How to build a company you’ll run forever | Zack Kanter (Founder and CEO of Stedi)

Zack: We've decided we're going to eat glass. That is what sets us apart. We'll go to the ends of the earth to do things the right way, even when it's not economical and it doesn't make sense.

Brett: For today's episode, I'm sitting down with Zach Kanter, founder and CEO of Steady, the only programmable healthcare clearinghouse. Before Steady, Zach ran an auto parts company and that experience heavily influenced his approach to systems efficiency and quality.

Zack: It took four-and-a-half years for us to launch anything publicly. We threw away everything that we had built every line of code, maybe eight times.

Everybody's had somebody say to them at some point, quality, price and speed. You have to pick two. That is like the most losing mindset.

Brett: Along the way, he's developed a philosophy about business that's both pragmatic and contrarian.

Zack: When you have bootstrapped a businesses, you're saying, I'm going to be constrained by capital. Being able to remove the fundamental constraint of capital as the limiting factor of a business is like the equivalent of going pro. All of a sudden you have a million dollars in your bank account, you're immediately faced with a question of, "How good am I?"

Brett: In our conversation, we dig into how manufacturing principles translates to software.

Zack: In software, you kind of have it on easy mode because the margins.

Brett:

What it really means to build a high-performance team, and how Zach thinks about creating enduring advantages in a changing market.

Zack: On everybody's first day at Steady, I say, "Welcome to Steady. Everything's your fault now."

Brett: Let's dive in. So what's the story of starting your first business?

Zack: Well, I was getting my first car, which was a 1995 Caprice. I don't know if you know what a Caprice looks like, is the old yellow cabs and police cars, the kind of big boat-y ones, and it's the cousin vehicle, so it's a Chevy Caprice is a cousin vehicle to the Impala SS, and the '94 to '96 Impala SS is kind of the last of the great four-door rear-wheel drive V8 American muscle sedans. So I was getting my first car. My first car was the $3,900 or something like that. And so it was in need of repair, in need of a number of things. And as I got the car, I started going on this car forum, the ImpalaSSforum.com, and I was basically looking for parts that I might want to put on my car. I didn't really know a whole lot about how to work on a car, but I was interested in learning and reading about this stuff. So I got kind of obsessed with it eventually was a satellite trade show of this show called the SEMA Show, which is the big show every year where all of the automotive kind of after-market business gets done every year. They had a satellite show in New Jersey about 45 minutes from where I grew up. I was 16 I guess at the time. And so I made up a fake business card so that I could go to this trade show and I basically went around from booth to booth and I asked every single person, Hey, do you have any parts for this car? And it was kind of an obscure car and not really a hot car like a Camaro or a Mustang or Chevelle or something like that. And finally, 20, 30 booths into the show, was not a huge show. I met a guy who knew a guy who had developed ball joints, heavy duty ball joints for the car. And now I didn't really know what a ball joint did or what it was, but I was pretty excited that he had something for my car. And so I reached out to this guy after the show. Interesting thing about the car industry, there's the kind of need-based repairs, which is somewhere as a $200 billion market. So your tire blows out or you need a new muffler or something. And then there's the want-based purchases, which are the upgrades. And I was more interested in the want-based purchase and I think it's a 40 billion market or at least that's what it was when I was in the industry. So I met this guy and called him afterwards and I talked to him and he said, I talked to him for 45 minutes about these heavy duty ball joints that he had made. Okay, it sounds like a great product. I'd like to order a set. And the guy goes, "What do you mean you want to order a set?" I said, "I want to order a set for my car." He says, "This is a wholesale business. You have to buy $2,500 worth of stuff." So I kind of did the math. I figured out that $2,500 worth of stuff was 25 sets of things. It was enough to do 25 cars. So I signed up for a credit card and I bought $2,500 worth of stuff and I basically packaged it into a kit, each one of these different pieces, 12 different components or so, and I packaged it into a kit and listed those for sale on the Impala SS forum and the kind of group purchases section, and I sold out of the kits. And so I got mine and I sold 24 other kits and that started to become a real business.

Brett: So to contextualize the journey of building the auto-parts business, tell us about the few chapters over those 13, 14 years and then curious to get your perspective on a few things.

Zack: In the beginning I was selling direct to customer, so I had a website, took PayPal and I would sell on these online car forums. I started selling these kind of auto parts for this one specific car, these few different years of Impala SS and Caprice. And I sold small, it was nice fun money, but it was not real business. And I skipped my senior year of high school and I went to college and when I got to college, I met a guy who was also interested in cars and he said, "You know what you should do is you should develop ball joints for snowplows because the snowplow, the additional wear that comes from having a snowplow on the front of the truck actually wears at the ball joints very quickly." So I asked him what he was using for a snowplow and he was using a Chevy S-10 two-wheel drive truck. And so I said, "I'll make a note to go look at that and see if I could maybe have those made," or something like that. And so I went and I did the research and it turned out that the exact same ball joint that was used on the Impala SS was used on the Chevy S-10. And it's because when GM or Ford or whoever goes and makes a new vehicle, they make something, like I said, obviously in hindsight they don't go and make every part from scratch. What they do is they look and see, "What do we have that's a similar weight, that has similar angles and whatever all the other requirements are," and they take that off the shelf and they use it. So that was an eyeopening thing for me. I never thought, "Hey, what else does this vehicle fit?" There's an interesting kind of nuance to auto parts in that one part can fit many vehicles and obviously one vehicle has many different parts, so it's a many to many database relationship that's not the easiest thing to model. Because that is difficult to model and it's very different from how you might sell a glass or a fork or a watch, which is like it might vary based off of having four prongs on the fork or five prongs or being up dessert fork or whatever, kind of different fork, auto parts both have those different types of dimensions, but they also have this different fitment criteria to it. And because of the complicated fitment criteria, the e-commerce penetration rates in auto parts are very, very low. So I don't know the latest numbers, but when I was in the auto parts industry a number of years ago, the penetration rates in automotive were second only to grocery, so it was like a 4% e-commerce penetration rate for auto parts at 2% e-commerce penetration rate for groceries. So I found out that it fit these other vehicles and I started listing more parts on my website.

Brett: You were contract manufacturing them or no, still looking, just being a middleman.

Zack: It was still, I probably had modifications done to it. There was a number of different changes. There were plastic bearings. I switched them to metal bearings, which were better. I made them greasable and I changed the coatings and a bunch of different things like that. Maybe six months after I did that, I got a call from this guy one day and he said, the guy in Indiana, he said, "I'm making a production run of control arms and I want to buy, I hear you make the best ball joints in the world and I want to buy." And I said, "Okay, how many do you want to order?" And he said, "I want to order 500 sets." And 500 was I was floored. This was more money-

Brett: And this out of your dorm room?

Zack: Out of my dorm room, yeah, absolutely. He said, "How much would it be?" And I think I sold them for 20 bucks at the time. So I said it would be $13. And he said, "How about $12?" I said, "Sure." And so he placed the order and that was what it was 500 control arms, probably 500 sets. So it was like a $20,000 order or something like that, which was just an instant $10,000 profit or something. It was more money than I had ever transacted at one time. And that kind of woke me up to the power of wholesale. I continued on this supplying these individual small manufacturers for a while. This is much longer story, but I eventually got into one of the major auto parts retailers company called Summit Racing Equipment. I learned that the hardest thing to do is to get your first customer in any one of these segments. Okay, consumer, it's not that hard to get your first customer. There's enough people out there in terms of wholesale, everybody wants to know who are you supplying now? And so you have this problem of, okay, you meet your first customer and they-

Brett: Cold-start problem.

Zack: ... [inaudible 00:09:47] cold-start problem, exactly right. And so starting with these kind of private label jobs or working with these niche players who had a big name but relatively small volume and were willing to be progressive was very helpful. And then one of the ways that I ended up getting into the first retailer was I went to them and I asked which of our competitors catalogs you in? And I said, "I'm not in any of them right now," but I said, "Do you have a copy of your catalog?" And I flipped through their catalog and I circled all the different products that they were offering that had my parts in it and that was one of the several kind of chips that fell into place of convincing them to carry the product line. And then once you get that, within three months I had their biggest competitor. Within a month after that I had Amazon and within probably a year after that I had O'Reilly Auto parts and the kind of dominoes just fall from there.

