yeah, it's got it's got like some soccer towers and stuff there. It's like under a mound. I think you're right. You know, one thing, it's interesting, there's this podcast called The think fall of civilizations podcast, and it goes through all of these ancient civilizations that didn't make it just like, let's see, you know, an arrow in their back or something. Right? And you realize, actually, in some ways, we're like, you're playing game Battletoads. It's a really hard game. And and you often get really far, and then you die. Right? And I'm like, is that what civilization is? And we've just gotten to like, level 80. And we just need to not, I mean, it's kind of an interesting way of thinking about it. Go ahead.

Tobi Lütke 46:57

Yeah, I'm just I just honestly, I'm just delighted what we're doing because the Hanseatic League, the East India Company, we do et Cie and WNBA. Also do Battletoads. I think that's the range of his podcast is fantastic.

Balaji Srinivasan 47:08

That's wonderful. This is, this is, this is the kind of thing I love. You know, the actually, it's funny is the one previous point that you made, where you said, actually, that it was very boring, the history of the corporation actually has actually been all that stuff is very applied right now, you know, why? Why? Because, you know, we are probably reinventing many institutions over the next currently and over the next years and decades. And so for example, the advent of cryptocurrency and the smart contract is a sort of, I mean, the corporation is a social abstraction, right to take this thing that just exists in our heads, and treated in some senses as a person that can own property and where you limit liability that was a social convention that people had to arrive at, you know, as you talk about, and similarly, like a, like a, like a smart contract is like that. It is a an abstraction, where we are essentially all agreeing to treat bits on computers around the world in a certain way, because they have certain properties are hard to muck with and whatnot, and they give us mutual value. And we cooperate in that way. So I think a lot of the stuff about how those things emerged, it didn't they didn't emerge in their clean, final debug edge case optimized form, they emerge in a very messy way early on are useful to learn about

Kaz Nejatian 48:22

crypto maximalist. And I think there's an interesting, interesting history of it like, like the way that people actually look at crypto right now and say, Oh, this thing is janky. It doesn't work, you know, was janky else doesn't work, the common law of the English empire, like you just need a starting point A to build upon it. And to build a plant. Like, people misunderstand how things are created. Things are created by tinkering, not by five year plans. Like you just have the thing you have and you tinker upon and you improve it, you make it better over time, and all of a sudden, you look back and like, oh, wow, this is like, the very interesting, like, British Commonwealth French civil code thing. That I think is like a difference between, like tinkering and not, and I'm

Balaji Srinivasan 49:02

quoting Apollyon and Academy front, say the top down versus bottom up, right.

Kaz Nejatian 49:06

Yeah. And I think there's something very wonderful about this idea that like, what cryptocurrencies and smart contracts allow you to do is to see the history of change. And knowing the history of change is as important as knowing state. Like is actually very important because the actually what you end up being do the project forward to this tinkering later. This, I can tell you like there's a part of Shopify that Toby himself wrote called active merchant. It's an open source library, you can actually look it up. That started, like how we how we think about payments, payment processing on the internet, took us take credit for this, but I will give him some active merchant really started much of commerce and internet. Like, if you look at all of modern PIM companies, they'll say, Oh, we saw active merchant, and they said, Oh, we should build something for this. And I was just, I was just tinkering, right, it was just a tinkering from the last thing they built. And I think that's what that's the most wonderful thing about thinking about certainly smart contracts, that it is a thing that you can tinker with.

Tobi Lütke 50:04

And I totally agree with what you say, in that fear of sort of reinventing, or at least re integrating novel ideas about history from these new primitives, right, like, I have to say, when I really started studying crypto, which, mutually into, like I had this, like, sort of rapid realization of things that I treated as almost just as constants were actually more interesting than they were like, like, I mean, just the concept of ownership, right? Like, you know, like ownership, like just dividing ownership of a thing into two bits into two separate aspects such as the utilitarian value of something and the status of something if I mean, this is something you talked about in length about I probably learned from from from from listening to

constants becoming variables, I was fascinated with, I own NF Ts, which I'm extremely, like happy about the like, like having, and I have a sometimes from the actual artists, like a version that I'm hanging on the wall, but like, if that's being stolen, it just doesn't matter, really, to me, like someone else can enjoy this, like I, you know, I own the right and, and that those things are separate, they were just treated as the same thing, partly because there was no other way in the physical world because it was like, like the ownership and for you to attain value over and like linked because you could you couldn't take one without the other, it's just like, tells us something new about the world, where this becomes the most interesting really assess to me is like building anything states, countries, cities, communities, villages, companies, is done with people. And we basically have largely free different ways of creating these things, we have a we have, we have, we have policies we have, like policies and processes. One we have culture, and incentives and like narrative, like as another one and then vessel. So instead of cinderford and an organizational design or incentives is sort of a third way, but like everything that the ever built in the human world is really as using these free tools. And solving large scale coordination challenges has to always be done like this. And you have to like, like, people then misunderstand some of the policies or like or end up bringing external incentive systems or hierarchies into the mix, and everything gets really, really messy. And leadership is largely like finding and so somehow changing people's minds with the same tools. They haven't had anything new come online until the smart contract, the smart contract is a way of like, large scale incentive alignment. It is amazing how many people have worked together. Because like, and believing the same things trusting like the trust into the system externalize because of this verse verse, a smart contract that it's verifiable. And then people believe in it. But like it can continue executing these incentive systems for absolutely forever. Even Even if a blockchain stops moving, you could be assassinated, interested in just the next transaction would interact with it, no problem, even 10,000 years from now, that question like this has been used for a whole lot of things, many of them are not actually valuable, but like, the new primitive is there, and we will figure out in the next 10 years, other amazing things that can be done with it. I think this is super exciting. And again, I do think every time something new becomes possible that was previously not possible. people overestimate the impact of that in the short term, but massively under is underestimate the long term consequences of it because at some point, you know, people have done their six seven attempts to do something with it, all of them failed, but they have no like an intuitive native understanding of the primitives and can can apply them to large scale problems.

Kaz Nejatian 54:15

I think because there's actually really interesting like, consequences to like how cryptocurrencies work in commerce, honestly, people usually first go up payments. I think payments is interesting on the payments nodes, we can talk about it, when there's a more interesting thing, which is what Toby was talking about earlier on, which is most, most of the things you value have a community around them. If you think of like, the things like, you know, things, you own things you value to have a community around them. And we're early on this journey, but we're really, really still seeing like, you know, really interesting use of cryptocurrencies for, you know, gating of things and FDA gating or creating communities that are almost entirely about being fans of things. And you look at like, these, you know, super traditional brands, like Mattel being, like, creating these interesting experiences that are about community, as are much more easily done in a wallet aware world. Because the original sin of the internet is like lack of identity. And, you know, cryptocurrencies actually solve it in a really interesting way that allows you to not just have a community, but to actually be able to recruit and trade that community way and like really cool ways. So I think I'm, I think we're gonna think about like, no, 5000 years from now, people are gonna look at us playing with these NF T's. And think about us as like, see if the NF Ts, like, like

Tobi Lütke 55:44

those people have arrows and obix.

Kaz Nejatian 55:46

But no idea what they were doing. They were just like, these were just toys to them. It's really, they couldn't be very clearly going to be a thing. Just don't know what timeframe?

