Introduction

Patrick: [00:00:49] This week's guest is Zack Kanter, the founder and CEO of Stedi. Zack and I decided not to talk much about his business on this podcast, and opted instead to explore business more generally. A bit of an introduction to what they do may be helpful here for some extra context. Stedi is a platform for exchanging and automating 300 types of business-to-business transactions, things like purchase orders and invoices. It's a modern take on an archaic protocol called EDI, Electronic Data Interchange, something I'd never even heard of until a few months ago.

Learning about EDI is a bit like finding out about the matrix. Every physical object you come across, from the food you ate for breakfast to the clothes you're wearing and the electronics you use, basically anything with a barcode on it was likely touched by EDI, often dozens of times before making it into your hands. Stedi is the first update to this messaging layer in decades. Our conversation in this podcast is about business in general, starting with Zack's fascination with Walmart and Amazon. I should also note that my family is a recent investor in Stedi, and I'm thankful to have learned a great deal from Zack over the last few months, please enjoy our conversation.

I want to dive into two businesses that you've already mentioned because of how extensively you've explored them. And because there'll be great excuses to dive into like 10 or 15 interesting bite-sized topics that some of which we've explored before together, some of which we have not. Those two businesses are Walmart and Amazon. I'm curious why you wrote this. And I know it was even longer than what you published. You sent me some of the stuff that got clipped left on the cutting room floor, which was even deeper into this topic. Why are those two businesses so fascinating to you and maybe draw a little picture of the differences and similarities between them?

Zack: [00:02:26] Walmart and Amazon, I think a lot of people appreciate Amazon on one or two dimensions, and that's usually maybe doing business with Amazon or you use something like Amazon Web Services. Or maybe you're just a consumer and you think that the idea of Amazon Prime bringing a package to your house in one hour is the greatest thing in the world. What I've had the fortune to see being a vendor to Amazon, so I've sold with auto parts business, I sold to Amazon both on the vendor platform, which is where they're at the seller of record and on the marketplace where Amazon is the seller of record, obviously as a customer in many different ways. Then also as a business customer in the sense of using Amazon Web Services, which heavily relies on Amazon Web Services for our infrastructure.

And so you just see that the surface area of this business is astonishing and it's not only the surface area, it's the speed or what I would say, the tempo, which we can come back to some of the differences on that later. The idea that Amazon seems to be picking up velocity over time is what makes it the most interesting business. It's the only business that I can think of that as it's gotten bigger, it's gotten faster. Faster not just in one dimension, but faster across a broad range of dimensions that are certainly astonishing to watch. You can't appreciate Amazon without appreciating Walmart because Amazon really took Walmart's playbook and then they brought it onto the internet.

And I think a lot of the things that Amazon has done really well in the ways that Amazon has differentiated itself from Walmart and built, what I believe is a really truly first in kind first-in-class business are things that were required to do business at internet scale. I'm obsessed with the founding of Walmart. Sam Walton's book Made in America, it's fantastic. Hidden beneath this veneer of foxiness is this unbelievable story where he just makes it seem like, “Hey, I woke up out of bed for 40 years and accidentally Walmart is the biggest company on earth.” When you look at Walmart, what they did is a feat of such scale that I think in some ways it's more impressive than Amazon because they reached the limits or got close to the limits of what's possible with intentional coordination.

Whereas Amazon, I look at it more of an organism, which is I think some of the things that got clipped to the cutting room floor in this piece. I think the differences in those models is just almost infinitely rich area to explore.

Walmart and Amazon

Patrick: [00:04:48] How much of what makes those two businesses and their history so interesting, do you think can be borrowed or emulated versus is just singular and very specific to those two businesses? Asked differently, how much of what you're going to do with your business is directly inspired by what you've learned from those two businesses?

Zack: [00:05:08] Of course, what makes the two businesses similar, when you start to align these up side by side, you say, “Okay, what are the commonalities in the business?” The thing that occurred to me when I was writing this piece called What is Amazon, is realizing that they're both optimization problems. What Walmart basically figured out is that their hypothesis in the beginning was rural areas of America can support much larger discount store formats than most people think. And so I think, is it Keith [Farboy] who says, “What secret do you believe about the world that everybody else thinks is false?” Or maybe it was Peter Thiel, I forget. And that was Walmart's core thesis. And at the time people thought it was crazy at 25,000, 50,000, 100,000 square foot store. They thought, “You put it in some town in the middle of America.”

They didn't think it would be able to support that. Well, Walmart basically set out to do is procure these stores and what Sam Walton talks about in the book is he would pick an area based on zip code and some demographics, and he would fly around in his plane and he'd look around and say, “That looks like a pretty good spot for a Walmart”. And so you would land the plane and then he'd get out and he'd negotiate the purchase of this place right on the spot. Had a whole store team. And this is actually, you could see the mirrors of how Uber expanded in region by region. They'd bring in this crack store team and they'd build this place and then they would order goods. They'd order a tractor-trailer load full of welders, or God knows what, and then they would have a grand opening and they'd start to sell this.

What you realize is that Walmart is this optimization problem. You have this footprint and maybe this footprint is a 100,000 square foot store. And what you're optimizing for is within this, assuming that the footprint is fixed, because it has to be increased by step function sizes or large increments, is that you're basically optimizing for the selection of skews, the quantity of each skew that you have. A skew is a product. Selection of products, the quantity of each product that you have, the price for each product and then the placement for each product. It's a pretty simple thing. As they say, “It's as simple and as hard as that.” You've got this optimization problem. Walmart executed on that for 30 years and they got to be the best in the world at that.

And what you see with this optimization problem is an entire organization built around optimizing for this problem. You could break this out as a formula practically, and there's massive trade-off. If you have the correct price, but too few of a skew, and in stock, you robbed the customer of the chance to buy that. If you're selling for too low over price, you've robbed the company of a chance of profit and therefore you've robbed the customer of the chance of investments in R&D that that would bring about. If you have too many of things, or maybe you take up too much shelf space, now you've robbed the customer of the chance of being able to see that product. Or if you screw up placement, you've robbed them of that as well.

If you look around a store like Walmart, what you should see, the correct way to see it, is this massively defective. You see defects everywhere because of course the only good news is bad news. If you don't get bad news, if you only get good news, you can't change anything. When you get bad news, when you find out, “Oh, we've been robbing our customers by pricing this to high or we've been robbing them because we've put this welder at the wrong end cap,” you have a list of things to remediate. The idea here is basically that you can never have the perfect mix of everything unless you have perfect access to information. Your job as the manager for Walmart store constantly to fix these defects. I don't know if this is how they think about it. This is what I think when I walk into a Walmart or a Sam's Club.

Everything else, I've heard these unbelievable things about Walmart that I feel like have gotten overshadowed by Amazon, but they built the first and largest private satellite network so that they could do store-to-store communication. They invented practically EDI. Although I think IBM got most of the gains from it because Walmart didn't really do a whole lot with this standard. Innovations in inventory management, they have a logistics fleet of 7,000 tractor-trailers and all this whole apparatus that makes Walmart look like a very complicated business. And of course it is, but it's all in service of this up until walmart.com, is all in service of this optimization of this footprint. And the whole organization got optimized around that. And then in 1994, I think when Amazon came about, Amazon had their own hypothesis. Amazon's hypothesis or Jeff Bezos's hypothesis was, “Okay, well, what if we did exactly that, but we took away the constraint of the floor space?”

And they said, “It's this idea of infinite shelf space, where all of a sudden, you start to see this paradigm shift where consumers are shopping online, what are the things that are going to stay the same? What are the things that are going to change?” The things that are going to change as all of a sudden, there's really no constraint in terms of shelf space. You don't have to figure out, should I have 500 units out front or should I have 50 units out front? So, it's a different optimization problem. And you also have the ability for customers to look for the specific thing because your customer, maybe they would have been willing to spend 45 minutes in a Walmart. Maybe they're also willing to spend 45 minutes on amazon.com, but they get a lot more browsing done in 45 minutes than they can in 45 minutes of walking around Walmart.

Of course, this brings up questions of discovery, where it's much easier to discover something in a Walmart than it is in an Amazon. We can talk about that. Fundamentally, Amazon came onto the scene and the distinction I draw is basically they came up with this unbound Walmart, which is basically they're taking the Walmart and they're dumping it online, not to downplay what Amazon did, but that's the basic idea. Amazon comes about and starts doing this, and all of a sudden, what you find is that when you take an infinitely massive, incredibly large market, which is consumer packaged goods, and then you collide that with the internet, and now all of a sudden, you're not geographically restricted in terms of … In average, Walmart has 100,000 skews or something like that.

Amazon has 488 million skews. And that was the last time I checked. It was probably another 500 million on top of that. All these weird things start to happen. And so this point of this piece that I wrote, is discussing what is the weirdness that starts to happen when you bring things into internet scale and how did Amazon solve this thing.