Brett: Through all the ups and downs over that 12, 13 years of building the company, what ended up causing you to want to sell the company?

Zack: I had bootstrapped that business, self-funded the business, which means basically funded out of revenue, funded out of debt. I hear a lot of founders talk about the venture capital backed founders saying, "Oh, next time I'm going to bootstrap or You should bootstrap your business, you shouldn't raise venture capital." There are a lot of downsides to bootstrapping and the fundamental thing that I think about when I think about bootstrapping is that when you have bootstrapped a business, you're making a statement of saying, not in every case, but in the vast majority of cases you're saying, "I'm going to be constrained by capital." Now, there are examples of business that just generates so much capital that they're not. Bootstrap business just generates so much capital, they're not constrained, but by and large bootstrap businesses are constrained by capital. So I would have this cycle where I would have millions of dollars of new products that I wanted to make, but I would have $20,000 of free capital that I could deploy. So you go and you design $20,000 worth of stuff. You place the order, it takes three months to get a sample, you get the sample, you approve it takes four months for the production and shipping and everything, and then it takes you some months to earn your money back on that, get your initial cash back out, and then you go and you take your $40,000 that you have now and you deploy that and you kind of do the cycle over and over again. And I did that enough times. You get to the point of saying, "What am I doing here? I think that I can do a lot more than I'm doing. I think that I have a huge bench of ideas that could be deployed," and it's kind of a question of how do you want to spend your time? Do you want to spend your time owning everything %100 but being able to work to 2% of your capacity or do you want to be unconstrained by that and figure out what you're really capable of? And so fundamentally venture capital is this idea of which I think is the most wonderful thing in the world. You raise an initial round of funding, we raised $1 million dollars in our first round of funding and all of a sudden you're faced with, you have $1 million dollars in your bank account, you're immediately faced with the question of how good am I, which is if you don't have engineers, it's like how good are we? I'm not good enough at recruiting. You're hiring engineers, but it's not clear enough what they should build. You're not good enough at product. The product is there and you have engineers, but you're not good enough at selling it initially. That's a hundred percent your problem. Now you get to this place, you hire four or five, six people and you realize, "Well, I can't hire too many more people or else we're going to have too short of a runway," and now you're not good enough at fundraising. Of course, another way of framing that as you're not good enough is it's within your locus of control. So as a high agency person being able to remove the fundamental constraint of capital as the limiting factor of a business, to me it's like the equivalent of going pro. I think that there's vastly more cases in which venture capital makes sense for a high-achieving person than for bootstrapping.

Brett: Why do you think this idea of bootstrapping being the way to go, at least in some corners of the internet, it's almost turned into sort of this ideology?

Zack: Well, look, to some extent the grass is always greener. I think there's also a lack of granularity of how you think about it, which is basically what are the things that you don't like about raising venture capital? People say, well, we were pressured to grow really fast. What does that mean? Does that mean that you were going to be fired as the CEO if you didn't do something? That's an issue of board control. Was it just uncomfortable that people were pushing you to move fast? It's like, grow up. You're going to be pressured for all sorts of things from employees, from investors, from customers, from peers to do things. And I think you have to have a backbone and decide what are you going to do? What are you going to stand for? What are your principles versus your preferences? And kind of own those things. There's an element of ownership, which is people say, "Oh, it'd be nice to own a hundred percent of the thing." And I think, look, that's a math equation. You can't be obsessed with owning a hundred percent of the thing. Would you rather own a hundred percent of a 10 million business or 10% of a billion dollar business? Of course, I think unless you really have some obsession with ownership. But I think most of these things are confusion about ownership versus control.

Brett: And why do you think people get confused in that way? Is it just the way the brain tends to work?

Zack: When I was starting Steady, I had a lot of friends who had started companies and I noticed something interesting, which is that founders wanted to often leave the company that they had started. So you see people romanticize about bootstrapping, but you also see people romanticizing about starting over from scratch again. The question is like, what are you hoping to do differently? And basically when it came down to it and you probe these questions of what they were looking for is basically they disliked the business that they ended up building. And to me that's kind of a crazy situation to get yourself into because it's your fault. You've built this business, you made all the decisions and everything. Yes, okay, maybe you had a board who's forcing you to do something you didn't like to do, but that was your decision to give up board control. Now you can get yourself into difficult situations where you have to give up board control, but again, it kind of all compiles back down to this is your fault. Just don't put yourself into those circumstances and you're not going to end up having those problems. And so I love the Charlie Bunger quote where he says, "All I want to know is where I'm going to die, so I know not to go there." I thought about that from the early days of saying, "Okay, what are the things that, as I was thinking about this question of do I want to pick a new business that I think has faster returns to cash or do I want to do a venture capital backed business?" I basically said, "Well, what are the things that I might end up regretting? What are the reasons why I would want to leave this company?" And as long as I decide not to do those things from day one, I'm going to end up building a company that I love. In other words, just don't build a company that you want to leave and you're going to want to stay there for a very long time.

Brett: What are the things that you listed out? And I should say do you think you could have listed them out if you didn't first build the auto-parts business?

Zack: I do think so, yes, because I think it wasn't really learnings. Look, I didn't learn that much about venture capital from the auto-parts business except for the fact that I've learned that I didn't want to bootstrap. And I think that having gone through that taught me, "Okay, how do I want to spend my life is I want to spend my life working on hard problems with smart people and I want to look in the mirror every day and be faced with the question of how good am I?" I think that that is the ultimate dream of a founder is to figure out what my limits are. The decisions basically that I made upfront were, well, one thing I liked about bootstrapping is that I didn't have a boss and to me I kind of went through this agonizing question of do I want to raise venture capital because is that going to mean having a boss and you can just invert that like everything else and say, "Well, how would I have venture capital investors where it didn't feel like having a boss When I could have talked to other founders about the circumstances that they had found themselves in most common things were that they promised things that they couldn't deliver on, got themselves into bad circumstances and didn't explain those things adequately or upfront to investors. They gave up control early, brought on investors that they didn't like. And so I worked backwards from this list and I said, start with I'm only going to tell investors good news after it happens and bad news either hopefully before it happens or as soon as it happens worst case. And talk about that first one a little bit. That I guess is something that I learned from the auto-parts business. My largest customer for the auto-parts business, which was like a customer that when I landed it, it kind of put the business into a different stratosphere. It took me five years to land that customer, and the other four times that I almost landed it, it was like on the one yard line and then, poof, something happened. You said something they didn't like and they would disappear for a year and you won't answer emails for a year. A year later they finally answer your emails and I kind of realized that of B2B deals that seem 99% likely to close, 50% of them fall through. And so I've Angel-invested in 50 or 60 companies, and so I get a lot of investor update emails and through angel investing I've realized I definitely don't want to be an investor, but you get these emails and they're like, "Oh, we're in a POC with this Fortune 500 company and we have this much in pipeline, and we're this," and I'm like, "Man, why are you telling me this? Just tell me about the things that happened before." I don't need to know about the things that might happen or end up not happening. And I think there's just almost pure downside in listing these things out to investors.

Brett: Was there anything else you listed out that were sort of non-negotiables in trying to create a company that you can run forever?

Zack: Well, yeah, I mean the run forever piece is interesting. I kind of realized with the auto-parts business and I had started a lot of other side projects while doing that, I was bored with it, that the returns that you get from working on one thing are like when you split your attention, you don't get 50 50 and so you say like, "Oh, we're going to do this and this. I'm going to focus on this and this," is never going to really work very well. With the auto-parts business. Having done all these things, I decided that I wanted to sell that business and start working on the most complicated problem I had ever come across, which is this problem of business to business transaction exchange, which I had come across through the auto-parts business and knew that I didn't want to retire. I never wanted to work for anybody else. I didn't particularly like starting and selling businesses. I like compounding and I think it comes in many ways, it's compounding of relationships and friendships and whatever it is. It's like yes, the beginning of something can be exciting and the end if something could be relieving, but by and large, the muck in the middle is where the value is built. I didn't want to retire, forget if I that part already, but if you kind of put these things together, which is I don't want to retire. I don't want to be an investor, I don't want to start and sell companies, I don't want to work for anybody else, you can default. You realize that what you want to do is run one business for a very long time. But when we build this business, it's gonna be, not a one of end business, but an end of one business.