Balaji Srinivasan 55:55

Well, it's interesting. I Okay, so I was writing out a bunch remarks and all those interesting comments. One concept, by the way, where there's sort of a difference between like ownership versus utility is like, let's say, federally owned parks, in theory, the federal government owns it, but anybody can walk in and use it. That's like one example of a owns it, but B, C, and D can use it as an example. Another thought, you know, was that your your kind of tripartite thing there, I might summarize it as policies processing payments, just to have as an alliterative thing. And you'd mentioned common law earlier. And I think one thing I've said before and thought a lot about is that you go from common law of the UK, which is all just kind of iteration, then the Constitution, which is like written down. And then smart contracts. That's like version 3.0. You know, where it's even more Universalist, it's even more auditable, and fair and so on. Right? Then, you know, it's like turning it into literal code, not just legal code. And I think actually, you know, somebody once said, why is it that all founders had a certain point, they all become like, philosophers, and they post stoic stuff, and Marcus or all this type of stuff. And I actually had an explanation for that, which is, the founder starts as lead engineer and ends up as chief psychologist. Right? Because you're, you're basically managing this gigantic community first within the company, and then also outside. And managing humans is very different than, you know, like working with machines, but you kind of need both right? The machine will do exactly what you asked it to do, if you shut down a node, boot up a node shutdown a node, but doesn't care. Humans aren't like that. They're highly stateful. You know, and they are, they're, they're not LT, they're not time invariant systems. And

Tobi Lütke 57:45

so deterministic turns out to like, non deterministic,

Balaji Srinivasan 57:47

exactly, that's right. And so, so on the one hand, anything that you can do with a machine, you actually prefer to do the machine if you can, because then it can scale you can iterate on it doesn't, it's feelings aren't hurt, and so on and so forth. On their hand. That's why founders actually like all this stuff, you know, all these things become applied subjects. Even if you start an engineering, you actually have to pick up effectively a humanities education. Because only history gives you enough, you know, game film on on what large groups of humans actually do. Let me pause there. Tell me if you guys agree with that. I totally agree.

Tobi Lütke 58:20

I think company building is supplied philosophy in a way. It's like if you hang out with people who, you know, good poker players, they are going to be the most interesting game theorists you ever come across, right? And they have no idea that they have any kind of aptitude for this. They just pick it up intuitively, because that fit, like the game they dedicate themselves to is applied Game Theory. They solve it in hardware. Exactly. Well, actually, it was super interestingly, sounds like when they analyzed with fMRI is like Go players and then poker players. They look at cards look at board and it's actually the visual cortex that lights up, which is a big part of our energy budget of a brain so like, we actually basically figure out how to use their GPUs. Yeah, have you ever been used for GPUs? And and it's actually funny because I mean, this is via no getting on a crazy tangent, but like, it's really funny because when you when you talk to these people and say like, why did you do this move? They actually say like, it seemed beautiful, like, you realize, yeah, Because aesthetics is the only way how your visual cortex can actually communicate with your reasoning brain. So it has to use like, basically smoke signals to send information back. And it kind of all fits together. It's kind of it's really neat.

Balaji Srinivasan 59:39

You know, the other thing that's like that, just to just to, you know, like digress on your digression for a second is, I think a lot of computations are like this. So one part is like the visual, like using your GPU. The other is people are very social animals. So there's fairly complicated things, if, for example, someone says, oh, a will be happy, if C does D with E, right? That's like actually a fairly complicated graph theoretic thing. If you were to actually like, write that down, you'd actually have to explain that, you know, a social computation that somebody will feel this, if somebody else does this to this person with this other person at this time, right? I've thought that there might be something like that as well like that, for those people who are good socially, there's like a graph engine, just like there's a visual cortex, I should probably talk to neurologists and see if there's something there. One part I just want to come back to for a second and move on is the part about the arbitrage point. And the community point, they're actually kind of related, where, you know, how much of selling is just arbitrage versus how much of it involves the community. And, and I do think it's both because if you're just doing drop shipping, you don't have as much of an investment, you're just seeking a return, you're not there for the long term. And the guy who loves it will, will probably Outlast because they'll the last even beyond where the market is temporarily down, or they'll figure out some iteration because they are just up at night working on it. And you know, that's also related to crypto where it's, you know, it's not just payment really, it's about community, and people who share the same values. And it's almost like, you know what the Yen is the Japanese a cryptocurrency is to its community, the Yen has value even if you're not Japanese, because enough Japanese people give it value. So if enough of those people coordinate, they can actually build something amongst themselves. Even if the rest of world doesn't really value it. Even if you don't hold the yen, you know that even if you don't personally hold it, you know, the Yen has value. So that's kind of a thing about crypto, where it's not just a technology, it's a community as well that I think you guys get that others don't. And that Shopify is also you know, your community is actually a big part of it. It's not simply just a tech stack,

Tobi Lütke 1:01:33

I have a little follow up, I agree with your job. But here's the other thing you definitely learned, it's like one of the common problems around offers for me as a founder of a company is, you know, I say things which are really important to me, like, we're here to change Shopify, as of Kobe, or Freiburg and change, we make a change. And I could supply this with all sorts of meta data as what we are doing, I know, in the end, this is going to be translated into a fortune cookie, because the packet size for MTU of like communication inside of a company is so small, that everything has to be packaged in a fortune cookie. And then unfortunately, this fortune cookie is when non deterministically unpacked by everyone. And so like I like one more thing, mistake a lot of founders do, versus a like a very complicated journey for for people is that initially, you have a very small group of people, we can move huge context over like pizzas and beers or in the evenings and you get everyone not just like the title of the article, you actually give me exactly like the exact wording of the article and sort of align for that, as things get larger, you will, you will have to consider your words very carefully in a way that you have to pick how you packaging them, because you need to like ideally, determine what fortune cookie will actually say. And then you will have to deal with other like variants of unpacking. Again, that's what the best we got in the void of how we are building organizations and communities. Again, for smart contracts vote can be an antidote here. Like if the particular thing that you're trying to do is a like can be almost arbitrarily complex if, if in the end, it tells everyone how to behave or what good looks like in the community, right. And I think that's exciting. I think that's like, this was completely new in new ways. Like you look at something like DoorDash or so and you realize like, yes, it's not a smart contract. It's a centralized system. And so in the middle of it, it was a two sided marketplace brokering, but there's like millions and millions of drivers working towards like an incentive system that they know works for them. And that's all coordinated by like something which like software, which is fully deterministic and provable, and observable. You could imagine the Victorian version of or Hanseatic version of DoorDash where someone tried to build this entire thing with telegrams and a very, very large middle management layer of people who are then trying to recruit drivers and like set shifts and move food orders around. I don't think this like even if someone would try, I don't think it would work, right. So like, you just sort of get a sense for that computing isn't just a computation isn't just like allowing us to become more efficient at things that we want to do. It's a really gets more interesting when we realize, hey, there's now things that can be done. Because we can solve problems at vastly higher coordination, tax and coordination challenges. I can suddenly this Hanseatic version of, of DoorDash. I don't know what fortune cookie could be crafted, that would unpack for millions of drivers or get horse drawn carriages drive like Stuart's to behave correctly in all situations. So just yeah, it can be created. Right? So I think that's awesome.

Kaz Nejatian 1:05:25

Absolutely. Even further, because he went to Victorian times gonna go back to Darius. Okay. But there's, I think there's a couple, there's a problem, honestly, I think is that the reason transfer knowledge has become harder is because communication has become easier. Like there's so much like this, this is the meta problem with Swift slack is everyone's got a problem with Slack, right? Because attacks on communication has gone down, we have a lot more of it, which means transferring knowledge has become more difficult. Now it's become more democratized. And then district is a good thing. But I think it has become like in a world in which poses like DoorDash works is because the communication channel between the people is very clean, there's no noise to it, go here for this price for that value. Pick this, like, that's just like, key imagine trying to coordinate and slack. Or even worse than office, like, you just never get anything done. And then there's a very, I don't know, I