Patrick: [00:11:08] What a fascinating sprawl, I guess Amazon has become and will continue to become. I don't want to lose sight of this speed versus tempo idea. And this is going to be good, excuse to talk about things like OODA loops, another interesting military analogies. What do you mean by tempo specifically of a business?

Zack: [00:11:23] When you think about any business, if you could wipe your mind clear and you come into this in a very naive way, you would think as a business gets bigger, it should get faster because it has more resources and all these things. And if you were this totally naive actor who walked in here, this is what you'd believe. Of course, any person who's worked in a small company versus a large company, knows that this is not the case, or even a larger group. HOA with three people or an HOA with 30 or 300 people, you start to see the point of diminishing returns pretty quickly. And I call this generally gridlock. Generally big systems end up in gridlock. And whether that's a feature or a bug of systems, you can argue. And some people say, “Hey, the US government's a giant system and it's in gridlock.

And that's maybe a good thing, but we'll try and stay away from that. When you think about Amazon, it defies these ideas of what should happen in terms of agility at large sizes. And I think the difference between speed and tempo is that speed implies some lack of maybe precision or insight. It's the Facebook idea of move fast and break things. And that speed, and it's not necessarily efficient speed. Versus you look at a company like Amazon, it has an emphasis on frugality, is like one of its main core principles. Yeah, 17 leadership principles. Frugality is one of them. And I think probably top of mind for a lot of people at Amazon.

And so Amazon makes bets for sure, Bezos always writes about this in his shareholders' letter in terms of if we want to make bigger bets and want to lose more money and all this stuff. And it's like, yeah, that's kind of true, but you don't really want to waste a lot of money. Frugality is baked into the DNA of Amazon. And what you see is Amazon, I think, is the best in the world. They've institutionalized the evolutionary equation. And what evolution is, is again, these all things just all breakdown algorithms, evolution is this idea of differentiate, select and amplify. Process differentiation happens by mutation. The process of selection is by fitness for given functions, and then amplification happens through sexual reproduction. Evolution is definitely not the most efficient system, but it's definitely a highly effective and it evolves these really interesting, unique designs that I think would be difficult for understated to be quite difficult for a person to come up with on their own when you see the way fish swim and all these things.

Amazon has taken this concept of differentiating, select and amplify, and they've added, so if you think about this as the branch of science of evolution, if you go back to these ideas of philosophy and pseudoscience of this idea of creationism, again, I'm assuming we're just talking about straight science here. I hope I'm not offending anybody. You talk about creationism is basically the idea of intelligent design. And everybody's probably heard the story of the blind watchmaker analogy and the idea that this was the standard philosophical argument may be up to the 1600s or something like that, where you are wandering in the woods and you come across a pocket watch and you pick up the pocket watch, and you would not assume that it was grown in nature because it looks like it was designed with some intention.

Versus you pick up a leaf, a leaf is just obviously natural. What Amazon has done is taken the best ideas from intelligent design and mixed them with the best ideas from evolution. And what I mean by that is evolution is a very dumb process. It's very smart in the long-term, but in the short-term, it's very dumb, like the stock market. What happens is Amazon is famous for this memo culture. And so for those who aren't familiar, in advance of every major meeting with every major new idea, employees have to write a six-page memo. And the memo goes through all the details of what they're proposing. What you don't see is all the selection processes happening in advance of that.

Let's say Amazon Prime, for example, started as a memo. And presumably there are many similar ideas that were discarded before they ever got turned into a memo because an employee has an idea and they start writing about it and they say, “Oh, this is dumb. This isn't going to work.” Or they talk to someone else in Amazon and they hone their ideas. And it's this idea of like steel sharpening steel. You've got a group of very smart executives at Amazon, all the way up and down to management chain and they're sharpening these ideas and finally the best memos make it to a meeting. And then these meetings, in the meeting, people spend the first 20 minutes of the meeting, reading it, and then they debate the memo and they talk about it. And they're basically all working towards this fitness function of customer value.

Anyone at Amazon tell you frugality, customer value, these are the things they're obsessed with. And the reason why all these things are so important and you had asked me, “How do these things come back to what we're doing at Stedi?” You can't just say like, “Oh, customer obsession, that's great. Let's do that. And then we'll have Amazon.” Or, “Let's just be frugal.” Well, that doesn't quite work either. All the customer principles or all the Amazon principles are in this elegant tension. And what the executives jobs are in these meetings is to be a proxy for customer value whether this idea, Amazon Prime, as an example, is going to deliver customer value while allowing Amazon to build a sustainable business over time. As you start to model this process out, now the ones that succeed, of course, the ones that fail the memo goes into-

Patrick: [00:16:25] Very quietly, yeah.

Zack: [00:16:26] Yeah, those go away. And now the ones that go well or the ones that are well-received, are given the two pizza team, which is the other thing Amazon talks about, which is the idea that when we have a new initiative, we're not going to have a team larger than can be fed by two pizzas. And again, people like, “Oh two pizza team, I'm going to bring that back to my business.” It's the whole set of ingredients that works in a very special way. What Amazon has done at this point is they differentiated through the mutation of memos. They've selected the ones that seem most promising, and now they're selecting those for some mild amplification. And those are going to be brought to market in some way, which is why, whenever you see an Amazon, in nine times out of 10, you see an Amazon release of a new thing and it's half-baked, and it's embarrassing.

And all these things that you could just not believe that a company of Amazon's resources could possibly roll something this terrible out, but they do. And then they're the best in the world at allocating the capital to the ones that work.

Patrick: [00:17:17] I want to ask about the lateral movement in all of this. Amazon started, if you think about an unbundled or limitless Walmart as analogy for what its original vision was, I don't think you would have thought about Amazon Web Services, for example. Amazon beyond its ability to eat more and more traditional retail industries has also gone into completely almost orthogonal spaces. And so I'm curious as you think about your own business and maybe as advice to other people, building ambitious people that want to build big businesses, how you think about that lateral ability that perhaps Stedi becomes something much different and maybe even more valuable you could argue AWS is more valuable than the rest of Amazon given its cashflow profile. How do you leave baked into like the DNA of a business enough flexibility combined with like enough focus, I guess, is what I'm curious about?

Zack: [00:18:09] We touched earlier on about this idea of Amazon as an organism. Charlie Munger and Warren Buffett, they talked about this idea of, we want a business so great that a ham sandwich could run it. The problem with a lot of ham sandwich businesses is that these businesses that are run this way, they're generally rent seeking. In some way, people are locked in and all the sorts. There's a lot of negative things associated with a business that runs on autopilot. When you look at a business like Amazon, which is not rent seeking, you could argue about whether it will be rent seeking in the future, it's constantly lowering prices with AWS and all these things. There's really only two ways of doing that in the short-term. And there's, I believe one way of doing that in the long-term.

The short-term way of achieving this relentless tempo of customer value, customer value, lower prices, faster delivery, all the sort, whatever that means for whatever your business is, any sort of business, is through Herculean efforts and the canonical examples of this are Elon Musk or the Collison Brothers at Stripe. Name a famous founder. And you've got this heroic effort that continues to go into building this business and through their efforts of sheer will, they build this culture and they ship fast over and over again. What you're relying on is this dynamic of successive heroes. And so other examples of this Warren Buffett, Charlie Munger, who are going to be the heroes to succeed them and have they created something, have they created a business that a ham sandwich could run and for how long? I'm not sure.

What Amazon has built, is like an autonomic nervous system. It has built into its DNA, a self-reinforcing property. And this to me is the thing that in terms of you want to build a business like this one of course, like if you want to build a business that lasts 1,000 years, you have to do it this way. What are the odds that you can get succession planning, which is like the hardest thing in the world. You can get succession planning right for so many generations, nothing, almost impossible to do. What you have to do is build a system that is not rent seeking, because if you build a rent seeking system, the business is going to fall apart once you see some skilled competitor come about, but it's also, can run on its own. And that's the problem. This is the million dollar problem. This is the multi-trillion dollar problem that I believe Amazon has solved.

How is it done? You talked about these lateral moves. Amazon, to go back to where we talked earlier. We talked about this earlier, which is this idea that things get weird when you meet the internet scale. Amazon started this unbound search. Pretty quickly what they found was, “Okay, we can't stock all the products that we want to stock. Not only can we not stock them, we can't even identify the totality of vendors that are out there and communicate with them. And if we were able to communicate with them, we can't negotiate with them to set the best prices. Walmart's the best in the world at negotiating, the best prices. Amazon is maybe average at it to start. And are they going to build this competency in-house? Can they even hire enough people to do it? Maybe not. What Amazon did is, they opened up Amazon Marketplace and for those who aren't familiar, Amazon Marketplace is Amazon's third party business, where Amazon sells tide on their website, but also you as a small business owner, or even as an individual can sign-up and start selling tide.