Brett: Is there any other hidden knowledge that you picked up through running the auto-parts business, the types of things that are important, ideas that would've been hard for you to figure out had you not had that chapter?

Zack: There's this tremendous alpha that can be found from doing things better than other people. There's this great, I'm not a sports fan, I'm in particular not a basketball fan, but there's this great converse ad from around when we grew up that had Larry Bird in it, and it's a picture of him diving for this ball that's going out of bounds and he says, "It makes me sick when I see a guy just watching the ball go out bounds." That extends to everything, the things that I did to build the auto-parts business. There was no breakthrough in manufacturing technology or distribution technology or anything. I basically said these parts are being sold very poorly on the internet and there are some basic things that can be done in order to make the products better. What were those basic things? You look to what the customers are saying, customers saying, "The original parts that were made on my vehicle had a metal to metal bearing service and I want it to be metal to metal, and the ones that are sold today are plastic." They complained about the parts rusting and they complained that the boots on the parts would tear. These are all solvable problems. This is not something you have to go learn how to land a rocket in order to fix. And then the pictures that you would see online for these products were scanned in from the old books that were in parts stores. And so these are multi-billion dollar market cap companies that their photos were, the e-commerce sites were scanned in copies of books. It was grainy, it was black and white. And so my basic hypothesis with the auto-parts business after I was through that initial phase of it was that by having a high quality product was better on these few clear dimensions. And by having excellent photography of each one of the products and clear explanations and then following that up with excellent customer support, that I could take some percentage in the market without doing any advertising. And that turned out to be true. The business was immensely profitable. It was like it had software margins that people told me were not possible to do in that business. I met, actually tried to raise money for the auto-parts business in maybe 2014 or 2015, and I talked to some guys who did a mini family office, private equity VC-type thing, and they asked me to put together projections, which I had never done for the business, and I projected the business would make such and such profit at such levels, and they actually kind of laughed me out of the room. They actually laughed when I presented it to them. They said, "You just don't get it." They don't. And I said, "What don't I get?" They said, "Well, when you get to a certain size, you're going to need to do more things. You need to hire more people. You're going to do all this stuff." I was like, "But I've built software for that. That's what the software is for." They said, "Yeah, it's in theory that's true, but when you get there, it doesn't actually work." I went out when I sold the business, I went back and I looked at those projections that I had sent them and the actual EBITDA margin that I achieved, the actual real profit, not just a fake EBITDA margin, was two and a half times what I had shown to them at that time. So it was wildly pessimistic. When I later read about the Toyota production system, you think that some sort of a major breakthroughs and they're developing these cutting edge things and robots and all that. And really what the founder of the Toyota production system said and what people said about him is he took basic ideas and he carried them through to an extreme degree. And basically that's the philosophy that I brought forward to Steady, which was like there is, yes, other people have attempted to solve business to business transaction communication before, but I can look at it and I can see there's hundreds of things of obvious pieces that they're not doing. And my hypothesis again is that if you fix those things, that emergent property of fixing all those basic pieces is going to lead to something that has a much better outcome.

Brett: Switching gears slightly and going back to talking about the journey of building Steady, the company sort of had this interesting path where it took something like five years to get to your first million in revenue and then you sort of went one to 10 in a surprisingly short amount of time. Maybe you can kind of contextualize the journey and sort of explain some of the higher order bit insights of that seven or eight year journey you've been on.

Zack: Now, being a healthcare clearinghouse, as we fundamentally sit in between providers, which could be your individual doctor or dentist's office, if they've got an independent practice, it could be a large health system, it could be an ambulatory surgery center or anything in between. It was a pretty broad spectrum of what these customers might be and they need to submit transactions to payers, insurance companies, they're called payers in the industry, Aetna, Cigna, Blue Cross Blue Shield, whoever it might be. And then they need to process claims and the transactions that come back in response to that, like claim acknowledgements and claim status, electronic remittance advice, which is record of the payment that was sent, those transactions all get processed by these clearinghouses, but that's not where we started. Where we started was a layer underneath that, which is with this archaic file transfer protocol set of standards called EDI, Electronic Data Interchange, and EDI was what I had encountered in the auto-parts business when I started selling to these various retailers, I sold to O'Reilly Auto Parts and Amazon, and when the volume got such that it wasn't reasonable to automate manually into the transactions anymore, I reached out to the different retailers and said, "can you send me your API spec so I can automate the transactions?" It's probably in 2011 or 2012, and they all wrote back and they said, "What's an API?" Things have gotten a little bit better since then, but broadly, I found out at that point that the way that these businesses integrated was using this technology called EDI, which was developed in the sixties for railroad communication, popularized in the eighties by Walmart. And then when Amazon came onto the scene in the '90s, they said, "Okay, what's the fastest way for us to onboard all of Walmart suppliers, Procter and Gamble and Rubbermaid and whoever else might be, Nike." And the fastest way for them to do that was to adopt the same standard of the Walmart use. And so you fast-forward to today and the whole world of retail logistics, transportation, supply chain manufacturing anything revolving around physical products works off of EDI. And so the way that EDI works is it's a file that Amazon or Walmart will send you for a purchase order and then you send back a purchase order acknowledgement and a ship notice and an invoice and these various transactions that flow back and forth. When I found out about EDI, I was very excited. I had read the 4-hour Workweek not too long before this and I said, "Okay, I need to automate everything." And so when you look at this list of EDI transactions, it turns out that it's not just purchase orders and ship notices and invoices as inventory feeds, I can automate my warehouse shipping orders, I can automate the load tender, which is where you tell the warehouse, the warehouse tells the truck to come pick the product up. You can file your tax returns via EDI, get your bank statements via EDI. So I said, "In six months I'm going to have this whole thing humming." A bunch of different, I was running on NetSuite at the time, implemented NetSuite early and there were a lot of EDI platforms that were out there that were pre-integrated with NetSuite and pre-integrated with these various trading partners, the Amazons and O'Reillys of the world. And I signed up, interviewed three of them, spoke with three of them was my first piece of software I had really bought outside of NetSuite that required an implementation and I reached out to three of them and I chose one of them and they walked me. This was great onboarding kickoff call. They assigned somebody to my account, seemed like a really great process and they said, this is going to take 30 days, this is going to take 30 more days and this is going to take 30 days and in 90 days the whole thing's are going to be done. 60 days goes by after that kickoff call and I reached out to them and I said, "Hey, it's been 60 days and just wanted to check in how we should be thinking about scheduling that last call." And they wrote back to me and they said, probably took them 10 days to get back to me. And they said, "Oh, we reached out to your trading partner and they never got back to us, so we haven't started yet." And that was where I first experienced this level of ball dropping in the software world. And you find out is that these, when they said they were pre-integrated, what they meant was we have done an integration once before. It doesn't mean that this is not like a plugin, this is not like a Zapier integration. I went through three of these. One was worse than the last and the last one was an 18-month implementation and everything could possibly go wrong, went wrong. And to kind of go back to our earlier discussion, there's just so many of the things that they were doing, I was like, I'm pretty sure these are very fixable. This is not rocket science. So I ultimately had my own EDI system built and then that automated the auto-parts business for the next five years. And in building that business, I realized, or in building that automation platform, I realized there's a big opportunity for this market. It's really, really hard to build an EDI platform and that's what I'm committed to do and we're going to win this by going through the muck for longer than anybody else. And we got exactly what we wanted. We did indeed go through the muck for a very long time. It took four-and-a-half years for us to launch anything publicly. And what you don't see during that time was that we threw away everything that we had built completely every line of code maybe eight times and you just get to a place and you say, "Okay, this is not the way," and you have to start from scratch again. And I think that you gain more of an appreciation for how the other EDI businesses ended up where they are because they got to one of those paths and they said, "All right, this is going to gets us kind of part of the way there and we'll fix it later." And they stayed there and ended up on this kind of dead evolutionary branch.