Balaji Srinivasan 1:06:25

think it's actually a really good insight that communication has risen, and then necessarily average context has dropped. And it's like, if you ever have a tweet that starts going really viral, it goes outside of those people who kind of know you or know what words you're using. And then it goes to like, a group of people who have just no context whatsoever, besides those characters of note, don't care about anything else. And they'll often attack it, or they'll go crazy about or not understand it, or something like that, right? I actually have a solution for this or a proposed solution that is actually technological one. But just to give a little more flesh on the bones of the problem. Ben Horowitz said that his biggest thing that he learned would being CEO was, or just really a senior executive needs to know is any statement, how it's going to be perceived by every single person in the company, the same statement that may make a ton of sense to your exec, who has lots of context is just going to be perceived differently by somebody who's just joined, versus, you know, like a merchant, or somebody who sees it online. And so you have to sort of build in armor plating, or what have you to make it hard to interpret. But you have to, you're doing a very space constrained environment. And I think of that as almost like a, it's like a one liner script with no local dependencies. You know, it's always like, like, it's like tweet like communication, because you're just, you're establishing a beachhead, and you're living off the land, and you have no local dependencies, right. But that does limit what you can do. Because really, what you want to be able to do is dependency check. You want a dependency check prior to installing to take that computer science analogies. How do you dependency check? Well, one way of doing it would be like, for example, you have an implicit one, if they're within your slack, they're within the company versus outside. So it's like a dependency check. But you can actually generalize this with smart contracts you're saying, and NF t's not in the sense of like an art NFT. But like a badge NFT, for example, you might have a post that's only legible to those merchants who have sold at least $1 on Shopify. Right, those are folks who have gone through the entire interface, because it's like a different thing to think about doing it versus actually doing it right. And then you might offer those people hey, can you give us feedback on it? Right? And maybe it's like you further limited only those people who sold one thing at $1. And did so in the last year, they'd sold their first dollar last year, because the older merchants have figured out the interface enough, right? So that's something which you could, you could have very fine grained conditional queries on on groups. If they have that particular entity, they can login, for example, you know, Alon has like 100 million followers. And I've thought for a long time, what if they could do something other than just like an RT? Maybe he could put up for example, a task that is limited to only those people who have the rocket science credential, right? That NF T and so those 100,000 followers of his who are rocket scientists, suddenly you can collaborate on something and it's muted for other people. Or a third example like you're an educator and you you limit replies to those who have done the reading, as demonstrated with them asking, you know, answering correctly, three questions and getting a little NFT badge for that, right? So the NFT's is little markers of I did this, I completed this, I have this, you know, thing, I think lens particles. So I think that's one possible solution to this context loss problem. Let me know your thoughts.

Tobi Lütke 1:09:37

I love it. Like, I mean, this is what we like NFT gating is what we have acquired and commerce, right. Like, it's people will be able to, like, you know, like, you know, I mean, if you can get into business, it doesn't doesn't matter. But like, I actually think, fundamentally, if everything is in the abstract realm of status and internet accessibility, and can I read something or not? I think that's valuable in a lot of sub communities. I think to make it mainstream, I think you want to make you want to bridge it to the physical world. So this is why, you know, this is, I think, a role that Shopify quite specifically had them. Initially, I really disliked but the Kryptos ecosystem was calling things wallets. So casually, I really didn't like the term even buy it sort of get it with like it being money. But I actually, I really come around around but it's actually was a genius. idea, because, but most of the use of a wallet these days is actually in the like, at a station carts that, like allowed to operate vehicle and these kind of things. And I think actually, that is the better, like, maybe not even a better use, but like a more consequential use in the long run right at the stations of skill sets. I mean, this is what degrees in school, it's read like degrees are just a really, really, really slow way of of getting an attestation that you are the kind of person who can complete habit, or whatever. And then other people can run a fortune cookie quality smart, like checks on that. And like, for instance, to get you into companies that some hope some companies recruit, like doing this way more fine grained is way more interesting. I think it just, I think this is forever, you will find some of them mostly, like just fascinating use cases, because it can solve the noisy channel problem, right? Like, it can mean that you can have a conversation amongst people who have proven uplight success in rocket science or half of respect, but by votes by two, and then you can have community and conversation filtered down to that if that's something that is of value to to some people, which is currently very hard to do other than in a sort of a temporal kind of, like array by Invite Only

Kaz Nejatian 1:11:57

problem, I think it'd be that like, you can actually replicate this today, in centralized systems, like a Facebook group is closed and has rules is perfect for this. But the problem is, that's not transferable in a way that is useful, right? Like, I would love to limit all my interactions with like, people who have never screamed like, cancel that person. Like, just like, across the internet, like if I didn't have like a browser plug in that says, okay, just like, remove those people from my like, thing that would just make my it was significantly improved my life. But right now, I can't stop adoption. And because I want to live in the real world, I have to abide by the real constraints placed on it. So I think there is a is a missing primitive in the world that I think wallets can solve.

Balaji Srinivasan 1:12:41

So I agree with you guys more than you. I mean, I think of this as computable context, computable context is a good term, because it's like, on this person or on this individual on this entity, I can look up their history of NF T's it's like their resume, but it's a computable resume, it's not just their GitHub commits, which is fine, or their LinkedIn self self attestation it is something that is both what they have, like themselves chosen to display and then also maybe what others have, you know, set, right. And, you know, it's funny, you also mentioned the wallet, you know, upstream of everything in crypto in a sense is identity. Because if you do that digital signature, you can send $1 But you could also have one vote in a protocol, you can enter you know, get into one door, or you can show one credential. So the ens kind of thing of like showing your identity up top that gates all this other actions is kind of like your, as you said, your your identity within your wallet that you kind of show your, your your card to get through the TSA or something, right. Anyway, so the computer will context thing is just something I think, takes your insightful points, you know, casual, your point on the communications reason, but content is dropping, maybe give some way of, of dealing with that, right, you know, it's kind of like you seen The Simpsons episode where he's like, a beer, the cause of and solution to all women's problems, right? And so it's funny, but you know, often, you know, technology the cause of and solution to all of these problems, right? I'm sure someone will be like, Oh, that's a perfect encapsulation, everything I hate about tech, blah, blah, like, you know, it's actually it's one of these things where people made cars and the first cars were not that great and they you know, the belt shout fumes and so on. And now we're actually solving that and, and so yeah, See is true. It's a cause of and solution to even if it sounds like a joke. Okay, let me pause there and get your thoughts and I'll move on to the next question.

Kaz Nejatian 1:14:30

I'm pro beer is what I want to go on this record.

Balaji Srinivasan 1:14:34

So we covered basically the scale of Shopify economy. We covered the point that you've built something literally on par with a not so small country, you know, your GMV annualized is on par with like Greece or New Zealand's GDP even it's not exactly apples to apples. It's in the ballpark. You mentioned Chavez collective revenue of all the merchants is 444. Bill, you know, which is, which is crazy. And actually, is it four to four pillars at 184

Kaz Nejatian 1:14:59

GMV in revenue, different things, because you're counting Jamia goes through Shopify, whereas the collected revenue is more than just that, right? They sell on many channels enabled by Shopify, I think many of them actually have a very large b2b business, for example, okay, you can buy, you know, you can buy some of the beauty brands that we sell in stores, to the revenues are hard and Archie.

Balaji Srinivasan 1:15:21

So like, even if we just took the limited number of 184 bill as your q3 annualized GMV, that's still massive, right? So then, you know, you've done this thing, you built this thing, you've also managed both, you know, like, internal and external pieces. You know, one thing is, given that you actually have a perspective that is comparable to that of a serious country in terms of like the scale of the economy. Um, you know, have you thought about putting up like a Shopify stats dashboard. For example, I put up a bounty a year and a half ago on an inflation dashboard, and I was thinking stripe or square, or Shopify, or Coinbase, somebody who has a long history of prices, you guys are perfectly positioned, you have a long history of prices across many different kinds of merchants, you can see how those have changed. You could calculate aggregate stats, and you could build a site, that would be much more granular than the CPI, because people could just punch in their own basket of goods, and you could actually see, just, okay, it's I'm not hallucinating it guacamole is way more expensive, you know, right? Because you kind of think, Am I Am I crazy? Or is all this food gotten way more expensive? Or am I just you know, like, it's in my mind, right? Because you don't save every receipt, you don't have all those graphs, right? But you guys do. So you could actually aggregate them to show graphs or individual items, I kind of think that would be just a huge traffic driver. And it's a source of shadow statistics. And maybe it's like an institution like an external institution, that replacing institution like, you know, maybe the Canadian government, even us, what do you guys think about something like that?

Tobi Lütke 1:16:55

I like the idea. We haven't written it like, it's, we have such stats, and totally for sure, because we will be very monitoring this like, I mean, I would save it as a couple of hurdles. Like, in a way of doing this straight out through, there's certainly a conversation we will have to have, you know, just like the merchants and how to, you know, like, how they would think about like, even aggregated versions of

Balaji Srinivasan 1:17:18

like, of course, you have to opt in and share it and so on. Yeah, go ahead.