And so what Amazon gets with this is they've got hundreds of millions of skews, is they outsource or eliminate the necessity of being the best in the world at something. So, most businesses, you have to be the best in the world at everything that you do. Otherwise, you're going to get chipped away at piece by piece. Amazon, even if they had just the same number of skews as Walmart, they would have to be as good as Walmart, which is pretty hard to do. They brought in the wisdom of the markets. They open up Amazon Marketplace, which allows anybody to create an account. The idea that memo that was written for this. They said, “What if anyone, anywhere in the world at any time could sign-up to sell goods and ship them immediately? What would happen if that could be true?”

And so Amazon solved the problem this way with a third party, what I'll call a platform model, which is, “We own the demand,” to use a phrase by Florent Crivello wrote this amazing piece about, “We own the demand and we'll outsource the supply.” Similar thing happened with Amazon Web Services where they say, “Look, we can foresee that we're going to outstrip the ability to procure enough services and at a good enough service level and compounding speed and all this stuff. So, we're going to do Amazon Web Services.” The problem with something like Amazon Web Services and building these vertical integrations, these are the things that play for generations. You've seen this in the world of automotive where General Motors buys a supplier and they gain 20% efficiency. And then three years later, they're a captive customer for their suppliers of the efficiency plummets. And all of a sudden they have this genius idea and they're going to spit it out.

And they spit it out and it's a separate company, and then they go back and buy it five years later. And it's just like never ending thing. Amazon or credit Bezos with it Bezos is key insight was that our system will be self-reinforcing if not only are we selling services to other people, but we're selling the services to ourselves. There's two sided benefit to that. One, if we start at Amazon Web Services and we're our first and best customer, which is what Amazon is for Amazon web services we will have the customer feedback. We can't say we're internally obsessed. We know we're customer-obsessed. We'll be fanatical about collecting this information as feedback from customers, delivering what they want, and we'll get the benefits of using the same service that our customers get.

On the flip side, sure you're customer obsessed, but there's a whole different level of customer obsession, when you have to use the same interfaces and tools that customers are using. And so what Amazon did is they exposed themselves to their customer's pain, which is the fastest way we talked about OODA loops, the fastest way of getting into this feedback loop. And I think I'm going to try, and it ties back to something that not Amazon and not Walmart, this is the key to building a lasting business and Walmart did not have this. And that's why I say Walmart's one of the most impressive companies ever built, because look at it, still today, it's still cheaper to buy stuff at Walmart than it is to buy stuff on Amazon.

The selection is still amazing. Walmart is an absolute eighth wonder of the world. It's incredible, what it's done for the world. And there's of course there's negative externalities as well, but what it's built as a business is incredible. As a business, as a start-up, as an established company, your reign is going to be zero months, or it's going to be 12 months or five years. To what degree are you a flash in the pan? You zoom out far enough and you're a flash in the pan, no matter what scale you look at, unless you can solve this problem of how do we force ourselves to continuously get better, faster, cheaper, faster tempo over time.

Patrick: [00:24:33] It's an amazing idea that perhaps all Amazon is, is the equivalent of a great constitution, that through Bezos's focus on words and communication and rules and principles and all these things that they're famous for, that he's effectively externalized culture, put it into formulas. And that that is the magic. I've never really thought about it that way, but it seems to be kind of what you're saying. Is that fair?

Zack: [00:25:01] I've never thought of it as a constitution, and I love that. He's built people talk about simulations is a popular thing to talk about these days in terms of, are we living in a simulation. Bezos has built a simulation or a game. Really what it is, is it's a microcosm of capitalism, except they own it. It's like, “Okay. Hey, a distributed ledger blockchain, Hey, that's great. But what if we did that and we owned everything and owned all the benefits that accrue to it?” And so it's this beautiful mix of centralized, some centralized decision-making and capital allocation and bringing the wisdom of the markets by opening up all these things to external services. And yeah, he's I think in some ways people say, “Oh, you take away from Bezos's genius.” Because I actually think that it's interesting to note, I don't know, in your circles, but in my circles, the subject that never comes up, nobody ever says, “Who's going to replace Bezos?”

When I bring that up, people say, “Oh, it's because he's young.” It's not because he's young. People talk about that when there's someone. People talk about, “What would happen to Tesla or SpaceX if Elon Musk weren't there?” These are subjects of conversation. The reason is, is because Bezos is not, I don't believe is existential for the business anymore. And people say, “Oh my God, how could you say that? He's the best thing.” Absolutely is best thing in the world, but he wrote the program and he wrote it in a way similar, does you bring up to the constitution where the course corrects. Yeah, hundreds of years later it's still working pretty well.

Patrick: [00:26:23] Absolutely incredible. You mentioned a term and it's a good excuse to talk about some of these interesting military analogies that we've discussed before. The first is the OODA loop. This is a, in certain circles that incredibly popular phrase, an idea from a guy named John Boyd, who if people have not read a book called Boyd about his life and three-part career. I highly recommend it, it's an amazing book, but at the core of the book is this philosophy, this OODA philosophy. And I'd love you to talk us through what that means and deconstructed it a bit into its most essential, I think first, O as what you think is the critical, the orientation piece of this. Talk us through the OODA loop.

Zack: [00:26:59] Boyd is a fighter pilot. Call him the Robert Coram biographies is Boyd: The Fighter Pilot Who Changed the Art of War. And Boyd is a fighter pilot, I believe in Korea. And he was an incredible pilot. And at the time, fighter pilot tactics, these air-to-air combat tactics, was like more of an art than a science, or maybe almost purely an art, not a science. Boyd basically came back from Korea. I think it was Korea. Was it Korea or Vietnam?

Patrick: [00:27:24] Yeah, I think it was Korea, but irrelevant.

Zack: [00:27:26] And so he comes back and he sets out to write a tactics manual and he writes this tactics manual and it turns out not a single new move has been added to it, substantial new move has been added to it since he wrote this tactics manual. He's an incredibly thorough person. And eventually came up with this idea of the OODA loop. And the OODA loop is observe, orient, decide and act. And still throughout this book, I'm like, “Oh, I really don't …” It's doesn't land with me. And it wasn't until multiple Boyd books and papers and stuff that I finally came across, what I think is what made the OODA loop click for me and maybe it'll help some other folks.

If you think about a fighter pilot, a fighter pilot is flying in three-dimensional space and maybe you're a fighter pilots in an F-4 Phantom and the other person's in a MiG-25. And you're coming at them at this air speed. And the position that you are angles relative to where they are flying in three-dimensional space also. And so a pilot is sitting there and they're observing, they're seeing what's going on. And one thing that Boyd came up with is this idea of energy maneuverability theory, where up till that point, there was no equations, no algorithms for figuring out what tactics and moves were available to a fighter pilot in any given situation. And what Boyd basically figured out is if you say, “Okay, if an aircraft weighs this much, has this much fuel and has this certain characteristics of lift and whatever else goes into an airplane and it's flying at this altitude, these are …”

Patrick: [00:28:46] Your menu, here's your menu.

Zack: [00:28:47] “Here's your menu of options.” Exactly. And if you memorize these same things for MiG-25, and you're flying at X altitude, the other person's at Y altitude and the air speeds and stuff, you could start to figure out, here are the things that I can do that make it impossible for the other person to respond. So you're observing. And the next step of this is orientation. And so, I don't know, I think you've read Ender's Game. So, Ender's Game sci-fi book where all the battles take place in space, where there's no up and down. And there's famous quotes, sci-fi quotes from his book about picking the correct orientation because if your enemy is disoriented in the space and the enemy thinks that they're upside down and you have picked an orientation where you're walking on the ground, it's a very mental exercise, but all of a sudden this person is disoriented and you're able to obtain that.

Similar with two fighter pilots, I imagine. And so as this the analogy and I thought I like to read sci-fi. And you see that they're in these two different orientations. And now once you've decided on the correct orientation, I'm banked at this, and this is up, and this is down, now you can decide of what move you want to do. And you can act, which is pulling a trigger and firing missiles. All sounds reasonable, but what it turns out is that, Boyd meant for the OODA loop. And there's a more complicated drawing of the OODA loop. I think it's in Chet Richards' book, which is called Certain to Win, a fantastic book. And as it turns out, deciding and acting is implicit. He says, the observation and orientation is explicit, but then once that happens, you decide knack. If you're a fighter pilot and you've got all this muscle memory going on, not saying like, “Oh, like eight seconds, I should pull the trigger and fire these guns.”

Similar in business. You spend a lot of time thinking about a problem. And then once it happens, you decide, “Knack, this is why it never really resonated with me.” I can't remember if this was in Chet Richards' book too, but as I've been thinking about it more, it's like, “Well, they're really observing is pretty implicit also, if you think about the fighter pilot, back to this example, not sitting there, I'm going to turn my head around and look.” You're looking around like crazy, trying to find out where the other person is. You're looking at your radar and all this stuff. Similar in business, unless you're totally asleep at the wheel, you are observing all the time. What is the OODA loop really about? I don't view it as a loop because observe, decide, and act are all happening automatically, if you have good instincts.