Brett: What are your reflections on this idea of building products for yourself? Because I think part of the reason that even though you thought the degree of difficulty of the build was going to be very high, you had built a version of this for yourself and maybe to a certain degree, you thought that your preferences were representative of the larger market and it felt like as you spent more and more time with trading partners as you were building and delivering product and building and delivering product, before you had any real launch, anything like that, that in some ways your preferences didn't externalize or weren't fully repeatable to an end market, that there were too many people that wanted to do all sorts of things in all sorts of different ways that made the degree of difficulty maybe even higher. I don't know if you have sort of thoughts or reflections on that.

Zack: I love the term product market fit. I think that the coining of product market fit is a technology in and of itself. When I was in the physical product world, product market fit is pretty obvious. Now, product market fit in physical products is not easy, also is also hard to find product market fit. Plenty of people don't find product market fit there, but there are some things that are much easier about it. And when you look at companies that are 150 years old, you look it up, Design House that makes clothing is 200 years old and every season they're shipping new designs and new stuff. You see people who are making TVs and they're making new TVs and you see people who are making all these different items. Furniture makers, it doesn't matter if the company's two years old or 50 years old, they're shipping new products. And then you come into the software world and the most common thing in the software world is that people talk about things. They know what the features are that people want and you can't ship it, you just can't. But why is that? That's kind of the trillion-dollar question, but companies get stuck in the mud, and so what you generally see is that companies are able to ship for some short period of time, two years, three years, and then the system ossifies and then they go and they're stuck with that system going forward, and that's basically what they're going to sell. The question is have you built enough in that cement setting phase to build a multi-billion dollar business or do you get stuck. Counter examples to that, of course, rippling would be one great counter example to that as this compound product that's shipping lots and lots of features over a long period of time. Anyways, when you get into software, the reason why product market fit is so complicated is because you're no longer dealing with a 3D substance, you are dealing with an infinitely dimensioned substance. And it's not just like the way the product looks, it's the way the product responds interactively. It's the way that the email notifications work and a text message notification, the notification settings, and it has to do with who the buyer is within the organization. And it has to do with all these kind of intangibles that are very, very hard to reason about. And basic reason for all this is that everything's happening below the line of representation. Mostly vast majority of what's happening in the software's below the line of representation, when you look at Uber and you're working on doing something on the Uber app, the vast majority of what's happening is not visible to you behind the scenes. So solving this problem of how do we make something that users want is really hard. And that to me, the easiest way of solving that problem is to scope it down, which is to say, if this is an infinitely dimensioned substance, at least let's make it small. At least let's make these building blocks individual small pieces. And that's why the idea of a minimum viable product is so powerful. It's why the idea of let's get something in front of users so that we can start testing the product infinitely dimensioned product and also infinitely dimensioned market where there aren't homogenous buyers and everything. You're trying to get the fit of these things together.

Brett: How did it instantiate itself in this path that Steady took?

Zack: So what we did was we first tried building a user interface and probably we got six months into that of saying We're going to build a user interface just for retail and logistics, and then we'll kind of expand out from there. What we saw was like, okay, you look at a purchase order from Walmart and then you look at one from Amazon and you try and combine the properties that are common and you build what's called a canonical model. You do that a number of times and we start to make these decisions, and I'm pretty comfortable making the decisions on a purchase order and a purchase order acknowledgement on the ship notice I was less confident about that because it's a more complicated thing, and then you start to have to deal with people like, "Well, I really need to do my warehouse shipping orders." You're like, "Oh, I've worked with those sometimes before." And you start thinking about, okay, there's 321 of these and I'm looking at some of these and they have to do with student loans or student transcripts and I have no clue about any of that. How are we going to be able to scale the quality of decision making that's required in order to not make a whole bunch of bad decisions here? And basically came to the conclusion pretty quickly, this is not possible. And this is one of the paths where people die. Josh Wolf talked about this idea of at some point he said something like, there's a myth that the best entrepreneurs are risk-takers when in reality the best entrepreneurs are risk killers and they're trimming away failure paths. And the thing is, when you trim away a failure path through odds of success go up. And so we found that within six months we were like, okay, trim that off. That one's a failure path. What we ultimately landed on was, okay, we need to strip this all the way back. We need to build support for all 321 transactions. That's all 30 plus releases, and we're going to do this by shipping Lego blocks. We're going to ship an API, a developer focused API that turns EDI this format into JSON a developer friendly format and vice versa, we're going to launch an API for mapping one JSON format to another and we're going to launch another one for spinning up SFTP. And eight of these APIs relaunched throughout 2022 could be stitched together in order to build anything. It's kind of like discovering the building blocks for DNA, you figure out what the building blocks are, you can create a great many different organisms or instances of organisms using that combinatorial structure. So we did that and what we found was that indeed, to your point, there were customers who could do that, and this is not really unexpected, but there weren't tens of thousands of them. There were small development teams who were super competent led by a CTO, you know, who was hands-on coding of four or five engineers. They could build it, but it would cost them tens of thousands of dollars of development time and months of work. And then sometimes they would pay us $4 and 18 cents in a month. So obviously it's a bad trade for them because they want to spend more dollars and less development time. Bad trade for us, because we want to earn revenue. And so we started working on the next layer on top of that, which was A, looked at all the common things that customers had to build over and over again, and we built the SaaS platform that actually wasn't too dissimilar looking from the original thing that we had started to build and through away and that SaaS platform under the hood use these eight individual APIs to stitch things together in a higher order structure. And it gave the ability to configure things in a user interface, still API driven, but you could configure things in user interface, you could retry things, it kept track of your all transactions and all those pieces, but it still took weeks or months of implementation time for people to get up and running. Maybe this time it was less about code, it was more about configuration, but we had customers who were using this for retail logistics. We also had customers who were using this for healthcare. And what we found in the back half of 2023, we launched this in March of 2023, was that nine of the top 10 biggest deals that we signed were in healthcare. And we started looking more deeply into healthcare space and found that there's a lot of things to like healthcare is 18% of GDP is obviously a very big market, but also the HIPAA Act of '96, which everybody knows is a privacy framework. You might have had to sign HIPAA paperwork at your doctor's office or gone through a HIPAA audit. Also mandated transaction standards. So we talked about this problem of Amazon and Walmart using purchase orders that were incompatible with each other, using kind of different versions of the standard. That concept doesn't exist in healthcare. So everybody is mandated by regulations surrounding the HIPAA Act to use the exact same schema. So yes, the schema is 650 pages long for a claim and you have to translate that from a PDF into something, but it outlines a superset of what is allowed, which cuts off a big part of the problem. And the second part of the problem is the connectivity piece. So no matter how big you get as an EDI company, you have to integrate every single one of your vendors one by one with Walmart or Tractor Supply or Target. There isn't this such thing as you connect once and the transactions just start flowing. In healthcare, the existence of clearing houses means that they're pre-connected to thousands of different insurance companies. And so you connect once and boom, the transactions are flowing instantly. Those two things combined were the big breakthrough.

Brett: And so if you started at the end, the product would have the same limitations as so many of the other products in the category.

Zack: Yeah, although I think a little better, you have the benefits of cloud computing and things like that and more modern software languages and libraries and stuff, but by and large I think it would've ended up in the same spot.

Brett: So how do you define product market fit? What is the Zach Kanter definition?

Zack: We thought we had product market fit in 2023. Things were going really well. We were growing and we were landing, customers we're getting great feedback and all this stuff. And then we launched the backstory of this is Change Healthcare, which was the largest clearing house for claims. They got hit by cyber attack and they were down for 60 days and even today they're not fully back up. There's the dental portion of their business is not back up. There's a bunch of things that are not fully up and it's the equivalent of it's like, it's like AWS went down for 60 days and everything comes to a stop and it was just a disaster for everybody. And of course in the days after we launched that, what we built was a set of drop-in replacements for their APIs so that people could migrate directly from them to us. And it was like the classic things you see, read about in articles or you see in movies of like, we launched this thing and immediately it spreads like wildfire. It was like you don't have the chance to go viral in the B2B world very often like people go viral in the consumer world. In this case it was viral, it was like through our investors' portfolios we were onboarding all the customers and then through our people who weren't investors in us were calling us and saying, "Hey, I heard you helped out this company. Can you help out our other companies?" Three weeks into this, I had not left my keyboard for a long. It's this thing, you wake up at 4 35 in the morning, you work till midnight, you get some sleep, you come right back to it. And my wife said, "You have to go outside, we have to go on a walk, you have to go outside." And I went on this walk for 15 minutes and in that time period I took three calls from CEOs and CTS who were desperate to get back up and running, people who were behind hundreds of millions of dollars on claim submissions, going to run out of money if you don't get them back online. And so more extreme version of product market fit that didn't continue forever because last for maybe three months and then Change Healthcare came back online from that point for the next few months after that, a snake swallowing a deer where like, yes, a snake can digest a deer if a snake is big enough, but this is not going to be a particularly fun digestion process. We had to digest all these customers that we had landed and then we had to build a repeatable go-to-market motion that was not dependent on your competitors getting hit by massive cyber attacks.