Tobi Lütke 1:17:21

I think like one of the quips early get by our sort of early investors, board members about Shopify, so Shopify is the kind of company that only Canadians can build, because it's like, fundamentally wants other people to look good rather than itself. And we really like to be a brand behind the brand. thing. Now this due to the successes, this has sort of evolved since and via, you know, putting shop pay more Center as a clearly a global thing, and the shop app, and so on. So maybe this might be a good time to rethink some of those things. I mean, for like, purely on merit, like I think would be super cool thing to do. Like I think having some kind of, you know, Google Trends or cloud for radar for like small medium business world, if you're hosting a significant percentage of the SMB world, and like Shopify runs a good percentage of internet traffic, is Shopify stores. So we are definitely at the stage No, but I would say, recent arrivals at the stage, that we can start thinking about sort of more, you know, representative, larger things. There is a one event Black Friday, Cyber Monday, which has always been, you know, some it's a hot internally, it's like there, there's a lot to it, just because frankly, we have to keep a whole lot of servers from going on fire. Because the traffic surge is so massively ahead, like almost a year forward during those times. And we've talked about this externally and kind of create, like, we actually do share, like sort of a real time dashboard. Like if you can see how, you know, you see, like different countries come online, different time zones and so on, and for sales per minute, and it's obviously a fantastically interesting thing. So there's certainly demand for this. It's, I guess, it's a great idea and maybe you should just like practice it.

Balaji Srinivasan 1:19:23

Great. Awesome. I'm happy to help. By the way, if you guys want, you know, maybe we can have this as the little launch clip for when it goes or something like that. Go ahead and sweet.

Kaz Nejatian 1:19:32

We should do it. I'm just gonna live, we're gonna send you to career website after this, and you can join a crypto team man.

Balaji Srinivasan 1:19:43

I am, I am happy to help. And we can we you know, let's see how we can collaborate. So, you know, one thing that's kind of the macro point on that, you know, I love the concept of Shopify trends. And what you might do is it may be that inflation free inflation stats were like your public loss leader, and then anybody could sign up for the pro version and just see all these trends across everything. And maybe every merchant that contributed their numbers got free access, something like that could be something as like a stone soup model, right? That's just that's just, that's off the top of my head, you know, quick thought experiment.

Kaz Nejatian 1:20:16

I think if it's a good thing for world, we should do it. But you'd be shocked how little time we spent thinking about how to monetize things, honestly. Like, it's not that, like, I think this is like one of those very weird things where like, time value of money is a hell of a drug for other people. And we tried to very much to particularly like, not on that particular drug.

Balaji Srinivasan 1:20:37

I agree. I was just, I was thinking about how you, you could have it be sustainable, since you're going to be running a lot of compute jobs, and so on and so forth on it. Right. But I agree with you. I mean, in fact, I was I'm a career academic, I was actually, I was certainly not, you know, I don't know if I'll give the exact dollar terms. But like, let's say I was not focused on making money for the vast majority of my life. And I'm still not focused on making money, but I'm focused on making meaning. And so I'm with you guys on that. So one thought, oh, by the way, I just, you know, the kind of macro point here I think, is international networks, should be able to collect better statistics than nation states. So like international versus national and network versus state. And the reason being, because you have the same format, you are digital native, you have data analysis jobs, you can do the user interface, all that type of stuff, you can you know, it's real time, none of that is something there's a few governments like Estonia, or what have you that have some of those characteristics, and what are some of those characters, the one thing you cannot do is you can't coerce people into just giving you the data, the you know, governments can do that with, you know, but But you don't need that. And so I actually think this is an important kind of growth area to get into, because it puts you upstream. It's like non obvious, but it basically makes you like a news outlet, where the information breaks first, at Shopify, and then stories are written downstream of that. And you could eventually put that into an oracle that then allows people to programmatically access and maybe do interesting things with it. So I actually think this is a potentially strategically important thing. Because you're used you actually guys define what is true from data. And then folks react to that, which is actually a good place to be, in terms of also focusing people's attention on what's important. If some shortage is happening, and people don't know about it, you might see it first. And you could alert people to that rather than, you know, the Kardashians, nothing against the Kardashians, but I'm just saying, like, you can help prioritize. Let me pause here and move on to there.

Kaz Nejatian 1:22:28

We have observed, we have observed preferences on our databases rather than state preferences. Like this is one of those things that like is real about like, Yes, I think like, three terabytes per minute of practical with you, I feel like it's not like a It's not nothing. When it was wonderful things about Shopify, ology is that we are much closer to a real world than most than much of the technology world, honestly, because we deal with this very real thing. It's commerce inventory. Yeah, inventory, and shipments. And like, this is like a real thing. So it's like, it's like, it makes our ability. We put our money where our mouth is on this, right? We lend money based on this data. Like we give people money based on this data. And so it's a very real thing. So I think I agree it is useful. But Toby not gonna have a layer conversation about whether it's useful for Shopify or whether it's just give it away to the world.

Balaji Srinivasan 1:23:22

Sure, great. Okay. Well, let's, let's do it somehow. But that's great. You know, it's funny, it's like TCPIP, you have to do syn and ack, when there's like a packet that's dropped, or, for you guys, the physical shipping aspect is way more complicated than just a SYN ACK. Right.

Tobi Lütke 1:23:37

Yeah. So like, like, I mean, one of the most surprising, consistent themes is like just how much more complicated atoms have in bytes. It's like, this is this is there's like five stages of grief related to this, but everyone's going through a chop fire. And then whenever we get into the realm of logistics,

Balaji Srinivasan 1:23:55

yeah, I think it's, you know, why is that? I think it's because atoms have many continuous failure modes. And digital things tend to fall into buckets to a greater extent. You know, or or you just have less control over the world outside and so on. Maybe that's maybe there's obvious reasons, it's not obvious reason.

Kaz Nejatian 1:24:13

Trucks listen less well than computers is my answer. I like to listen to instructions less well.

Balaji Srinivasan 1:24:22

But you know, my concept of Like, you talked about, like the concept of generalizing and you're printing things out, you know, obviously, you know, the kinds of printed printed document, right? And electromechanical process Springs Life, it was your thinking. In the same way you can almost think of, you hit something on UberEATS. And you're like printing it out, right? Everything happens in the digital world first, and you hit print and it comes out, right? Or you scroll through Shopify website, and whatever you find interesting, you hit Buy, and then materializes, prints out at your doorstep, right? So everything starts in digital world first, and it's really important you hit print and materializes. Now, of course, the critique of that as well, obviously, the printing, normal sense is fully automated, whereas the, the package appearing door is not. But every individual step of it has been like there are, for example, robots that grow food on the farms, there's things that will package it, you can see clips of each of these things, there's things that will delivery robots will bring it to your home, and so on. So the thing is, you couldn't build a robotic Shopify economy globally. But maybe you could build a neighborhood. I know Amazon tried something like that with drone delivery. Have you thought about doing something like that,

Tobi Lütke 1:25:27

like the would probably play the role of, you know, providing the demand liquidity for the people who want to build neighborhood freely additive, you know, printing,

Balaji Srinivasan 1:25:39

too complicated to do, you can't rent trucks and all that type of stuff.

Tobi Lütke 1:25:43

I mean, maybe in the future, if not yet. You have plenty of downstream jobs from

Kaz Nejatian 1:25:49

also it's not our job, but it's not our job. It's our job like this is a really important thing about Shopify is that we're not a company is trying to take over every aspect of the world. Honestly, we really aren't that we really are trying to create this. This like, like, if you look at how Shopify built payments, for example, like, like most companies want to build payments for cool, how can I go full stack all the rails, I just like owned the entire thing myself to extract every ounce of value. Shopify built payments on back of 720 payment providers, including one main partner and Stripe, like, like I think our job honestly, just like we think it's our job to make commerce easier for entrepreneurs, we think our job is to make it possible. So developers can do things that we don't have, like lots of things we just don't want to do. But we want to make it possible for other people to make an insane living doing it. So if there's a drone delivery startup out there, it's kas@shopify.com. Send me an email, we'll create API's that make your life easier. That's cool.