What you're doing is you're honing your ability to orient. And that's where people talk about, Shane Parrish talks about mental models. You're talking about mental models a lot on here. You talk about all of these different things. It is pattern-matching to give you the correct orientation for whatever problem you're coming about. And when we say tempo, tempo means … Boyd talk about what's tempo, is going through the OODA loop as fast as possible. If observe, decide and act are all implicit, what this tempo really is, is orient, orient, orient. You're orienting over and over and over again. In my business, in software, I believe orientation is more important than anything else. And it's actually orders of magnitude more complicated than orientation and other businesses. I can talk about why, it might be some analogous things in investing.

And so you have to maintain this mental agility, mental flexibility, and then externalize that to the rest of your team, to the rest of your company, such that the whole team can operate at this tempo of orientation that is faster than anybody else thinks possible.

Patrick: [00:31:50] So, if you think about orientation as if you got some platform or collection of skills on the team, where you're pointing the laser, so to speak, pointing at the problem to be solved and keeping the focus there is orientation. I guess one thing I would say about observation is it is implicit, but sometimes it's helpful to remind oneself to look around. And I always think about this little business parable called OBVIOUS ADAMS. It paints this guy that's Obvious Adams is almost like a schmuck that is not especially smart, but he's just good at observing. He just happens to notice things and thinking about the system's Bible again. One of my favorite lines in there is, people in the system slowly gets sucked into that system, so they can't see anything outside of it. It becomes their world.

And maybe the observe for business people out there is really just making sure that you're not suckered into the realities of the system and don't realise that you can do something different. That's like the only thing that I keep from the observed section of Boyd's work. I'm curious, you mentioned software, this special category maybe when it comes to orientation. Could you give us an example of specifically what that means, because this is something, software specifically that I've personally gotten really interested more recently?

Zack: [00:32:58] Absolutely. And I think as you're talking about the systems Bible, he often talks about the systems delusion.

Patrick: [00:33:03] Yes. That was the phrase. I couldn't think of it.

Zack: [00:33:05] It's so great, which is you're stuck within the delusion. And it's actually one of my favorite concepts in terms of relationships, situations too. And I think business, everything's relationships. You have business relationships, personal relationships, family relationships, investor relationships. And you often find yourself in a shared delusion with another person. And so here we are, maybe we're out for dinner late and we get into an argument about something. And really when two people are getting into an argument, it's rarely about an argument versus a debate. An argument is rarely about the actual subject matter. It's more about I'm having a reaction to something and you're having a reaction to my reaction. And so now we have these reactions to reactions and pretty soon you're like way off the subject matter.

My favorite metaphor for this is that when you find yourself in this space, they're usually characterized by stress, anger, frustration, any one of these negative emotions, which are fine to feel, but these negative emotions that what you have to try and do is climb up to the next level. Imagine you can visualize a ladder. I'm trying to get up the ladder to the next level so that I can look down on the delusion. And when you finally get to the place of looking down on the delusion, it's hilarious. And this is why you say like, “A year from now, we'll be laughing about this argument.” And why is that? It's because it wasn't real. We were in a shared delusion together. Of course this is helpful in relationships, but businesses, we're all in some delusion, the pain of reconciling reality with the delusion can be substantial, but I very much agree that observation is important to get you out of this delusion.

Anyways, let me answer the actual question you asked. The actual question is, why is orientation so important in software? I think it was Marc Andreessen who coined the term product-market fit and Marc Andreessen's famous venture capitalist and founder of Netscape, and now Andreessen Horowitz. What product-market fit is, is basically what happens when whatever the start-up has built has crossed this threshold into being the function of it is obvious. So, the market is responding. The product has met the market and it is working. People are starting to hand over money, or if not, money hand over data or whatever it is they're supposed to be doing. And of course it brings in other interesting things. People have expanded on this, saying like, “founder-market fit.”

And so like you could say Stedi as an example of founder-market fit, where I'm happy to be one of the only people on the planet earth who are interested in EDI and have some experience in it. And so like, this is a good founder-market fit, but I actually love the image that comes up when you talk about product-market fit, which is, if you think about the market as this amorphous three-dimensional thing … Did you watch the Apollo 11 documentary?

Patrick: [00:35:32] Not yet,

Zack: [00:35:33] I'll spoil it for you, they land on the moon.

Patrick: [00:35:35] Or so they say.

Zack: [00:35:36] Yeah, allegedly. They have to perform this docking maneuver where the lunar or the LEM, the lunar module blasts off from the moon and it docks with command module where a Michael Collins is waiting there as the loneliest guy on earth or not on earth for the LEM to come. And they have to perform this docking maneuver. Of course, like the docking maneuver you start to wonder, you look at this and you're like, “Oh wow, it's amazing.” And then you start thinking about this, you're like, “Oh my God.” They had the entire computing power of NASA was like smaller than my iPhone. And they're looking at these dials and switches and analog systems and trying to figure out what's going on. It's a miracle that they were able to do that. Similar with start-ups. You have this amorphous market blob, and then you have also a product that is occupying three-dimensional space.

I don't know if these analogies are landing for you, but there's so many factors to a market. What are the competitors that are out there? All the market forces. What's a regulatory environment? There's so many things, your product as well, and software is particularly complicated because you are building, you and I want to build a glass. We're going to create a mug. We're going to build a mug and it's going to hold 12 ounces of coffee. And we're going to work on this as a side project. We're going to talk about this today. And then we're going to come back in a week and you're responsible for everything, but the handle. And I'm going to work on the handle. Well, chances are we're going to come back and it's going to work pretty well. What are the odds that you're going to screw up the mug part, I'm going to screw up the handle part?

And that's because what we're talking about is something that's observable. Mug is observable. It's quite simple. It occupies three-dimensional space. When you talk about software, let's pick a piece of software that most people have heard of, Slack maybe. Most people have Slack at their company. How would you describe Slack? Or more importantly, if you were sent 10 years back and you just had to describe it in words, do you think you'd be capable of building exactly what they've built, that functions all the same way? Well, you might say, “Yes.” It might not sound that hard, but I've seen the severity in the flow chart of what goes into whether or not to alert the user about a Slack message. It's this interesting, complex thing, just like down to the smallest detail, this infinitesimally, small details are super complex.

Anyways, the basic idea here is you're trying to build this multi-dimensional jigsaw puzzle. And so if you're off by one degree, it doesn't work at all, colossal failure. Versus you have a mug and it's like the angles aren't quite right. You might be like, “Man, I really need to nail this in a V2.” But it still performs its function. And you see that it's just the certain narrow miss where software doesn't have and that's what makes building software so hard.

Building Software

Patrick: [00:38:03] What do you think are the prime lessons you've learned about effective software building? So you and I have talked about this. We're building some software. You kind of quickly diagnosed when looking at it, what the key areas to focus on are, things I want to think about, that sort of thing. What are the top couple things that come to mind that you would encourage anyone that's building software let's say for the first time to consider?

Zack: [00:38:27] Well, I have this three-step thing and it just leads to guaranteed amazing software. No, it's complicated. I think we can tie this back to some of the military strategy stuff. When you read Certain to Win, which is the Boyd book by Robert Coram, the biography is good for understanding why should I care about Boyd? Why was this an exceptional person? But then you need really a lot more application of the theory to understand why all this is important. And you hear people talk about Boyd and you're like, “God, why are people talking about this?” But it's recognized, at least in some circles as the most substantial contribution to military strategy since The Art of War by Sun Tzu, 2,500 years ago. And Boyd talks about this idea of a patchwork of strategies, which we can come back to, which I love.

And he weaves together strategies by pulling them from other places. And it reminds me of this quote, which is, “When it comes to your product innovate where you need to, and then copy everything else, because there's no sense.” You don't want to reinvent your whole, unless you're doing a PhD paper and reinvent everything from scratch. What he talks about is the four elements that made the German Blitzkrieg, the start of World War II successful. And there's some German words, I'll probably butcher, but I'll try. And they talk about it as this idea of mutual trust. And I think, now I'm going to probably mix these words up, but I think [Einheit] is mutual trust. Trust is the most important thing on the team. And so you say, “Okay, I get to go back and I really have to like work on trust on my team.”

Well, how do you work on trust? There's trust exercises and we need to communicate more and all this. Well, it turns out that people have generally gotten the arrow of causality with trust wrong. And what the Germans figured out is that trust comes from bottom up. And what I mean by that is they had this thing called, there's no exact American word for it, but it's our English word for it, called Fingerspitzengefühl. What Fingerspitzengefühl is, is the literal translation is something like the feeling that you feel at the tips of your fingers. And what they mean by that is somebody's intuitive ability to do their own job. And what they found was that this mutual trust, which I believe is Einheit comes as an emergent property of having a team of people with Fingerspitzengefühl.