Brett: If that exogenous sort of change brought into tight focus, what extreme product market fit actually looks like, the Delta between that and what you had before? Is it urgency in the buying process? Is it the amount of people lining up for the thing that you've built? Is it the repeatability across all of them? Is it a little bit of everything? How would you compare and contrast?

Zack: A combination of things like speed to close.

Brett: Urgency?

Zack: Yeah, urgency is when people are buying the product despite all the things that it doesn't do and the deals are closing anyways, the deals are closing anyways and everybody's so happy. And I originally started Steady because I said it would've said at the beginning, it was like nobody's writing love letters to their EDI platform. And today I say the same thing, nobody's writing love letters to their clearing house. Everybody really has a hard time with the other clearing houses. And the most common thing that we hear is that by the time we have a deal signed with somebody, the other clearing houses often haven't returned their email, like their initial email. And so that's what product market fit feels like is I would hesitate to compare it to that time after the Change Healthcare hack because it's so extreme, but what it feels like now is just you have so much to do, there's just so much obvious stuff to do. The features are getting pulled out. We could sit down and come up with a list of 2000 features that our customers have asked us for and all of them are valuable and you just kind of have to ship them as quickly as you can. But customers are the product anyways.

Brett: One of the things that you've touched on sort of in a reverse way is that stylistically there's two dominant ways that you can go about building a software company. One is that you down scope product surface area and down scope the use case and down scope, the end customer ship something as quickly as possible and then layer cake the company for the next decade. The other is one where to get to sort of the minimum usable or sellable product, it might take a year or two or four or five. And there's different reasons why. Sometimes the long build shaped products, there's a real set of technologies that you have to assemble. That's probably like the Figma case. It was clear what the thing had to do, but that you needed to assemble a lot of technology to get something like that to work in the browser. Another is just that the core product surface area is so large. Like, Workday couldn't be built in two weeks. You ship it on a weekend type thing and there's obviously consequences of starting in either position. We can obviously come up with examples of either position that led to meaningful companies and come up with either position of the company that never ultimately shipped after five years or the wedge product that was basically a bridge to nowhere. But I'm curious because you've spent a lot of time thinking about this. Stylistically you've been a slow build style company.

Zack: I think you kind of need a fit between what sort of problem you want to solve as a founder that people talk about founder market fit, but I think kind of also founder approach fit. And you get into business and you start reading about Warren Buffett and Charlie Munger and they talk about moats and they say like, "Oh, a moat is a business that's so great that a ham sandwich can run it, because eventually a ham sandwich will win it." And you kind of look at the businesses they cite as having these, Coca-Cola and these different businesses that have these moats and your conclusion is, by these guys' definition, there's basically no great businesses. They own half of them and there's basically no great businesses. And you see all these companies getting built around you at the time. I remember Slack being built at the time and you're like, "Well, Slack doesn't really have any of these characteristics that these guys are talking about. I don't think ham sandwich could have run Slack at the time." And so you kind of draw the conclusion that moats must not be that important. Very common thing people say in software in venture, "Oh, the only moat is execution," which is just definitely not a moat. And then you get into, so you go through and you think like, "Okay, execution's the Moat and I'm going to kind go through just move as quickly as possible." And then you see these companies that initially seem promising either just stall out in terms of growth or get obliterated that don't have moats, and now it kind of put Slack into that category, although they had a really wonderful exit story. But you look at Slack, and I remember when Slack came out, I knew a guy who was an investor in Microsoft, which at the time this guy didn't know anything about tech and he was a huge investor in Microsoft. And I was like, "Man, you have no idea what you're talking about." He is like, "Teams is going to kill Slack." And I'm like, there's no self-respecting tech company on earth that is going to use Teams instead of Slack. What you don't appreciate is that it's not about steady or first round or Uber or whoever it is using Slack or not using Slack. It's about the marginal Fortune 500 company. Whether they're going to decide to use Slack or whether they're going to get Teams for free or bundled in. And if they make those decisions to get it for a lower cost, it puts downward pricing pressure on Slack. And so the price that everybody pays Slack goes down. The reality is that most people don't care about the product only being 80% as good. For most people, that doesn't matter. They're not willing to pay a lot and more for that incremental piece. So as you start to see these businesses not really achieve this sort of dream that they were set out to achieve, and it seemed like Slack was on this meteoric rise and is going to be the fabric of everything that we do things in, and started looking for explanations as to why this is not happening. And that led me to, you read about Michael Porter, Porter's Five Forces, which is something you would read about in business school and it's fascinating. He basically boils it down five forces and all this stuff. But what it compiles down to is very useful, which is that if something doesn't lead to lower costs, sustained lower costs over time, or the sustained ability to charge higher prices over time and ideally both, it is not a mode. And so when you look at execution's a moat. Okay, is execution going to lead to sustained higher price or lower costs over time? No, it's not. And of course everybody else is trying to execute very quickly too.

Brett: The biggest downside of taking on one of these big builds is that the company ends up never shipping or they end up shipping the wrong thing. Right? Part of the point you're making is if you have something that is a very rapid, basically timed to first product shipped, you can get in front of customers. When you get something in front of customers that gives you data, you take that data and you iterate and iterate and iterate. If you look at the long build, you have sort of these two dominant problems of building the wrong thing or just never shipping the thing. Do you have any thoughts on how do you avoid those things other than a founder and CEO with very good judgment or who has well modeled the problem in their head?

Zack: It's tough because I think that for every actual instance of this is a long build that you have to build all this stuff, there are probably hundreds or thousands of people who think that they have a long build that could be done incrementally.

Brett: And that if you have a long build that can be done incrementally, you should go do that.

Zack: Absolutely, yes. If there is the option to ship things incrementally, it is, I think virtually always better and that's how we ship things now. We we don't ship big bang features. In fact, we de-scope features to the point of where we think it will be barely usable by one customer and we get it there because what we're optimizing for something a little bit different, which is that we will never have one individual product that we can go with one couple of features that we're going to be able to build a billion dollar fee stream on. We have to build many, many, many products across a broad surface area. So what we're optimizing for feature velocity over time, and in order to do that, I think incrementalism is the only way to go. It's very hard to ship big bang, big release features in succession like that without incrementally releasing them to customers.

Brett: What is it about breaking things down and getting it in front of customers and listing to feedback? Why is that so valuable in building software?

Zack: All software is a cascade of miracles. When you look at something, I think you go in Vibecode for a weekend and you're like, "Oh my God, why doesn't my company ship more?" Because I just shipped. Look at all the stuff I shipped into the weekend. And then you keep running with that project. You realize like, "Oh God, I didn't think about how am I going to do the database migrations and how am I going to do this? And oh, and I shipped this regression, I don't have any testing in place and oh no, I'm not using the SDK, I just use this portion of the API spec." And you gain this new appreciation for like, "Oh, it's actually really, really hard to build and maintain a complex code base." And so a complex code base, a complex piece of software is basically any piece of software that you use today is pretty complex, is a set of miracles and any feature then has a set of miracles. And so when I'm reviewing our tickets or our design docs or whatever, I'm going through and I'm just looking for miracles and I just want to cross, I want to get us down to one or two miracles and I use a broad definition of a miracle. I think as a miracle when a feature ships and works for a customer and solve some basic problem and doesn't back people into a corner and all the studs have all these downsides. And that's why I think overall this idea of how can we scope things down, how do we get to ship things by Demo Day? How do we get things in front of customers is either basically just eliminating miracles. But going back to this product market fit analogy of you have a many, many dimension product, you have many, many dimension customer configuration in market. Basically what you're saying when you're saying I'm going to go build something small and talk to a customer is you're reducing the number of fit surfaces. So you're coming down, maybe you have 12 fit surfaces here and 14 fit surfaces here versus 300 fit surfaces here and 3000 fit surfaces here.