Tobi Lütke 1:26:47

I think I think that's a really good visual sort of, when you sort of think about this, like as concentric circles like, there's like the world of technology, Turing machines, and incredible amount of opportunity space that's theoretically created is sort of like the inside of a cell. And outside of a cell is like a call of a real world for lack of better term like maybe the void of atoms. Shopify is kind of like via not an academic Ivory Tower, we like lab code, kind of Eureka company, they're like a bunch of a craftsman tinkerers and our Livia very firmly on the membrane, like, we are basically funneling, obtaining, like everything that opportunity space creates, and try to get best to people on the other side as quickly as possible, because we want, like, we want their businesses to succeed, right, like, so. The theory of a large, I like to think of ourselves as an aggregator of opportunity and value. And, you know, sometimes we take active roles and creating some of those technologies. But that's the video versus a, like, only if the rest of the industry kind of doesn't do the job you want it to do. So. Right? Maybe not that hubs,

Balaji Srinivasan 1:28:08

that makes it makes total sense. And, you know, it's actually, it's interesting, I have a better sense now, like, basically almost anything digital, you're like a, like a connective tissue between the bits world and the atoms world like that inner, that inner ring of the concentric circle, it is something where owning trucks and planes and so on is like a very capex intensive thing. And it's probably better just do that with a partner or something. But you can dispatch them. And you might say, hey, implement our API, and then you can more easily you know, talk to our network or something like that. That helps me get a sense of it, if I'm not mistaken. Yeah. Now, one of the things kind of related to sort of managing an economy level thing is, you've you've had a great balance of the centralized and decentralized now, often these are just just to kind of motivate the question often centralized and decentralized are nowadays portrayed as totally antithetical right? Now you have to be either one or the other. And often, though, like, it's useful to have some balance or to use the right tool in the right, you know, circumstance. So you have this decentralized, you're, you know, you've said many times, you're arming the rebels by allowing people to sell online, but that's a balance of your centralized scale and the sort of decentralized entrepreneurial energy. And so it's like a state setting the rules of the market, and then you let people interact. And then the the state takes a slice of that in the same way that Shopify sets the rules of You're economy, and then you let people sell and interact within fairly large parameters, and then take a cut of that. Right. And so do you have give thoughts on like when you do something centralized versus when you do it decentralized?

Tobi Lütke 1:29:36

Yeah, I think I mean, I think key thing is, I mean, this is true of all votes, like votes, pointers at regions of the brain, the brain is set to private, rich, like, it's like other people have no observability of what the concept really encodes in your brain. And like, basically, both are pointers into private memory regions. And it means something different to everyone. And then, unfortunately, people end up fighting a lot of these definitions, like centralized decentralized, as I think has just given that you're talking about a podcast of a lot of people from a crypto sphere, like mean something specifically, I think the broader term like the shortfall was thought of itself as actually an agent of decentralization, even though clearly we're running centralized service like that ended up being like I started off in 2004, there was certainly no way to, to build it in a completely decentralized way, given what we weren't trying to do. But the closest thing would have been for the source code up and like do WordPress thing and let everyone install it on their own servers. That was the definition of decentralized in both days that wouldn't qualify as decentralize investment, newer censor, where Shopify like like Petrova does pray on the altar of decentralizing like, like, opportunity, right? Like, this is like, like, the one that I mean, odd, quite democratizing, like, like opportunity is the thing that people really want to push everywhere, right? Like, because, again, winter is awesome. And entrepreneurship is important. And it would be weird if the internet would be a bad place for entrepreneurship. So so this is, this is like everything we are building is basically so that the people we're starting out have a fair shot at building something that conforms to the expectations of the customers that might exist like find the 1000 free true fans,

Balaji Srinivasan 1:31:34

I was gonna give a concrete example. Here's like your obviously you know, Uber, do you ever hear the company called sidecar?

Kaz Nejatian 1:31:42

Like positive pre existing to look like what no one talked about was what Lyft became became Lyft as

Balaji Srinivasan 1:31:47

we know it, well, it's actually so Lyft Lyft survived sidecar did not but I actually believe sidecar was even started before Uber, I think about sidecar is, like, you know, good try, you know, company didn't work out or, but what they did was they were actually leaning into the Hayekian decentralized model in the sense that they said, Look, every driver knows their own supply and demand, we're gonna let every driver set their own prices. And every customer wants to satisfy like, they want a specific car, some person wants a Mercedes, and some person wants this, and so on, and so forth. So we're gonna let people scroll through and find the driver at the price they want and the make and model and they treat it almost like a shopping cart kind of thing. And you pick and then you book a ride, right. And it actually turned out that the Hayekian model of like, decentralized local computation didn't work as well, you know, the central actor of Uber or Lyft, seeing every ride, and all past people who had booked their, and aggregating them, such as you didn't see Mercedes or BMW, or you saw, like a luxury version, like Uber Black, or you didn't see, you know, Toyota or Ford, you just saw Uber X, right. And then they would actually also just quote a straight price to that. And they essentially centralized the market. But they the difference versus like 20th century communism is that they weren't just doing that without information, they actually had better information, not worse information. And so I think about this as a concrete example, sometimes of where sort of libertarian ish Hayekian theories which are, which are, you know, I'm sympathetic to, you know, don't always work. And actually, sometimes the centralized model actually produces better economic results for everybody. So let me provoke with that. And that's very similar to some of the decisions I'm sure you've had on, like, do we set this as a default? Or do we set it and it's not even able to be edited? Or do people actually be able to edit it you know, this guy's you have to make decisions 1000 times and running something what you guys do

Kaz Nejatian 1:33:40

it dislike it when a good people who are Hayek, so it's this, this is bad for my DNA biology and I'm trying to figure out in what way you are wrong, so I can defend Hayek, but I can't quite think of what I'm gonna tweet at you later.

Tobi Lütke 1:33:55

It's a common, like cast has like casting I have a long standing proxy war between like, Friedrich list on my site and and

Kaz Nejatian 1:34:10

Adam Smith Yes, this is this is this is a very common conversation that Toby and I have between lists in Hyack and Marcus Aurelius and Julius Caesar. It's a common conversation between him and nine that we disagree with over frequently. Luckily, it usually works out in his favor.

Tobi Lütke 1:34:31

It's a well, I European kind, I don't actually think that's true. But like, like, I mean, the list even, like economics. Like, again, I'm not gonna get too much into it but like, the main difference to the philosophy is just that there's a role in the fight like, like he approaches economics and state design much more in in a way that you define a game. That is leaves enough space for the market, but also biases the self act like for self dealing, or the self interest in the market towards the greater good, which is like endlessly Theatre's shiz away from and takes on faith. So like, just as a local, it serves offers a similar story to one your, your your shared, Shopify launched with a theme store, just because we wanted to, you know, allow our designers like, just like, clearly, we want to create a two sided marketplace where so that people have these amazing designs, the first version of adverts like purely everyone sets whatever prices, and you had like this massive draw down, like it just ended up everyone under bet each other and no one ended up making money and the quality just suffered enormously. And then in the next version, what we just did is like, hey, we just set a minimum price of, I think of as 180 or $200, for a theme, which was $160, more than the average team was being sold. And of course, you have to go through a bit of change management, like like an angry emails about this kind of thing. And, and for like, three months, no one posted themes. And then month, four, suddenly really high quality things are being submitted, right, like, so I've seen, like creating parameters for the health of a game to be a useful thing. I don't think that has like, again, I still don't know if this is truly a centralized versus decentralized debate at all. Because again, you can actually set parameters for decentralized scales with smart contracts in the future. Previously, you needed to be like, like it like a central central actor to do this. So. But I Yes, I think this is, I think, I mean, this is sort of getting really into the network state. And then, like, internal economics and conversations and of I think there's very little experimentation along those lines in actual economics. And I would really hope that there's more in the future, because I think we actually learned a lot more about incentive at group design and community, even by studying what's already happening in crypto, especially during the last summer. And I think that's really exciting.