So, you might be a gunner and I might be a tank driver, and we don't really need to trust each other a lot in terms of having done X, Y, and Z together. If you believe that I'm great at my job, and I believe that you're great at your job, we're going to probably work pretty well together. This is particularly important when you're dealing with anything that folds out of this easy realm of tangible objects. Think about a map or think, “Okay. We're going to go in and we're planning a military operation. We're going to look at this map.” And a map is a very rough approximation of reality. If a map were a perfect reflection of reality, it would be reality. Yeah. And so what you need to do is paint a general picture of a map. We think they're going to be over here and they can have a force about the size, but when we get there, we need to make a whole bunch of decisions.

And in order to make a whole bunch of decisions, we have to have a shared consciousness. And a shared consciousness, I believe is the most important thing in software development is because we're talking about things that happen below the line of perceptibility. Meaning I can't even write code, but if I could write code and you could write code, we still can't write the code in our heads and run it in our heads and get guaranteed outcomes. I need to trust that when we break up into separate places to do our separate work, that the decisions that you're making are at least directionally similar to the decisions that I would make if I were making the same decisions with the same set of information.

A shared consciousness, I believe is the most important thing in software development is because we're talking about things that happen below the line of perceptibility. Meaning I can't even write code, but if I could write code and you could write code, we still can't write the code in our heads and run it in our heads and get guaranteed outcomes. I need to trust that when we break up into separate places to do our separate work, that the decisions that you're making are at least directionally similar to the decisions that I would make if I were making the same decisions with the same set of information.

Patrick: [00:41:52] Do you think that comes back just again to shared consciousness specifically to this constitution idea that we talked about with Amazon, that it's really just about setting and communicating a common vision at the company level? Is that the best, most effective way to create this?

Zack: [00:42:07] I think a lot of people talk about vision. And so I think vision can't possibly be the answer because when you think about this, if it were that simple, every company that was great at vision would have taken over their market by now, and we will be talking about them instead of someone else. I think it has to be more than that. Vision is what I would call a Northstar. And when you think about this, people generally need to know what I would call vision is if I went away for a month. Would people on the team be able to make directionally correct things? Well, by definition, if they're going to be directionally correct they have to understand what direction it is. And so the wonderful thing about the business that we're working on is that it's a vision that we can execute on for so long.

And what we want to do with vision is give people a Northstar they can go towards, and our ultimate purpose is to facilitate the exchange of all global trade. We believe that we can sit like a clearing house in the middle of every single transaction that happens back and forth. Now that's a nice vision to talk about, but that's not actually actionable. And so we have a lot more, that's a very low fidelity definition. We have a higher fidelity definition that talks about what our platform is and the different ways that we expect people to use it over time. Bezos is famous for talking about, “Look, it's much less interesting to talk about what's going to change over the next 10 years. It much less interesting to talk about that, than it is to talk about what's not going to change.” And he says what's not going to change. People want low prices. People want fast shipping speed. People want vast selection. The selection one maybe is questionable, I think there's value to curation, but it's this basic idea of what can I bet on over the next 10 years? And that sort of thing is the Northstar.

For us, it's things like, okay, we know that there's this set of 300 different transaction types. We know, there probably isn't going to be a new replacement for a purchase order over the next 10 years on invoices. Somebody's been around for a very long time. When you think about these things, that's the vision part, which a lot of people have. Now, what you need to work on is what's the other part of it. And that's where I think the principles and the values start to come in and the principles and values in terms of like, what are the ways that we think things should be done? So, we have this idea of what needs to be done. It's the how and the why I think, and what's the general toolkit that we're bringing to in.

Patrick: [00:44:14] So, just to close out our military analogy, section of the conversation, one of the points you made when we were talking at Capital Camp was this notion of potentially Google being like an Army, Stripe, being like the Marines and the desire to hopefully it's steady or the ability to build a business. That's more like the Navy seals. Would I take this to mean is fewer and fewer smaller and smaller team that are more and more precise, perhaps. I'd love you to flesh out this idea and maybe use it as an opportunity to create a commentary on the changing dynamics of teams and the size of teams and what they're able to accomplish.

Zack: [00:44:48] I love this topic. The thing that drives the trajectory of humanity is this basic idea, that's very simple, that we use the best tools of today to invent the tools of tomorrow. And that's what turns the whole thing. If we had to go back and build every new thing with the original tools, we'd never get anywhere. When you think about the military analogy, you can't say like, “The Navy seals is the best fighting force on earth. I'm starting a new country. I'm going to build the Navy seals.” Well, it doesn't quite work that way because, and again, I'm going to probably show some of my military ignorance here for people who have served, but my basic guess is that like, they have to be delivered somehow to the places that they're going. And so maybe that's being delivered on a plane, which is run by the Air Force, or maybe it's on an aircraft carrier or a submarine.

And aircraft carriers don't travel alone. They travel the whole group of things in a plane requires mechanics and all these things. What you need in order to build a more precise tools, the surgical instruments or what I'll call the high leverage instruments is a unit of it, a lone assassin if you could give them a machine where they can teleport anywhere in the world would be pretty high leverage Army, maybe less high leverage because it's a whole lot of resources go into it. The way this ties back is idea we stand on the shoulders of giants. Well, more than that we'll use those giants tools for all sorts of things. And Google built a whole bunch of infrastructure. Amazon, AWS built all this infrastructure. And of course then Stripe comes and says, “We're going to build on top of all the things that they've been, literally, not just the code libraries and stuff that are available because of those, but also the physical infrastructure that delivers these things.” Well Stedi is starting I don't know when Stripe is founded, but it's been quite a while.

Stedi has this tremendous advantage of starting in the next generation. And so not only are we able to build with AWS and Google, but with all the SAS tools that have come out since then, so we can use these best in breed products. We can use Stripe, for example, for payments. Stripe, couldn't use Stripe for payments in the early days because Stripe didn't exist. AWS couldn't use AWS because you can't build AWS on AWS. What we end up with, which I believe is the first time this has been possible is having an effective, like a shadow force that works on your behalf on a transaction fee basis. And what I mean by that is we use tools like we use something called AWS Cognito, and this is like a boring little corner of AWS's vast swath of services that gives users the ability to choose a username and password and login to our application.

And so we don't build that. We just use AWS as off the shelf tools for that. And AWS, I don't know if it's dozens or hundreds or thousands of people, but they have lots and lots of people working on that. And we pay some fraction of a cent for every user that we can figure. And every time they log in. And so probably 24 hours a day, seven days a week, someone is working on that section of our code base. We don't pay any debt associated with that. We have no liability associated with that. We don't have the code on our balance sheet. We don't need to improve it over time. We don't need to focus on security. We don't need to worry when they roll out the latest version of some JavaScript framework, whether it's going to work with that. All this stuff happens automatically. And we just pay this tiny, tiny fraction of a cent

Patrick: [00:48:00] It's a fascinating leveraged story, and reminds me of two other things I wanted to ask you about, there's somewhat related. The first of which is this notion of a to-hire list. And the second of which is a simple heuristic for build versus buy. You lay the groundwork for that second answer. Maybe start there. for business people out there. How should they think about whether or not to build or buy some portion of their business?

Zack: [00:48:23] This is something we've talked about a bunch, something we talk about a lot internally. I think a hundred years ago, or 50 years ago, or 30 years ago, you could get away with having portions of your business that were not something that you were particularly proud of, certainly ethical and above the line and all this, but it's not the most impressive part of your business. I think that times have changed. Things are hyper-competitive. You can look at the turnover on companies in the S&P and all these sorts of things as indicators of the fact that a company's reign is much more tenuous than it used to be. And you have to divide everything your company does you think about your company as a surface area and everything your company does is either getting better or it's getting worse. Things don't really linger in the middle because you're working against other people who are potentially getting better.

On a relative basis, it's either getting better or getting worse. Every piece of your business then needs to be getting better because presumably I don't think there's anybody's listening as podcasts who's saying like, “Yeah, I'm just willing to have my business fall behind.” So the idea with build versus buy, you look at every piece of surface area that your business has and you decide, is there someone out there who can compound this at an effective enough rate? Or maybe you could say, “Who could do it better than I can with the resources that I have, limited set of resources?” I'm quite confident that Stripe can compound the payments platform that they've built at a faster rate than Stedi code if we built it. A statement, one about the product and two about the founders and all the velocity that they have behind them.

AWS is another great example, where the rate at which their services improve, they launch a service, then they launch it in multiple regions. Then they roll out server-side encryption. Then they roll out price decreases and speed increases. You could just watch these things get better and better over time. Those are actually two interesting examples because I can't go through a list of a dozen examples. There's very few companies that I believe are compounding faster, or even at an acceptable rate. And that's basically the list of public companies that I own and pass that we have to make the decision of either build it ourselves or don't do it at all. And most often we don't do it at all. We just say, “We're going to leave this. We're going to leave this till a day, either that somebody else has done it and we could buy this primitive unit, that's going to be compounding at a fast rate, or we have the resources to do it ourselves.”