Brett: What are the most important ideas that you've taken from other places or other industries or other books that are a lot of the input drivers to the way that you have built the company?

Zack: I think that there's more to learn from the category of discount retailing than there is in software than there is to almost any other category. And you look at companies like Aldi and Trader Joe's or kind of some of the newer ones, but then you have the Sol Price business and you have Costco and you have Walmart and software. You kind of have it on easy mode because of the margins. And I think we've learned the hard way over the last 15 years that anytime there's a new venture backed business that looks like there's a software like Growth Metrics but not software like margins, things end up poorly. And so they say product market fit is the antidote to stupidity. I think that product market fit is the antidote to stupidity, provided that as product market fit with high gross margins and I think product market fit with low gross margins is a tough business, but if you can learn the lessons from people who've played on hard mode, I think that you're going to have a leg up in the software industry and discount retail, they're working off of not just retail, which is low margins, discount retail, which are working off razor-thin margin. These people have built enormously valuable businesses. Sol Price built such a valuable business and then went and taught the Costco founder and I forget if he taught Bezos, but maybe the Bezos went to the Costco founder, but there's this whole network of all these businesses that were built based off of these ideas and what does it look like? It looks like lowering the cost, so figuring out how do we sell things for cheaper? How do we reduce the selection that we have? How do we simplify the process of buying from our company as opposed to making things more complex? How do you figure out what the right packaging is and what the right units that people want to buy in? How do you drive your cost down? I think all of these are themes that come from the world that was like this was honed in the trenches of discount retailing that are applicable.

Brett: So take a couple of those and explain how they express itself specifically in how you're building steady.

Zack: When we were in logistics and retail, people would riot over paying for these things called transportation status messages. And this was like when you see an 18 wheeler going down the highway every 15 minutes it has to emit a status message telling you where the truck is and all these things and that happens whether the truck is moving or not. And people didn't want to pay a hundredth or a thousandth of a cent for those things and that's because they were making a really small amount of money per load. They were not making very much money. The gross margins are tight and that amps up all the pressure on the business and you come into healthcare and people say, "Well, the transactions are commoditized that clearinghouse is the process." And you look at eligibility checks or claims are commoditized, but when you compare the price of a claim or an eligibility check to a text message or to a transportation status message or to an S3 read or write or storage or something like that, orders of magnitude more expensive. I love the Jeff Bezos quote of there's at some point someone wanted to raise the price of something in Amazon and he said, "Look, there's two types of companies. There's company's always working to charge more and there's companies who are trying to charge less, and we will be the second. We'll be the company's working to try to charge less." And we're doing that. We're pulling costs out of the transaction.

Brett: That's the your-margin's-my-opportunity.

Zack: Your margin's my opportunity. And I think that it's not a matter of saying, "Hey, we can operate on a lower margin than you can, but maybe even we can operate on a higher margin than the competitors because they're based off legacy tech stacks and legacy support models and things like that. And we can do that at a lower transaction cost. And so I think it's a basic question is are you going to be a business that's going to pull cost out or are you going to charge more? And I think those things can be both. I think that we hope over time to be a company that charges less for transactions that should be commoditized and charges more for the optional value add functionality that people can have on top of that.

Brett: Yeah, AWS is a good story and that costs every quarter basically forever have been driven down and gross margins are very strong in that product.

Zack: That's right. And people don't churn. Saying people don't churn from AWS because of costs. They look at this and say, "Okay, my costs are probably not going to go up here. They're probably going to trend down over time."

Brett: What about the, you mentioned this in passing, but the Toyota production system, it seems irrelevant to building a software business, but I think you'd disagree.

Zack: Fundamental tenets of the Toyota production system, everybody works off Kanban boards today. It's like Kanban board is an invention of the Toyota production system that is the Kanban board is based off of the Kanban bins where there's a bin moves throughout the production line as parts are needed and it goes and fetches the previous part. I think that the biggest lesson, well there's two big lessons from the Toyota production system that I took is a lot of lessons, but there's two wins that stick out. One is that people often think about Toyota reduction system as something that was done to reduce costs and when you read these guys talk, they hate talk about cost reduction. Now I've just been talking about cost reduction for the last 15 minutes, but they hated talk about cost reduction. They said the things that you do to reduce costs, the cost savings almost never materialize. And you read about this like, "Oh, we're going to do this to save this money and blah blah," and then everybody's margins are all kind of always the same. It's never really materializes. There's another variation on that that they say when you buy a machine in order to save cost but the volumes don't materialize, it wipes out the potential savings many times over. And so an example of that is you're using an on-demand service from AWS for queuing and you're using this menu. You're like, well, we did a million messages for the last three months. We're going to set up a server instead and we're going to in that server for the same cost can process 5 million messages or whatever, and then you set that server up and you are running RabbitMQ or whatever you can in order to save money and then the traffic on that goes down from a million to a hundred thousand because it turned out that someone figured out how to batch them, but now nobody's looking at that server and so you're running the server and the costs never go down. And so the cost savings often don't, first of all, when you buy a machine to do something, when there's this whole trend about people going on-prem off of AWS over the last couple of years, which was like, okay, you buy all these GPUs and you don't need them, you buy all these servers, you don't need them. It becomes fantastically expensive versus the cost savings that you would've had. So what they say instead is don't focus on reducing cost. Always focus on improving quality because the quality gains usually materialize if you say like, "Okay, we're going to reduce the defect rate or it's a very measurable thing," but what you find is that by improving the quality, the costs usually go down and I think that everybody's had somebody say to them at some point say, "Well, I want this done, I want this done, I want this done." They say, okay, well quality price and speed, you have to pick two. And I think that that is the most losing mindset in business and engineering in design and whatever the thing is. And that the trade-offs between those three are usually an illusion that usually the thing that can be achieved at the lowest cost and the fastest and the highest degree of quality. I think that the best people will find the way of delivering all of those things together and it becomes a virtuous cycle. If you go and you cut corners of quality in order to ship your code faster at a lower cost, then your code is something like you have to live with that forever. You have to go back and fix that later. And so I think that generally you have to find crafts people. You have to set reasonable scope is generally the thing to adjust and say we're going to scope things down in order to move it faster and you can kind of get the best of all the worlds as a company, we've decided we're going to eat glass. That is what sets us apart. I won't call it a moat because the eating glass is not a moat, but we'll go to the ends of the earth to do things the right way even when it's not economical and it doesn't make sense. It's kind a matter of principle, a matter of taste if for no other reason than the accumulation of not doing things the right way is the sort of thing that would make me hate the company. And if you want to build a company that you want to stay at for a long time, just don't do the things that you want to hate.

Brett: Where does your source of agency come from and what are your reflections on what it means to be a high agency person? Because I think you seek it out as a top level criteria when you're adding people to your company.

Zack: On everybody's first day Steady. I say, "Welcome to steady, everything's your fault now." And then I say, "It's not as bad as it sounds. Actually everything's your fault now also. Everything's your fault and your manager's fault and everything kind of compiles down to being my fault." But to me it's a great thing about having a company. I think being an adult is the greatest thing in the world and to be in total control basically of your life. I guess maybe a lot of people feel like I have to be responsible for things and I largely feel that I get to be responsible for things. There's a book that we've both read called The Courage to Be Disliked, which I think is a terrible title because that doesn't really, it's not really the message of the book, message of the book is that people are getting exactly what they want all the time. If you're not getting the things that you think you want, you probably don't really understand what it is that you want. And so you talk to founders, you're like, "Oh, it's the investor. They're being so difficult," or whatever, or, "My employees, they want this policy or they don't want this policy." It's like, you must want to be miserable in your business. By revealed preferences, that is what you want because you could just broadly just fix these things. And does that mean you're going to have a bad relationship with some investors? Yes. Does it mean you're going to have a bad relationship with some employees or have to part ways? Absolutely. It's just hard to imagine, I think, being a different way as opposed to, I guess just going through life and thinking that things are happening to you.