Kaz Nejatian 1:37:15

That's a really interesting book, Toby, I'm going to make your case for you. Now. She's gonna bother me. But the guy named John list who has no list and in relation to fedlink, LISREL method can tell who was who was who became the chief economist at lift has a book called The Y axis. It's about like real behavior of bits, but behavior of real people versus but the economic theory tells you should do, I think the best case for why Frederick Lewis was right. It's, it's nearly convincing.

Balaji Srinivasan 1:37:47

I'll make I'll make a quick remark. And we'll move on from there just to, you know, I think the reason maybe I don't quite say Hayek was wrong. But when he was he was right, in the 20th century, and might be wrong this decade, or this century, is the whole thing was the calculation problem, right? The economic calculation problem, but the calculation problem is arguably an information problem. And now we're really good at solving information problems. And we've got gigantic stores of data. And this is something I talked about in the book is maybe like China's centralized model, with their cbdc seeing every single transaction, you know, they could do all kinds of machine learning and actually optimize their economy in a way that would like I wouldn't want it to work from an ethical standpoint, but it might right like, I cannot say a priori that it won't. Right. And, you know, the thing that's funny is like the Shopify themes are that's like a price floor. That's actually Steve Jobs that that also, as I recall, it's like he said, like 99 cents for like, you know, songs, he didn't have them priced lower than that. And he understood that people would anchor on price, you know, Medicare in the US, in theory, it doesn't set prices, because private insurance can set prices they want in practice it does. And all these charge masters and all these insurers anchor on that because the thing that everybody sees this, it's like it's got that hub and that Hub and Spoke position. And so you know, the other aspect here is when you're talking about why do people underbid is you're optimizing for that irrationality and low context, right? They're coming in, the first thing I'll do is sort by price, that's what's legible to them. Then they see everything is crappy, then they leave, and then nobody buys anything. And then you're actually literally acting on behalf of both the customer and the vendor by saying, hey, actually, we want you to get to know each other a little bit better and we're setting a high Are price, but you'll both like it actually overall and more transactions will happen. And ultimately the justification for that is the GMV. And the fact that there's a better customer experience, even if in the, in the small your, quote, violating capitalistic laws in the large, you're still, you know, serving it. So there's like, you're, you're violating Hyack. At the small level, perhaps, but you're delivering on at the large level, that might be one way of resolving it to free cash. Go ahead.

Kaz Nejatian 1:39:48

I mean, I think there has to be incredibly careful. But if it was worth this is true. at Shopify, we're incredibly careful when we set defaults. Because the thing Hayek had called pretense of knowledge, which is the thing, the thing you think, you know, but you actually don't know. So if you think you have perfect information in your design system, you almost always end up being wrong. As you know, we've seen in lots of cover era policies, we think we know things, but we actually don't know them. So this is actually really important for software, because when you set defaults we actually are doing is setting the exclusionary policy. And you want to be really careful. And we are this is why like, look, I think I think Shopify payments is the best payment product on the planet. I think it's very, very good. I think there's no one who should use it. But you're free not to use it in Shopify, that you're free not to use it, you can use something else. But because I think you have to be very careful when you decide you have all the information you need.

Tobi Lütke 1:40:44

And also, like, I think you want some internal like, I mean, Shopify payments is very good. Partly also because it actually, even though it's an internal product that clearly is attached, like, do you have an incentive to, I mean, if we could make a unilateral decision, that's the only thing anyone uses. But again, company building is hard. And having some pressure on this product having to succeed on merit is actually good, right, like so even even bad is like the team should have pride in building something that is superior to the other options.

Kaz Nejatian 1:41:25

It's an odd choice that way, like, literally everyone else's face has made a different choice, either people have made a choice, they will have only one default, and only allow that, or people have chosen, they will have no default and your product, or shopper is alone in the spacing, we will have a product, you don't have to use it, it will compete with every other product on the space. Like if you think you have like, I mean, not direct companies, I think they've made some choices, like square has made a choice, you can only use square payments and square. Most commerce platforms have matrix, they will have no payment product. Whereas Shopify has no choice that we will have a payment product and you can compete with it, which is like very, I think it's a different choice. This is being self aware to know you may be wrong.

Balaji Srinivasan 1:42:06

It's interesting, because it's like, you know, Bayesian statistics like Bayesian software development, like you get the first one and then you update when you get the next measurement and you update and so on. Now, you're probably solution of how people actually want to use it, maybe it's 87%. And you want to actually, you know, move it up each time, their thing is, you know, the so with agreeing with what you're saying, the alternative view is centralized, the simple, Apple just makes the decision for you. And you're going to like it, right? The the prefix A, the Kamikaze model of software development, and it just reduces the combinatorics. And you don't have three choices here. And fortress is there and fighters is there and you know, all these different things you just get one thing is basically just work. So that was actually Rails is design philosophy. If I recall, at least early on, it became more configurable, but like, the golden path is supposed to just work right.

Tobi Lütke 1:42:56

Kamikaze recorded, like it's, you know, what the fuck that right? But like you can change things, but we call that, like web design philosophy in Shopify. Make the important things easy and everything as possible.

Balaji Srinivasan 1:43:09

Yeah, yeah. Okay. Right, exactly. And so, I think I just want to identify, I'm not saying I've got an optimal or, you know, anyone has an optimal solution along that axis, but identify that x is a very important one for like, managing people at large scale, like, what do you do, centralize where you decentralize? The thinking recurs. So, you know, we've basically talked about how you guys have managed and built something that is on the scale of a small country economically. What if you could start a new country a small country, right? You know, like, what would the Shopify country be? What are the Toby contribute? Well, because Country B and doesn't have to be just one thing you know, just like that guy. You said we had five things that failed and one succeeded right? You know, I would I you know, the president of Canada could be you know, the the largest company in Canada or it could build new Canada right? What would you know, perhaps press leaving Canada out of it. Okay. What would you know, what kind of thing if there's a new country Toby, what would you build a castle or?

Tobi Lütke 1:44:10

I'm actually very interested because what builds I think, I think that would be vastly more interesting when, like anything either. I mean, what about us versus like, I I have not spent 50 minutes, I think Think about it. This is why I love your book, because it's like, it's like, oh man, there's like so much more left in the tank in the future or things to explore, like, but I like I would like to see people apply themselves to this, but it probably would probably not be my project

Balaji Srinivasan 1:44:45

I have, I have one thing that helps focus it a little bit like, you know, with starting a company, it's this huge thing you have HR and you have, well, it's not the first thing that you have, you know, payroll, you have a zillion things to take care of. But usually, the way you describe it in a one liner is, you know, here's our first product. And here's what we're selling. And here's why people are buying get into the, quote, elevator pitch, and so on, right. And the kind of equivalent thing that I have for, quote, starting a new country. Just and again, this is like a distillation of lots and lots of intentional communities and things that have been started, right. So a lot of history gets packed into this not really just my invention, but I call it not not coming up with the holy new 10 commandments for your new society. But one commandment, like one big, moral or social thing, some issue that you think society is not solving that you'd solve with your new community, just like a founder of a startup is like, oh, there's some economic issue that I'm going to solve in my new company, you solve this moral or societal issue with your new community? For example, it could be okay, I think I'd like to explore what a full keto community would look like, and people, you know, they drop the cookies at the border, you know, step away from the smugglers, right? You know, and you actually have something where you give people willpower within that community where there's just like, no sugar, no high fructose corn syrup, nothing like that. And probably people would be very healthy. And I give other examples that are like that. So if there's like one feature, if there's one thing policy thing that you could maybe wave a magic wand and change, you know, you've talked about for example, fusion, right? You've talked about all the tech stuff that's being held back, you know, many of your tweets, I was like, oh, it looks so good. Like, I would tweet this or, you know, think about think a lot, a lot of same things, right? You know, fusions being held back and there's a bunch of things are getting held back. So technologically that might be an answer. But there's got to be something you think if you could wave a magic wand, you would you would do differently.