Patrick: [00:50:37] When we've talked about this, it's that last piece that has stuck most in my mind, which is very often the result is just leave it alone, don't do it. It's a great way of preventing systems bloat in the [John Gof] framework again, and makes you realize very quickly, I've kind of adopted this way of thinking that if there's not someone doing it really well, that you can hire or buy so to speak, you don't think you can do it really well. It probably isn't something that's actually critical to the business and it's a great way of reducing waste across the ecosystem, which is fascinating and keeps you very focused. You've laid out a couple of interesting let's call them grids or two dimensional ways of thinking about different companies and I'm kind of rounding now back to this, you mentioned earlier you've been in this software VC way of thinking about things. I'm going to ask about categories or frameworks for placing companies.

One that we were talking about just before we hit record, was this notion that Sarah Tavel from Benchmark shared with me about the power or the magic of companies which are delivering a better product but also for cheaper that when they see that, that has led to some fantastic outcomes. You had an interesting retort to that idea. I'd love you to lay out that simple framework.

Zack: [00:51:44] What we're talking about is this basic idea of value propositions. And so you've got the idea of more for more, which is a standard luxury products something like a Porsche or a Lamborghini maybe. You've got the idea of more for the same, which is something like Costco. You walk into a Costco, you're going to spend the same amount of money that you would have at Walmart, but you're going to get a 50% larger size. You can do less for less. Less for less is really popular with any budget airline; Spirit, Southwest, JetBlue. And of course you could try less for more, but that's not going to go over very well, which leaves this last one of more for less. And historically has been very challenging. More for less is challenging because customers often don't believe it. Almost sounds like an infomercial when you try and pitch the more for less idea.

I believe that that's starting to change a bit because this is like Benchmark's favorite business is the more for the less than Benchmark seems to know what they're doing. You just see from the magnitude of successes that have happened, that it's working. I think that it has something to do with the leverage of technology. It's a more believable value proposition, particularly when you're competing up against some legacy incumbent, which is we're going to deliver more value what we're also going to do it for a lower cost. And it'd be interesting if consumers and businesses start to get trained to say, “This is the only one that I'll accept.” And you might look at a Tesla, it might be an example of that, where an outperformed similar cars and its thing, and it class, and it's less money.

Patrick: [00:53:07] There was other interesting grid that you drew for us, which was around this idea of utility. Maybe you could describe that four-part framework and the types of companies that fit into that slot as well.

Zack: [00:53:18] Yeah, absolutely. We were talking about how to start a start-up I think, and you do hear some success stories of people who are just like, “I wanted to start a business, and so I start a business,” but that typically doesn't work out that well. What you look at in terms of, if you want to build a business that you're going to be good at, and you're going to enjoy doing over time. I think it has to fall into this in a certain spot in this two by two matrix and what I called it is the first axis, is it a positive? Does it materially improve the life of the user? Yes or no. And the other one is, would you yourself be a user for the product? And you sum this up in a much nicer fashion, what you said global utility and personal utility, low to high on both. And so if you look at something that does not materially improve the life of the user and you yourself would not use the product, this would be something like Angry Birds, something that just kills a lot of time.

The problem with that is two things. One is it's really hard to build something great. If you, you wouldn't use it yourself. Also, if it doesn't materially improve the life of the user, the chances are that you're in the business of hits you're in a fad business and you have to continue producing hits otherwise you're out of favor. If it materially improves the life of the user, but you yourself would not be a user for the product, that would be something like, I think the one I used last time was helped by fallen and I can't get up. I don't have elderly grandparents and I myself don't need that product so I wouldn't be very good at building or scaling that company because I'm not going to be able to see around the corners.

You might have something that you yourself would use, but does not materially improve the life of the user and an example of that might be, these are always the hard ones to come about because you like to think that the things that you're using are materially improving-

Patrick: [00:54:54] I think the example you used the last time was McDonald's.

Zack: [00:54:56] McDonald's. There you go. Perfect. Yeah. Are you saying McDonald's is not healthy? And lastly, is the magic category, which has materially improve the life of the user and you yourself would be the user for the product. You do see counter examples to this, you see things primarily in the category of people, just very smart people who go out and say, “I'm going to build something that I wouldn't use myself or I don't really have a personal experience in.” But it does materially improve the life of the user. And these are these mega successful consumer companies typically, but I very much view, and this is controversial, but like I very much view these consumer tech start-ups as lottery tickets and probably to many people on the show, they're listening and saying, “Aren't all start-ups lottery tickets?” And I think that that's not necessarily true and it brings us into this interesting subject where you look at the consumer business. The problem with the consumer business is that anybody can start one. Not anybody can be successful at them, but if you are, I'll use the Ben Graham quote, where he says, “Anyone can achieve …” I'm going to butcher the quote. He says, “Anyone can achieve market returns. But if you want to exceed market returns, you have to have something like a considerable amount of application.” This is typical understatement, but Ben Graham along with a certain trace of wisdom.

And it's that trace of wisdom part that I feel like gets overlooked, where you're telling people like you just need to go out and you need to work hard and you need to do all these things. When you're talking about docking this three-dimensional product into a infinite dimensional product into a infinite dimensional market, it takes more than a certain trace of wisdom. You need to be pretty good at what you do. Now, there's counterexamples to this, which is the consumer market, where you can just go out and build something and see massive adoption and sell it or not sell it as something, like an enduring company like Facebook. I believe that there's a more probabilistically reasonable way of building a start-up by using some of these techniques from value investing in the B2B space, but the caveat being you have to be willing to be a user of the product yourself and have this deep business expertise, domain expertise in order to do it.

Patrick: [00:56:58] It would be fun if some crazy enterprising student out there or somebody wants to do this to go through, let's say the top 50 or a 100 stocks in the S&P 500 and try to categorize them in these four buckets. Another question, a natural question is I think that upper right quadrant, we'll call it, would be the thing that everyone wants to work on. Another interesting question is that where's most of the value? Because I'm sure Facebook McDonald's, some of these companies may not be in the right quadrants in terms of what we'd want to build, but are nonetheless incredibly valuable because of like base protocol in humanity. One of the things you always read about is the strength of social connections, being a clean proxy for quality of life. And it reminds me of maybe the topic I was most excited to talk to you about, which is this notion of network health. I think I asked you the question, something like, “What were the best questions that you've been asked from VCs in the fund raising process?” So, maybe you could start the story there and describe why this question of network health was such a good question.

Zack: [00:57:57] So, the network health question, I talked to Alfred Lin at Sequoia, super smart guy. One of the reasons for this is he had a cheat code in our conversation, which was Alfred was, I think was the COO of Zappos. And so I sit down and like, I'm used to you just pull the wind up string on my back and we'll talk about my backstory and what EDI is and how it fits into the world. And we get on the call and he sits down. He's like, “Okay, great. You've come up with a new version of EDI. What happens next?” And I'm like, “Oh, all right. Well, I guess we can cut out the first whole bit of the preamble.” We really were able to really get into it. At one point in the meeting, he leans back and he says, “Okay, so Stedi is a network.” And what he meant by that is we have suppliers or vendors, and we have retailers. We might have third-party logistics, trucking companies and all these sort of stuff. And of course, everybody can play multiple roles too. You might be a buyer and a seller.

You might be a warehouse and a purchaser and all this stuff. He said, “So if you have this network, how would you measure the health of the network?” And I start answering. And I said, “Well,” and I stopped and started a few times until I realized that I didn't have an answer to the question at all, because I started answering with top level metrics in my head. And then it's short circuited, gross merchandise value. What's going through the platform. Oh wait, that's not health, maybe the performance of the business. You talk about take rate you can always … Stop, that's not the health of the network. And so I said, “I understand what you mean, but can you give me more? Ask me another question.” It was this nice meeting where it was not adversarial, so he was not asking this question to stump me. I gathered that he didn't have the answer to it either. And so he ended up maybe, maybe I'm wrong. Maybe he's sitting over there with my perfect answer for network health is holding out on me.

But he said, “Okay, Airbnb probably looks at things like the percentage of listings that have never been booked.” And I started thinking about that. I said, “Okay,” you think about that. It's an unhealthy node, so each one of these things are nodes. And unhealthy node, why? Well, one the seller's having the renter will have a negative experience, but the more important thing is that they've got finite shelf space finite browsing space. It's theoretically infinite, but you're limited by the number of minutes that someone's going to spend looking at this. And that's actually not my insight. There is David Ma from Composite Capital, who we started talking about this and he reads my Amazon piece and there's not many people who read that and come instantly with more information. He goes, “Yeah, well it's not infinite shelf space like you're talking about you're limited by the amount of time, the number of minutes that people are willing to spend browsing, which is this.” And I'm just like, “Oh.”

Patrick: [01:00:34] Quick parenthetical, he was with us as well in Columbia. And I think I asked him just to pick a random business and talk to me about it and like an hour later I couldn't believe that he could possibly do this about other businesses. What a fascinating guy. I'll get him at some point as well.