Brett: Is it something you actively seek out when you're adding people to study?

Zack: It would be very noticed in screening. We just don't have that. Maybe because of the type of people that apply, I don't know, but we don't have that problem. But there's all these things you learn about interviewing and I think a lot of the common wisdom about interviewing is probably a lot of it is correct in terms of it's a bad sign of someone's complaining about their previous employer. What we think about during interviews is what I look for during interviews. I want to get a sense that the person is held back by the environment that they're in, that this person could be achieving more, if not for the team that they're on or the organization that they're in. The idea of this is like a caged animal who needs to be unleashed is something that we look for. I think there's a fine line between that and somebody saying like, "I'm held back by my team and my company sucks," and-

Brett: "If only this and this and this, then I could have done this, this, and this."

Zack: You just find the generally high agency people just find ways to make things work and yeah, it certainly something that we look for. Somebody's like highly resourceful.

Brett: In the years that you've been building Steady, have you changed your mind about specific things you're looking for in people, the traits that tend to tie the people together that you want to hire, not a specialist in this area or sort of that type of thing?

Zack: Some of these things are trite because everybody says the same thing in the beginning. You look for people who have a lot of experience and then you realize you hire someone with 20 years of experience and they don't have 20 years of experience, they have one year of experience repeated 20 times. By contrast, you hire somebody who's two years out of college who doesn't know anything about anything and you can have them performing well in a job. The thing that we've ramped up over time in terms of what we're looking for is desire to work by and large. We've gotten a lot more comfortable screening for that. I think during the process of people who are work enthusiasts, people, what does that mean? You're working a lot of hours. Does that mean that we're holding Sunday morning engineering meetings? Definitely not. Definitely not. But I think that we went through this phase of the zero interest rate of 2020, 2021, whereas you couldn't bring those things up otherwise people wouldn't work at the company. I think people have realized a couple things which is, "Hey, it's nice to be if you enjoy those things, if you want to be part of a winning team, it's nice to be part of a company where people also are enthusiastic about putting in hard work and putting in more hours." I think this stuff can get performative with all the 996 stuff on Twitter, which I think is more about broadcasting than actually doing work, but that's been something that we've looked more for and I think that the desire to work also kind of tends to be correlated with other things and I think that that has become one of the traits that we look for the most.

Brett: In the context of an interview process, what are you looking for to give you confidence that somebody's a work enthusiast?

Zack: I find that if you lay it out, people are pretty happy to select into one of the categories that you lay out. Generally, people are not going to tell you that hey, they're in that category if they're not. And I very much believe this about hiring I think is true about friendships, relationships, investors and everything. People are very loudly and obviously going to broadcast to you who they are and what they believe in and what they want and what they want to do and how they're going to act. And it's up to you to listen to it. And the reasons why someone doesn't end up working out as an investor, as an employee, as a whatever are almost always the reasons that you thought they weren't going to work out. And this is one of our, we have four or so questions that we answer internally after an interview panel and one of them is on what dimension do you think this person might be a superstar? And another one is if the person doesn't work out, what is the reason that you think why? And you find that just almost every time when somebody doesn't work out, it's the reason why people put down in the thing.

Brett: The way the world actually plays out is not you ask somebody, are you a work enthusiast or where would you put yourself? And they're directly lying to your face to get the job. Actually in general there's very little lying. It's that you're lying to yourself basically or unwilling to sort of listen to the actual data that's sitting in front of you.

Zack: Absolutely, and what does that look like is you meet someone and they're the perfect person for the role and everything's great and they've got the right knowledge and they've got this and they're at the prime point of their career and everything and they're enthusiastic and you're like, "How do you think about work-life balance?" They say, "It's the most important thing to me, but I believe in work smarter, not harder. And of course when crunch time comes, I'm willing to put in the time and if something's needed after hours, I'm willing to do it." And you're like, "oh, that's a pretty good answer and this person will fit right in." And then you realize later you're like, oh man, they told me right up front that the number one thing that they were optimizing for was work-life balance. And that's fine.

Brett: That's the beauty of a free market system that we have here. Over the years, there's lots of things that you've chosen to do your own way at Steady that maybe is not perfectly aligned with best practice, be it the way you think about having managers or zillions of other topics. I think you build a very merit oriented company for a very long period of time, that sort of genre of topic. Do you have a way that you go about making these decisions? In what way are we going to have a conventional sales team, conventional pricing and packaging versus we're going to reinvent it and do it the steady way? Is it just a very organic, "This just makes sense, we're going to do it this way. This doesn't make sense." When you zoom out and you think about all the choices that you've made about who works at the company and how you work, how you organize, how you build products, pricing impact on and on and on, how you support customers, is all of that just your Intuitive taste and judgment and this makes sense to you where if you reason about it and maybe a more mechanical way?

Zack: We landed one big customer pretty early on, coincidentally in the auto parts space, and we basically took over their EDI department and we used through that four-and-a-half year period of where we were figuring out what to build and how to build it, we were not building without a customer, we were building with one customer just not building for the public market. So we took over this customers EDI department and figured out how do we build something for them? And we went to really extreme degrees there and invested many, many, many millions of dollars in building that. In terms of estimating the complexity of did I think it was going to take that long? It'd be that hard. I'd spent $50,000 of my own money with the auto-parts business building that EDI system, that homegrown system, and I said, in order to make this generic and work for everything, it's going to cost at least 10 times as much. I said I was going to raise a million dollars just to be safe to have a two x margin of safety and now we've raised $92 million and still lots of stuff left to build. So I was off by a couple orders of magnitude with this Automotive customer we've taken over the EDI business. We just did an unreasonable amount of work in order to make things work and everybody told us at that point, "Well this level of support is not really going to scale," and so it's not going to work for 10 customers. And then over the years we got 10 customers and we kept offering this level support, hyper-responsive and everything and then we got to a hundred customers. People say, "It's not going to work when you have more customers. You can't do this with hundreds and hundreds of customers." And then we have hundreds and hundreds of customers now and we do it and we don't have a different support program, any internal designation for support between customers who pay us seven figures per year and customers who pay us $500 a month. This is the exact same support. It's like an example of something that people said just wouldn't work, and that's something I learned with the auto-parts business. You know what I did is I said, "When people have a warranty claim, I'm just going to replace the part for free," and I'm just going to send them, basically no questions asked and people said it's not going to work in the beginning. "People are going to learn about it and take advantage of it." I was like, "Okay, well I'll figure that out as it goes." And just you find that most of the things that people warn you about don't come to pass and most of these edge cases don't come to pass. I think it's a little bit different in consumer with fraud and stuff like that, but B2B, generally don't need to worry about those things. Overall it's just been picking the philosophy, the things that just make sense where I'm like, I think that we wrote a blog post about this, David and James on our team wrote a blog post called Do Support That Doesn't Scale, a play on Paul Graham's Do Things That Don't Scale blog post and talked about this kind of philosophies of support. I think a lot of things have fell into that category of it's just reasoning through from first principles and doing things in a way that we thought made sense and being open to if it doesn't work, we can always go back and do it the other way. It's like, it's very easy to regress back to the mean of what everybody else is doing.

Brett: What are some of the other things that you've done your own way at Steady, or maybe the most important things that you've chosen to do your own way that if you just followed best practice it would be severely suboptimal for the company?

Zack: We've biased to really heavily towards hiring individual contributors. I think that there's pros and cons to that. I used to think what are managers for and is that really necessary? And you could build borderline the holacracy model and that would work and I've come much more around to the idea of you just need a system that's consistent and you asked about sales hiring and when it was time to go build a sales team, which I pushed off for a very long time. I wanted the product to be working and selling itself before we hired a sales team and I went out and I surveyed eight or nine of the smartest people I knew about enterprise sales. One person says, "You should hire someone who's in your industry who understands your domain, is the best on lock." You talk to the next person and they say "You should hire based off of last year's W2 and that's basically what matters." Maybe he didn't say that. Somebody else might have said that. You come through and you come away from this. My conclusion of this was at the end of this is when things are working it kind of, not that it doesn't matter what you do, but there's lots of different ways to be successful. The important thing is that the decisions that you're making are consistent and that they're well implemented. So some of this stuff is like, I'm not sure how much it matters as much as it matters, picking a system that works together and I think that our system all kind of works together really well.