Kaz Nejatian 1:46:31

I read your book, and one of the things I really thought about was how to French, like the French word Natsuo is a better word than like the English word nation. Like because it in French had this idea of PE, which is what we think of as a country. But it's necessarily just a group of people that have like a shared history or shared culture. And like, in fact, the Academy of many nations in one country, I think, like, that's a very, that's an interesting thing to think through. So I'll engage with the, you know, the third rail topic directly, because like, you know, why not? Okay. I do. I do wish people were like, I could have a community that was more curious and less sure. And more open to being offended. I know, it'd be like, if I could like have a community. I mean, look, I am alive only because can't exist. I would have been frustrated that with Okada like that, or like riding around in jail cell. So I have a cat has a special place in my heart, partially because it's the country of nations and you can be a lot of different things in one country. If I could create like a political party, it'd be like, hey, everyone here in this in this discord channel in my like, in my part of world be more curious and less easily offended

Balaji Srinivasan 1:47:51

the tribe of people who are not tribal.

Kaz Nejatian 1:47:54

Yeah. Yeah, right. Or at least tribal. But just one thing only.

Balaji Srinivasan 1:47:58

Now I sit in a somewhat joking way. But it's also I understand what you're saying. Go ahead.

Tobi Lütke 1:48:03

I think I think it's actually not being tribal. It's like, tribalism is super fun. If it's like, high. If it's sort of like the scientist version of being trapped, like, you know, like, then revert cups on I'm like, German, not Canadian. I've both passed. It's like, I'm going super tribal. If Germany plays Canada, it's not close. But like if Germany plays Canada in hockey, I'm Canadian, right? So like, I think there's like, like tribalism is like, in our like, it's built into our, into our brain. It's part of our virtual machine that we're running our brain on and it's like, it's like, it should come out in jest rather than in like, the demand blood, like Oh, tea or absolute absolute conviction. I think the I mean, like if you're if you're talking fancy edits to do to the psyche that you can incorporate into our community. I tweeted about this a bit. I think one of the biggest issues in overall despite the or not like the builders are special in a way like that they need to right back into the story like I really do like founders create Just people who, you know, going on journeys, discovering things, and then bringing back our lessons. We are currently in a world that has that treats the people who build things, and the people who critique things equally. In fact, the airwaves are completely owned by the people who critique things. And because they get to write the story, it's actually basically more powerful. And I think, like, I'm not against criticism. In fact, I think high effort, criticism is one of the most important things to inspire builders to do their thing. But I think critics kept in the environment of value of a concrete for society, by holding builders accountable, or inspiring builders to do better. So they were really a supportive tissue for builders. And I think that I think this is something that ought to be reinforced in the narrative for the good of all, including the critics. Because it just, I think, I mean, I started a company in the sort of magical five, six years, in the middle of, like, after 2000s, read, like they like, almost name any company outside of the fangs. And it got started in this sort of, I mean, probably six years time span. That just ended up being like, right when the internet grew up enough to know that software service was a business model. And advertising is good. And we like, the iPhone came, like what was soon after, and, you know, web two, before red three, was sort of a rallying cry of like, hey, we know how to get into software. And so we should build the important bits of internet software, Shopify is the retail, internet Ripto kind of thing. Like a lot of us hung out together in IRC channels, because a lot of us use things like rails, and we knew each other a lot of us, like know of each other. It was a fairly small community in, in those days of builders, and talking stuff, a lot of them and it's just like, so many of them opted out, like so like, obviously, the best businesses vocht people made money, but no one I felt I knew did it for money. It's a surprise to all of us. This ended up being useful in terms of, or remarkable as a money, thing, like a lot of us, sort of, I think made choices to start our own businesses, saying no to Korea, so if you otherwise would have had it just like I think you people would want, like a community and in a society where it's hard enough to build these things. And it shouldn't be artificially, like harder by the only people who were asked to solve all of society's problems are also the people who will already dedicate themselves to like actually building the stuff in it. Like, we should get politicians to solve society's problems. Instead, we should simplify, and so many, it becomes so hard that so many opted out and I think be would have gotten massively better companies massively better products, massively more product progress, if not, I think some of them luckily, all realized, like sitting on beaches, sipping where he does is actually super boring. And starting to come back and become active. I think crypto has seen many join, I think AI is putting some of us back to become builders again. And that's really good. And I think that's this, for how exactly to do it. I don't know. But like, I think this would be like something I would experiment with.

Balaji Srinivasan 1:53:17

So if I'm to summarize, like cultural innovation around curiosity and cultural innovation around builders, and really, it's a community first kind of thing, more than a policy or technology, per se, and I agree with that, I would add a couple of thoughts on this, which are I've actually thought about some of the same many of the same issues. It's funny, like a lot of our tweets are just like this. Yeah, right. It's convergent. Exactly. So, you know, one thing is, I'll just make a few remarks. I was thinking when you said this, you know, first is I realized, you know, the, okay, let's go back and beat up on our beloved high EQ. And so, comparative advantage makes a ton of sense if person A is selling apples and person B is selling oranges, right you grow your apples, I grow my oranges and you know, hey, we can trade or it doesn't work. If one guy is growing apples and their guys writing laws or, you know, generating media, because that is something where that by outsourcing that by just focusing on what you love, which is the apples, you have outsourced the governance of San Francisco, for example, to people who are bad at it, right. And they are actually you know, you If everybody's making a rational decision that, hey, I'd have, you know, be able to do something positive some and be able to build great stuff and so on and goes and starts a company, then the state is left to those who make irrational decisions. And who didn't, who cannot make good financial choices, because if they were they would be at a, you know, a fund or something. They they're not good at engineering, if they were to be at a tech company instead, what they are, they're optimized for politics. Right. And that is a skill, but it is. It is a skill that is that shouldn't be on its own. Right, you know, like, Lee Kuan Yew was good at politics. He's also good at actually building right. And what I realized is actually, that that's another place where the doctrine of comparative Vantage breaks down, we have outsourced something that we should not have outsourced. And so then what I think that means is I've got like a v1, v2, v3, v4, they'll bounce off you, right? So v1 is let's say, um, I don't mean this in a negative way. This is kind of where I was as well. And so v1 is the naive optimism of the builder I should just be able to build site should be set up so I can build right? But this is like, you guys play games. You play a game civilization. Yeah. So what happens if you just put everything into technology and commerce and you're just peaceful and building? Great guy next door? Yeah.