Zack: [01:00:48] Unbelievably smart. We had just fascinating conversations about, he filled in the whole missing piece for me, which is how the Chinese ecosystem works. Yeah, of course the person looking for places doesn't want to see all these useless places that nobody's ever going to book because they're overpriced or because they don't have photos or the photos are bad or whatever. And so I said, “I'll think about that.” I hypothesized a few metrics on the call that I think two that I came up with, I came up with two and I'm not sure I can share them. And I went and I started, I talked to someone I know at Uber and I said, “What does Uber look like in terms of network health?” And he has an instant answer for it. He said, “They look at the number of zeros. And what a zero is, is the number of people or percentage of people who open the app and have no Ubers in your area.”

Columbia, Missouri, when we were there, you try and leave for the airport at 4:00 in the morning, there are no Ubers in your area. That's an unhealthy node. The second one, and that can be solved with network density, the second one is this idea of the amount of time that drivers spend around driving without rides, which is the other side of the unhealthy node. The really interesting thing about network health and you and I riffed for 15 minutes or five minutes at Capital Camp. You just start to mind explodes. When you think about this concept, is that you can start to model because everything is networks. Most things fall into some a network paradigm. You and I right now are a network of two. We're somehow communicating with a network of people on the podcast. If you could theoretically measure this, the amount of time that people are spending zoning out on this podcast is a network health metric. If you could measure it.

You start to realize that all the things that you care about in terms of a network are actually incidental. The things like take rate things like gross merchandise value, all of these things are proceeded by these leading indicators of network health and so you may have very impressive top level network metrics, company metrics, but network health I would suppose is the Canary in the coal mine.

Patrick: [01:02:43] I think it's an incredibly portable and useful tool for talking to almost anybody, whether it's investors, business with a product, whatever, to ask them to define that idea because very quickly it's like one of those questions that let you get five levels down very fast, because on the one hand, it's very simple there's nodes and there's links between the nodes. You can define those things typically pretty easily, but then thinking about like, what would mean a high quality link? Is it frequency? Is it by direction? Is it intensity? Is it clustering? All of these things become fascinating exercises. And of course you could then apply the Peter Drucker idea of what you then start thinking about it measures, what you can manage towards. By identifying, it's almost a self-fulfilling prophecy by correctly identifying measures of network health, you're going to make a healthier network because that's the thing you'll tend to manage. Just a fascinating idea for people out there.

Zack: [01:03:38] Certainly exercise use with caution because you see the perfect examples of these gone wrong is engagement metrics. And so if you've logged out of Facebook, Facebook sends you these increasingly desperate emails and they say you have a notification because someone you once knew thought about liking something that was on your feed and you get an email notification for that. That's someone trying to improve the health of you as a node in probably not the most helpful way.

Patrick: [01:04:03] You mentioned something earlier. I love this idea of Tim Urban, the Wait But Why write calls it, the Human Colossus, this compounding set of technologies and know-how on top of which we can build further technologies. You've got this lever that's getting bigger and bigger and bigger. And yet there's this counter problem of our inability to do some big coordinated things quickly. The Empire State building was an interesting analogy that you had brought up in our correspondence. I'm curious your take there, those seem to be opposed that on the one hand we've got more ability and technology than ever. On the other hand, something like building the Empire State building quickly, or the interstate system or big infrastructure plays or massive physical accomplishments seem to be, if anything less frequent. What do you think about that idea?

Zack: [01:04:48] This is something I've thought about a lot, something that I had really interesting conversations. This is where David Ma and I instantly, where we talked about being able to go five levels deep you mentioned this, this is the thing that we went an hour deep on. The basic problem, if you think it's a problem and possibly it's just some sort of behavior of a system that's expected, is that what I frame maybe unfairly as is we can't do hard things anymore. And people say, “Well, what do you mean by that?” Well, we built the Empire State building from start to finish in 18 months and we invented all of the techniques for building skyscrapers while doing it. Today it takes five years to build an interchange in a major highway. I think the example that is nearest and dearest to my heart is if you came to me and said, “Look, I've got this growing software business and I need a financial package. I need an ERP system, an accounting system.” I would say, “No question best in bridge, you've got to use NetSuite, NetSuite founded in 1998. This is a business that raised a $100 million from Larry Ellison from Oracle.” It was a couple ex-Oracle execs or current Oracle execs at the time, I'm not sure.

Who went out and built this accounting in ERP. And today, 20 years later it's the best ERP system on the planet. And I want to put this in delicate way. It's not because they've compounded faster than anybody else. It's not, this is the AWS of accounting systems. It's clunky. It has an outdated API they're running in their own data centers, all this stuff. Of course, it got bought by Oracle, I think for $11 and a half billion, so maybe for you years ago. The question is you see these unbelievable suite of tools that are coming out, you see all the things that AWS has, got new languages, you have new monitoring tools and all these things in software. Why has nobody built a better NetSuite? That frames problem. I reject the idea that we don't do great things anymore because we certainly do. Stripe is an example. I'll just try and keep using the same examples over and over. It's a phenomenal payment system. You look at Stripe compared to Cyber source or Authorize.Net or whatever these legacy tools were and you can't even put them in the same solar system. Stripe has a suite of tools, but fundamentally they do one thing, they do payments.

And not to say that's easy. They have 1,800 Engineers and they're run by all consensus two of the smartest people in Silicon Valley who have an Army of people and read about management and read about the Apollo Space program, how do you coordinate people to do hard things and all this stuff? But the defining feature of the company is this payment system. Slack, similar thing, it's a messaging platform. They send and receive messages. Google. Google now they've got a suite of tools, but Google's got a really hard time rolling out new products, most of these have been done by acquisition. And so what, it seems that as we've traded off the ability because I don't think it's a choice to do broad things for doing deep things. We're very deep in one area, we have the world's best team. You have all these best in breed tools, but the reason why things like, and I assume the cinder blocks we make today, it got to be amazing. I don't know anything about cinder blocks, but I have to imagine you probably got factories that make unbelievable cinder blocks and we're probably more efficient at making steel than ever before and all this stuff yet it takes us 10 years to build a building.

What I call, these are called primitive units, these things that are discrete, well-defined, clean interface, interface meaning the way they interact with other things. Stripe is a clean interface for payments. We've gotten very good at these. We've moved away from this, what I'll call monolithic application. I've long said, I think it would cost a $100 million to build a modern NetSuite, a $100 million dollars in the initial round. And I think it would take, and you'd have maybe a one in 50 chance of success. There's terrible odds. Versus you want to build one specific piece of CRM or something like that, even much better odds. I'm fascinated by this question as to why I have some potential thoughts, but there's probably people who are listening to this who have more well-formed thoughts.

I hypothesized that it's a problem of complexity in that when you're building the Empire State building and you have zero ways of doing it, you invent one way of doing it and then you do it. And you've got this paradox of choice problem, where now we have all the best in breed, project management software and this cinder block manufacturer and all the stuff and architectural design firms and all these things, and everything gets slowed down even with best intentions. And I think this is the problem to solve. And there are counter examples. There's SpaceX, Tesla which are doing enormously hard things that require enormous complex coordination. One could argue though, that these both maybe fall into the category of inventing the techniques while you're doing it. Automotive, maybe not so much, but maybe in terms of their distribution network and whatnot.

Patrick: [01:09:37] Do you think that this could all be summed up with this just-in-time versus just-in-case mindset for learning and doing? Maybe you could describe that framework.

Zack: [01:09:47] Going back on this thread of reasons why school was so frustrating. School is an exercise in just-in-case learning, what I mean by that. You sit there and you learn about clouds and rocks, and God knows what in case someday I have to look outside and say, “Hey Patrick, here's our wonderful co-working space with a cumulonimbus cloud sitting out front.” Versus you get in a business where you talked about this idea of feedback loops. You start to learn things just in time. And so people joke around the most important skill is knowing how to Google, because that's what you do. If you can go out and Google and you can teach the thing yourself, you can get good at it. I'm not sure that it fits exactly in this paradigm of, we can't do hard things anymore. I think it's more of a paradox of choice than adjusting time versus just case. Although I see where you're going with it, because you've got this idea of we're going to invent the things we need just in time.

And with the whole Toyota production system is about, which is what just-in-time came from is this idea of eliminating waste. And that's the thing, I've heard that the Toyota production system is a commercialized application of the John Boyd theory, except they emerged in parallel. It's not lean and Toyota didn't borrow from Boyd and vice versa, but I think that's what Boyd missed. And Boyd if you read the biography, a lot of what he does fought waste at the Pentagon, but he never explicitly said, “Our goal is eliminating waste.” And that's what Toyota has done or Toyota did. And Toyota did that but with this inversion mechanism of Munger's the famous inversion thinker who says, “I can take a problem and put it on its head.”

And what Toyota basically said was, “Okay, instead of taking things and pushing them through the production line, we're going to instead start with basically what the customer orders. And then every time they remove a wheel from the cart, the cart is going to go back to the wheel manufacturing center and they're going to produce another wheel to put on this cart.” So it's this idea of just-in-time, exactly when you need it. The whole thing gets done just-in- time. And the whole idea behind this is eliminating waste. I could see how building something, they say a start-up is jumping off the cliff with the parts of an airplane and trying to build it on the way down. That's part of the reason why start-up works is also a part of the reason why a lot of them fall at a 100 miles an hour directly into the ground. But it's this idea of doing things just-in-time. When the just-in-time things work, they work like magic.