Brett: How do you get other people to have high quality judgment in the way that you do. Maybe where so many of these decisions are situational and not just this is a correct way, this is an incorrect way. There's a correct way and incorrect way to do something in the context of Steady, but I would assume at the 96th employee, it's so much harder to get that unconscious alignment that maybe you have with five or seven people when you're getting the company going. Where ultimately as you get to 100, 500, 1,000, the whole point of culture is that we have a consistent way that we basically problem solve and accomplish work together. Have you done anything to try to explicitly get the 96th person to have the same way of reasoning through problems as you do?

Zack: We have hired managers, now engineering managers and I talked about this idea of we were hyper-focused on hiring individual contributors and for a long time that people were, our team leads, which functioned as engineering managers spent 90% of their time coding and we got to a place of, the ultimate reason why we needed to hire engineering managers is because we had to double the size of our engineering team who's just in the space that we're in. You talked about first mover advantage like OpenAI has. We have what Peter Thiel talked about as last mover advantage, very clear what the features are, very clear what needs to be built and we just had to double the number of engineers or more and you can't just double the number of engineers and not add additional nodes. It's like you're going to get to a place where there's too much product context to which engineering context, the on-call rotations get complicated. So we hired engineering managers and I got into a discussion with one of the engineering managers and I asked, we do a Demo Day every Thursday and we were talking about something about planning meeting and I said, "Well, why do you think that we do demos? And why do you think we do Demo Day?" He said, "Well, I think we do Demo Day for accountability." And it was funny, that wouldn't have been in my top 10 reasons. I don't do Demo Day to see is this person working? You have other ways of telling if somebody's working, we do Demo Day because it's a way that everybody can get a shared understanding of how the product works and what's changing. And also because it encourages people to ship in incremental units that are viewable on a week by week basis. It's not the purpose of seeing if work is getting done. And so it's a hard problem of how do people know what's important? I need to do more writing in order to clarify somebody saying, I think as you identified, having a close cohort of people that you trust with really high quality judgment and we have people who have been there four or five years. We have a mechanism that we use extensively, which is called the decision record. And a decision record is a very much written culture, we write things, write important things down, and decision record follows a template and it says scope, context, so scope of what team it applies to context of the background of it, decision, which is the actual meat of the net of what's being decided action items and notable comments which was harvesting the comments that happened on the side of the Google Doc and anything that is done of consequence at studies basically driven by a decision record. And so we make decision records for SLA's, make decision records for SLAs who make decision records for hiring certain roles who make all sorts of different decision records. And so in terms of scaling judgment, I think part of it is scaling judgment in terms of more people with judgment, but then also scaling your own judgment. The other thing I do is I look at take at least a cursory look at all the pull requests that are made throughout the organization every day.

Brett: You've shared a lot of wonderful things with me over many years to read and one of my favorites was this short essay. I think the title is Reality is a Surprising Amount of Detail. I think it's something that you've thought a lot about as you've been building study and sort of earlier parts of your career. What's sort of the big idea in that and something that'll get somebody to go want to read it.

Zack: Reality Has a Surprising Amount of Details is an essay by a guy named, if I remember, John Salvatier and it's really wonderful. The basic premise of it is this. You've walked up and down staircases many times, then you could probably maybe draw what a staircase looks like, but if somebody gave you the tools and sent you to Home Depot and said, go build a staircase, you're going to find that it is a lot harder than you think. And it's not just building code. There's angles so that it's comfortable and how big does it have to be for your feet and how do you make it stable and all these different pieces. And another classic example is you ask people to draw a bicycle. Most people have ridden a bicycle. Most people with the bicycle pictures that people draw are hilarious. I mean this is contraptions that would never work in a million years. It's this idea that basically almost everything is a lot more complicated than you think. When you say something like, to go back to the example of the voicemail. If you were to write out a standard operating procedure for updating the voicemail, it's going to be like 25 steps long. And now yes, can you do that? Write the SOP for it? Yes. But now you want to keep it up to date and you're using a voiceover IP phone system or using Zoom or whatever. Now you have to keep that up to date and then you're putting it in Slab and then you decide, "Oh, we're going to switch to Notion." And then the links break and everything is just, there's fractal complexity throughout everything that you do. And it kind of comes back to this concept of the Cascade of Miracles problem, is these ideas, which Boyd talks about this patchwork of strategies, ideas are all closely related. You have this idea that you don't want to do too many things because they're cascade of miracles because multiplicative probability that lessens your chance of success. And similarly, reality has a surprising amount of detail. You have to be very careful what are the things that we're signing up to do because every single one of these things has a lot of detail in it. And I think that's why this incremental approach of what can I get done for Demo Day is like, it causes you scope and scope and scope and scope until you get to something that's manageable.

Brett: Is there anything else that you think you figured out in the process of building this company that is generally useful to other people at the early stages of building a company? You talked about a bunch of these ideas in various ways. Part of the point that you make that I think is certainly correct is that so many of the right things to do are right in the context of the machine you're building. And so the most important thing is to actually have a point of view on the machine that you're building, not just generically correct things, but are there other things that are broadly useful that you figured out in the process of building the company thus far?

Zack: Look, I think there's a difference between dogmatic and principled. I think that when you think about basic way, I describe principles versus preferences, is a principle is something that you're doing no matter what, that you are dogmatic about it. And the preference is something that you want to do, provided that the circumstances are, if the trade-offs are reasonable. Think about you find, you think that you are principled, the principled person for honesty and integrity and then you step into the back of an Uber and there's a suitcase of cash with $50 million in it. You're about to find out very quickly, is this a principle that you have or is this a preference that you have? And what most people define, think that are their principles are actually their preferences. Now. For example, I have a principle for doing things the right way, like these dependency updates and security and things like that, that we really care about doing things the right way. And when I say no matter what, I mean no matter what. There's other things that are preferences had a preference for not building sales function in order to sell something that wasn't working yet. And once that the trade-off became clear enough, I built a sales team. The sales team has been amazing, but we also did it at the right time. I think it's the right thing at the right time.

Brett: I want to wrap up with the topic that we always do, which is who's the person that have the biggest impact on you, the way that you think or in this case build the company? And what is the specific thing that was in part a person you could know or personally or know through a book that you read.

Zack: I think it's probably Jeff Bezos who I don't know. But he, maybe five years ago I wrote an essay called What is Amazon and a 25-page essay that explained my understanding of Amazon, having sold to Amazon as a retail seller and as a marketplace seller, and then used AWS and various different pieces, advertised on Amazon. And what I think I either talked about it in that piece or in a podcast or something about it was that what Bezos came up with, with the Amazon culture of the six-page memo and the two-pager and doc reads and the Amazon principles and the various pieces that make their culture, the service oriented architecture and everything. And when I looked at Amazon this, I think this was when Bezos was still CEO, I said, "The thing that I admire most about Bezos is that if Bezos were no longer CEO of Amazon, I think the business would suffer. But I don't think that the business..." If Elon were no longer the CEO of SpaceX, I think that would be a huge problem. Amazon is in really tough spaces, tough, fiercely competitive spaces, but they have a business organism-

Brett: Ideology.

Zack: ... ideology that has transcended just him or any one person. And I think that that's an unbelievable achievement. I'm very happy with our way of doing things and I feel like we have a special way of doing things. And I think that our customers recognize that there's a special product being built in a special way that it's being built in a special way, that they're being serviced as customers. And the next challenge among the other things is figuring out how do you distill this down into something that can be understandable and teachable and learnable. And that is when you think about everything's my fault, it's definitely a failure of mine to not have recorded all of those things so far. And it's challenging and that's something that I'm looking forward to, is figuring out how do you institutionalize that knowledge.

Brett: Good place to end. Thanks for joining.

Zack: This was great. Thanks for having me.