Tobi Lütke 1:55:36

So comes you have something that someone else wants? Yes,

Balaji Srinivasan 1:55:41

exactly. They call it with a spear man. They just have a little spearmint. And they just come in and they invade your thing. And now you're dead. Right. And so that's actually how I think of a lot of the establishment, the especially the media, corporations, and so on. They don't like these tech interlopers, who are often demographically different, like we're all immigrants, right? We're all from overseas. The old money in many ways, young people are like, Oh, the tech guys are so rich. Their problem isn't actually quote, tech guys are rich. The problem is that it's new money. As opposed to the old money that owns the newspapers. If you go and look at all of these, like media, corporations, they're all owned by like, these nepotistic families that have inherited from generation generation. Often the owners are in like, you know, the back you hear, you know, for example, everybody recognizes Mark Zuckerberg, right? For better or worse, he's out in front, he's taking the hits. He's a founder, he's the son of a dentist built from nothing. Nobody recognizes Arthur G. Sulzberger, who inherited the New York Times Company from his father's father's father's father's father, right? Once you apply that lens of like new money versus old money, well, actually, you can actually look at this kind of conflict happening many times in history. And so if v one is build, and then v two is oh, let me be a critic and take my spears, because what's the point of the critic, right? Their point is to take your stuff, if they can criticize you and get you to apologize. What comes after an apology is some form of reparation of some kind you're giving, you're giving up something, right? Because if I apologize, well, how are you going to make it right? Okay, I'm so sorry. Let me give you something. So it's essentially highway robbery on the internet, right? Where they can back you into a corner for building something right with this criticism. And it's actually think about how low capex it is, right? You had to go and take all this risk. And, you know, spend all this time and write all this code. And this guy can just tweet at you angrily and get you to back down and give you some money. And he he didn't do anything. Right. Extremely great. It's like, it's like the patent lawyers who go look up. What are they called? The lawyers who go around and chances. Yeah, the patent trolls. Yeah, exactly. It's a it's a low cost structure of patent trolls. Like these are like the the pirates who would just go and attack ships that were traveling right there like bandits, essentially, right. And one way of looking at that is, think about how hard it is to let's say, do a deal with I don't know, I'm just picking a name Coca Cola, right to use Coca Cola as logo for something to advertise it. But a media corporation can just take Coca Cola as name and put it in their headline, and use their brand equity to sell their papers, right? Or your or your name, and use your brand equity to sell their papers. They can just kind of steal. They're optimized for all that stuff, right? And so then once you kind of realize this, well, what do you do in civilization? One option is, and this is V three and V three is I'd call it like reflexive reaction. v3 is you just become a critic back and you spend all your time attacking them back, right? The problem with this is, then you become them. Right? So what I've come to after a lot of thought is V four, which is something like 20%, right? 10% Fight 70% built, you actually consciously think of this as a defense department. And you allocate budget for content, just like you have to have a legal staff, right, you have to have some degree of content budget, you have to go direct, you have to actually tell your own story and have your own community or someone will do it for you. Right. And at first, it's a necessary evil and then you actually might get injured you're like, actually, that's pretty cool. You know, once I'm set up to do it, I've got like a nice, you know, studio or I've got a nice, you know, template for our blog, I've figured out a nice story. For example, we talked about earlier Shopify trends, that generates a story every week. And it's an awesome one. It's something that's on topic. And so it's like, it's something that you like, but it's also driving the new cycle and you've got, you know, social media and people are subscribing to you because this becomes important information, right. So you're doing the service the community, as we talked about, you're also building your defense department because you're building in your community of folks who are tuned into your version of events, you're helping your community defend themselves. And that's kind of what I think of as like the four. Alright, let me pause there. That was a download. But I love your thoughts on that

Tobi Lütke 1:59:38

sort of baiting cast into it, like

Kaz Nejatian 1:59:42

I'm very easily baited. I mean, this is interesting. But as you were describing, I just realized that, you know, more or less got rid of piracy, like spicy piracy in the world, by not any one person doing anything, right? We got rid of them by declaring a persona non grata. Anyone could just kill a pirate and you'd be fine. Like it was a whole society thing. Ah, those people are all reflectively bad. Let us all attack them.

Tobi Lütke 2:00:09

Because I disagree. Story,

Kaz Nejatian 2:00:12

you'll think we've got one of pirates, the British Navy plus like persona non grata. How do you get rid of pirates who cool,

Tobi Lütke 2:00:19

I think I think I think all startups a straight lining to the power chips, they have as share operations, people get partly recruited on merit people, like you got to shave off a bounty, it just there was like, there was no opportunity space for this kind of grouping during those times, and therefore of a pirate ship is was a pressure release valve basically pushing entrepreneurial activity into the underground. I think we ended it by actually creating America, basically,

Kaz Nejatian 2:00:53

I think that I think there's an interesting thing to visit the system designed thing that happened, which is, if you take a pirate ship, you could take everything they had, as long as you take the pirate ship, and that could very good incentives for people to attack pirate ships. And that's actually what's needed in the world today. Like of critics, all you need to do is for a couple of people to stand up and say, Ah, I stand up against you with this person who you're criticizing, because what usually happens is someone very loud yells at someone who's building and everyone else cowers. But if someone loud yells at someone who's building an every other builder, the bad signal goes off and starts attacking that person. What ends up happening is the incentive structure changes. The instead of just sort of like picking you off one at a time changes. This is what I think is so important. I disagree with some things Elon does or says, but so point to defend him publicly. Because otherwise the incentive structure is screwed up.

Balaji Srinivasan 2:01:50

So it's funny because there's actually a resolution of both of your guys views into one thesis, antithesis, synthesis. And what that is, is there's two ways of defeating pirates, which are defense and conversion. Right? So are the pirates, the bad guys who will just fight us to the end? Or are could they become one of us with the right incentive structure? And you know, the second model, actually, I've seen that in academia, you know, the fundamental similarity between the socialist professor and the capitalist CO. So, they both deeply want to be in charge. Okay, there's actually something there there, actually, I mean, a lot of the social oppressors are pretty smart, right. And they have frustrated executive authority, and they try to do everything through the state. Right. But once they joined the network, and you actually give them that CEO title, they have more authority over fewer things. But now there's no one to complain to. So they open their mouth to you. And then because they have to close because they have to do it themselves. And now they start figuring it out. Right. And so that's I've seen this over and over again, so many folks become once they realize how hard it is to actually build something how competitive markets are, it converts a lot of them, it may be the best way to convert them, because it's a minus one for the pirate side, and a plus one for your side. Now, and then those that are left are actually those that can be mopped up and dealt with, you know, via coalition, right. Basically, that is that is a, you know, kind of resolution of both your views, which actually, I've seen both of those in practice both the unionization of tech and the conversion of former critics into tech people, for example.

Kaz Nejatian 2:03:26

This leads me to where I started, obviously, you're good at joining our one on ones you should just put you into careers. When I was perfect,

Balaji Srinivasan 2:03:33

I definitely want to talk to you guys about Shopify trends. I'm very interested that Okay, guys, I think we have covered almost everything I wanted to cover. I thought this was great. There's tons more we can talk about. But if you have any final comments, thoughts, anything you want to leave people with, stuff you're excited about.

Kaz Nejatian 2:03:50

Let me let me let me make a shameless pitch for something if you have on to this point is podcast, the odds are that you're a square peg in a round hole world. And Shopify is a company built for those people. It's a company of like ex founders who really believe in changing the world and taking ownership over your own life and helping other people take ownership over their own lives. It's not a typical tech company and it has a has ups and downs. But if you're open to being a typical and if you're deeply curious, Please look us up. We're building something special, and we could use more special people to build it.

Balaji Srinivasan 2:04:32

Well, the best ad for working in Shopify is just you guys. So. So that's, that's great. And I will put that link in the show notes. Toby, you have any thoughts?

Tobi Lütke 2:04:44

was a new podcast, I am amazed at your restraint of like, like, because I know like literally every thing. I read your essays and like we were talking about things, partly poorly, that you have written incredibly eloquent essays that were owed, like, for instance, I find, I was just reminded, like, in our conversation that what I was drawing on was largely inspired by your founding versus inheriting essay, which I really think is one of the top essays that I've come across in the last couple of years. So I'm amazed at your range, because you're being you're being a good podcast host here and not jumping in. And I think it's great fun. I, you know, one of the I was earlier critical on media being clearly an inherited world that is largely staffed by by Labor critics. This is something I've said publicly, I think podcasts are the are so different. And because you hear about the potential of new ways of creating countries even plus all the various histories, or how empires fail, but we can learn about them or like, thinks about optimal like, like optimistic messages, because podcasts are run by the founders read like a baby, thing like that. We are missing in other mediums. And I think it's great that just like you're running this, I think it's great. It's a great topic. I'm glad to be here. It has made a great pitch for for Shopify, for Shopify, and working for Shopify. So I feel like I should meet her pitch you on podcast at the end of one of your episodes, that's not super useful, but like, well, maybe there's something I think is back here.

Balaji Srinivasan 2:06:34

I'll make I'll make one comment on your last comment, and which is basically that, I think, with the tweets, which are the ultra short form, and the podcasts or the ultra long form, those are relatively less occupied by legacy media. And that's one of the reasons we made a lot of progress. And the other is that unlike a normal interview, you know, the quote, interview is like hard hitting, and they're supposed to like ask interrogating questions, and so on. And that was based on a distribution asymmetry of media corporation in the in the subject. Ours is based on a distribution symmetry, it's actually peer to peer. And so I connect to you, and I connect with metallic and we connect to others and so on. And so we're building actually something new together. So very glad to have you guys here. And thank you for being on the podcast.

Tobi Lütke 2:07:17

Awesome. Thank you so much for having us. Cheers.