Patrick: [01:11:54] We're going to close with a couple of fun questions. The first is, to ask you what your favorite question is and why.

Zack: [01:12:00] So my favorite question is this question of what's not being said. I think that this is the question that kills companies, kills relationships, kills products, kills investment strategies. Often what you find is that everybody thinks that they're holding secrets and the secrets that they're holding are not actually secrets. Example you're standing in a group of people and you're thinking somebody is like, “Man, this guy really hates me. This guy just can't stand me.” If you don't want to call attention to it. Well, the chances are that guy's probably thinking, “Wow, this guy talks too much.” Or “This guy reminds me of somebody I hate.” Or something like that. And so we're often holding counterweights or counterpoints in the same secret, or we're holding the same one, which as you say like, “I'm worried such and such, person's not working out. I'm worried that they're not in the right role.”

They're probably thinking I'm worried. I'm not in the right role and I'm not performing as well as I should be also. I look at my job as a founder is basically harvesting our collective unconscious as a group for these things that are not being said and saying them, it's actually pretty simple. And it works really, really well. And I try to use that as much as possible in relationships, in friendships, in business and everything. And you could find that like when we're talking about go back to the systems delusion problem, you can almost always discover the system's delusion by talking, by talking about these things that are not being said by naming them and then they basically go away.

Patrick: [01:13:19] I love that idea you have, a physical ladder climbing up to get the perspective? Another question that works here is what is the hardest conversation I could have right now that I haven't had. I think that always yields, one it's interesting how easy it is often for people to answer that question and how hard it is to then go and do it simple but not easy. The second to last before my traditional closing question is you mentioned this a little bit before or at least you hinted at it when you responded to my question about focus and how if everyone believes that it can't be the right vice, about why most commonly heard advice is often wrong.

Zack: [01:13:53] I think about Google AdWords comes to mind. I know a dozen people who got rich off of Google AdWords in the early-2000s. If you heard about Google AdWords and you had anything to do with anything, selling products, anything you can get rich on Google AdWords at that time. And of course, since then there's been Facebook advertising, Instagram advertising, Influencer advertising and all this stuff. The unifying feature of all these things is that by the time you've heard of them, it's all priced out. This is why Eric from Benchmark said, “An innovation product tied to the innovation distribution strategy,” because otherwise you're just giving away most of your gross margin to a platform, unless you are the platform, then you're doomed to give it away.

Most advice that you hear about it's all backward-looking. Blitzscaling if you heard of the term blitz scaling is the popular thing in Silicon Valley now. Blitzscaling is so obviously the right way to build a business. You can't be questioned. Look at the companies who've done it. Uber, Slack, Airbnb relative newcomers, someone like Flexport, but it can't be that simple because you can't just go out and buy the latest book and implement the advice and get rich it doesn't work that way. And generally, we'll go back to the systems Bible book, the Army is fully prepared to fight the last war, that basically sums up all advice. This is backward looking. Blitzscaling was the absolute correct answer over the last 10 years. It's a back testing thing. And so that doesn't mean that the opposite of all current advice is correct. I remember the Ed Thorp book, my big takeaway from that book, probably something you knew already, but for me it was the one thing that book is, never invest in anything in which you don't have an edge.

If you don't have an edge in it, don't invest in it. And that changed everything for me in terms of like, why would I do anything other than start an EDI company? It's the thing that I had the biggest edge in. When you start to think about advice, unless you need to have some insight, some edge that gives the ability to do things in a different way. People say, we have to be contrarian and right. But you have to do that many, many times in a start-up in order to have outsized results. Go back to the military stuff, which I know we closed out, but I want to open it up just a little bit at the end. In the Art of War, Sun Tzu says, “If you prepare for battle and you build a big army and you go up against another army and you win, you have not one you've done …” In modern terms, they call it attrition warfare, which is the dominant strategy for a long time. When you think about attrition warfare-

Patrick: [01:16:21] The trench warfare World War I or something.

Zack: [01:16:23] Or yeah. Civil War, you line up in lines and you shoot at each other and whoever is left standing is the winner. Sun Tzu said 2,500 years before that is not really a good way to win. When you want to win. What he said is you want to bring a boulder up to a high hill and roll it down the hill, is what the war should be like, onto the troops below. And to me, you hear Blitzscaling, Blitzscaling which sounds a lot like Blitzkrieg is not the same as Blitzkrieg, Blitzscaling is attrition warfare for start-ups. Was the correct answer before but now that it's commonplace is not the right way to do things, you just as a start-up you can't hire a 1,000 great engineers. Maybe you can. It's so probabilistically difficult that it's not a good strategy to do. When I think about these things, I'll steal the title of that Chad Richards book again, which is you want to do things in a way that makes you certain to win.

And that's what value investing talk is about is like, “Hey, screw being right. Let's just figure out ways that we can't possibly be wrong and we'll get rich.” And I think there's ways to do that for start-up. For sure there's ways to do that for a start-up, to maybe not get a 100%, but to approach 99%, limiting burn. Overpaying people in order to get the best possible people, extreme focus. There's a whole menu of things that you can do, but you have to have taste. That's hardest part, is this idea of what Ben Graham talks about, that certain trace of wisdom. It has to be more than a trace of wisdom. You have to have the taste to look at a whole menu of strategies and say, “This one is right, and this one doesn't work for me.” You can't just go out and read seven Powers and say, “I'm going to do this one. And this is going to be in my business and Blitzkrieg is good,” and so such and such. It's about weaving the correct Patrick of strategies to find the perfect orientation and then you take off.

Patrick: [01:18:00] It's a fascinating place to close and what's highlights the nuance and difficulty of building a successful business. My closing question for everybody which is to ask what the kindest thing that anyone's ever done for you is?

Zack: [01:18:10] I have one that's top of mind based on the conversations that we've had tonight, which is when I was starting Stedi, I didn't know anything about start-ups. I had an auto parts business, though it was business it wasn't a start-up is different in my mind. Didn't know anyone in the world of venture capital practically. And I said, “There's this one guy on Twitter who likes my tweets a lot and I'm pretty sure he's a VC.” And so I look him up, his name's Zach Ware at the time he's at a fund called VTF Capital. I sent him to like a two-page DM, which is like something we didn't get to talk about as this idea of all the common advice in terms they say, “Don't write long emails.” Well, it depends on how good of a writer you are. Sometimes if you feel like writing is an edge you have, maybe you should write long emails.

I wrote this long thing. I'm not even sure at this point how coherent it was. And he said, “Let's talk.” And his background is at Zappos. He sort of knew the EDI terminology and everything and totally got it. And I didn't have a deck, I didn't have a name. I didn't have a website and he said, “I'll commit for a 100K.” First call, that opened up unbelievable doors. They eventually went on to invest more than that. Unbelievable doors and if I could cheat by offering a second one, which is the second thing that sort of, I forget who was telling me about this. It might have been Brett from First Round, with this idea of kindling. The beginning of a relationship, the beginning of a business is very delicate. And it's like, you ever see a Castaway Oysters, my agency's just trying to blow on that fire and get it to work.

I think people don't realize how fragile these things are because the most little time they fall apart before they even happen, it's like the Fermi paradox. And so I got this first commitment from Zach who believed in me and has believed in me more than anybody, more than anybody had in my entire life. He just the conviction, boom, “I believe in you.” And then the second one was Ryan Petersen from Flexport by happenstance of too long of a story to tell here, I ended up with some early adviser shares in Flexport. I was an early Flexport customer. And Flexport is gone on to just raise a billion dollar round from SoftBank. And I told Ryan, “Hey, I'm doing this EDI thing just figured I tell you about it.” And he said, “I'm pretty broke right now, but I at least I can do 10K.” And this is not those three years ago or something like that. And I said, “okay, it's amazing.”

He texts me. And he says, “Here's a link to my Crunchbase page. Just tell me anyone else you want an intro to.” And that was, I mean, just like what an unbelievable thing of saying, like, “I'm going to open up my whole cap table to anyone you want an intro to.” And response to that kindness. I said, “Here's a list of all of them. Please send me an intro.” So, those are the two things. You look at these things, the trajectories of life and Stedi has been the most meaningful thing I've ever done and I'm only two and a half years in because it's changed my whole network health in terms of the people I've met, quality of thoughts that I'm having because I'm being pushed to such a limit of what I'm able to do. I can all trace back to those two people who said, “I believe in you let's do this,” and opened up their networks to me.

Patrick: [01:21:03] Taking a chance, a common answer for kindness. Well, this has been awesome. Thanks for all the time, the variety of crazy interesting topics. I hope to do it again.

Zack: [01:21:10] Yeah. Thanks so much for having me.

Transcript by @ScarrottKalani