Brian Bell (00:01.197)Everyone, welcome back to the Ignite podcast. Today we are delighted to have Jason Cohen on the mic. He has founded four companies over 25 years, including two unicorns, Smart Bear, and WP Engine, and has invested in 60 more startups. He's written about building companies at a smart bear for nearly two decades. And his new book, Hidden Multipliers, shows founders how small, repeatable changes can grow a company with the budget and team it already has. Jason, thanks for coming on. I'd love to start with your origin story. What's your background?
Jason Cohen (00:26.936)Thanks for having me. I've always written code like a lot of people have that same story. I mean I've I'm old, so I've done lots of things. You don't want to hear the whole story, it's not interesting. I guess with WP Engine, which is my latest company, which is now 16 years old. See, I'm old. what happened is I had this popular blog about startups, as you mentioned, and every time I get on hacker news, it crash. So I'd call up other bloggers and say, Hey, how do you keep WordPress running? Because WordPress is the most popular Software for making websites in general and blogs in particular. And people would say, I don't know, but if you find that, tell me, because I need it. Aha. So of course, just that is not a validation of a business, but it is a spark that you might want to fan and see if there's something there. So I did 50 customer interviews, 5-0. And that took months. But I was able to determine that if I did a few things, I could charge $50 a month for hosting. where currently they would spend, let's say, two to five dollars at a shared hosting, Company. And those things were things like making the site fast, making it scale under traffic. So stay fast even if you get a lot of traffic. Security, still an issue today, of course, across the entire internet, especially with AI. And service. actually having human beings that know what they're doing and take the time to answer your calls. And of course, all that does cost money, so it makes sense that it would be expensive. So for people that care about those things, it would be worthwhile to pay for it. And that was the idea of WP Engine. It's still the fundamental. Reason why people sign up today, although we do a ton of other stuff now. Cause again, 16 years later, of course we've done other things. Also, there's 1200 people there, many hundreds of millions in revenue, 200,000 customers. We have we run some of the most popular development tools for the developers that make websites. So there's a lot that now it's a lot, but at the beginning it was just that.
Brian Bell (02:18.061)And so you initially walk me through how a kid from Austin e ends up writing a book on code reviews and r and r you know, what is that?
Jason Cohen (02:26.529)Yeah, so now we're rewinding some more. So I wrote that book in 2005. I started that company, Smart Bear, in 2002. And that's how I have my online name of Smart Bear Smart Bear is because of that company. And I stumbled into peer code review because what I originally did at Smart Bear was made a my idea was a tool my idea was that version control systems are write many, read never, meaning There's all this information about like how we're coding and who did what and all this kind of and when and we don't really use that for anything. There's no reporting. So I had ideas like, hey, so if there's code that's been changed over and over again today, and you make another change, that's just code that's under development. But if you haven't touched this code in five years and you're making a change, and especially if other things are connected to it, isn't that scary? Don't you want to know that? So that was the original idea. It wasn't a very good idea, by the way. It sounds okay. No one cares. So, yeah, some people bought it, but not a lot. It wasn't doing that much. But then what started happening is that people would abuse the product to do code reviews, meaning engineers reviewing each other's work, just like authors and editors do, and for the same reasons to find issues, to help new people get on board with the platform or help junior developers get better. All the usual reasons why you one might review code. And they were abusing my tool. They were like making reports out of my tool and emailing it, and people would write on it and send it back. So I get these feature requests like, can I write on the HTML that you make? And I'm thinking, what are you talking about? Right. Again, this is also 2002. So this is you know, we don't have like fancy websites or software like the it's old school stuff, right? This was in software you installed. This was not SaaS. This is pre-SAS, right? And so they were abusing it to do code review, and that's another interesting insight in general. When customers are abusing your product to do something, one of two things might be happening. One is Those are not your right customers. Those are not your ideal customers. Who knows why they bought your product, but they're they're doing something weird with it. And if that's a different direction than your strategy, then you need to ignore them and they'll get mad and cancel, but that's because it was never right for them. And you have to have the fortitude to realize, wait, it's not right for you. But that means you have to know what your strategy is and who you're selling to so that you would realize this is a distraction. But the alternative is you have found a new vein. And in my case, like the original product was. Okay, but not really that interesting. And all the excitement was coming from many people, not one, who were trying to effectively do code review. So I thought, okay, maybe I should make a code review product using the same code base. So I did. And that was correct. That was something people wanted. So then the company did well. And of course nowadays everyone uses GitHub to do code review. Like again, but we're talking about twenty, twenty five years later. So to be fair, it was a good idea then. It's okay if twenty five years later it's not as relevant.
Brian Bell (05:01.877)What's up? Yeah, it's incredible to think that, you know, just you know, twenty, twenty five years ago, SAS was like this new shiny object. and yeah. Yeah. How fast. I mean, I I wonder what will happen twenty years. It's hard to imagine, but magazine ads. Direct mail. Direct mail's back. so let's talk about the multiplier idea. So the book opens with the story that four changes to pricing and pack packaging.
Jason Cohen (05:25.045)No SaaS, no iPhone, no social media. yeah. I had magazine ads. I mean, yeah, I did. huh. It is back.
Brian Bell (05:46.529)Hook WP engine from one to five million in a single year. Well, walk us through that.
Jason Cohen (05:52.162)Well, there's a whole chapter talking about that, but I'm trying to think. I've written so many articles online that you can get some of the stuff for free. Cause I don't I'm I don't want this to turn into like, there's this great thing, but you have to buy the book. I don't want it to turn into that. so I'll just tell you what the four things are, even though it's a chapter of the book. I'll tell you. I'll tell you, I'll tell you what it was. What happened was about two years in, so because so like I said, people would spend 10 times as much. Okay, so my thought was our pricing strategy is.
Brian Bell (06:05.643)book it book the book you just package up all the stuff that's already it's already out there. Yeah.
Jason Cohen (06:20.759)We're the most expensive and the best. So if your website's very important to you, then that's a good trade. And that's what we did. And to be fair, like we're still probably that now. But about two years in, I got a visit from Erica Douglas, who herself is an entrepreneur and had built a hosting company and sold it. And she was moving to Austin, where I am, and she said, Your prices are too low, and you should hire me to figure out what your pricing should be. And I thought, I don't know if they're too low, but I mean that literally. I don't know. And Erica's she seems like she knows what she's doing. And so we did. And we did things like we interviewed a customers about many things, pricing among them, but just what they're dealing with. And with the result of those interviews, we learned some things, which we then use to change a few in retrospects, it looks simple because many times it takes a lot of work to find a few simple things to do that are right. That's that's what good strategy and good decisions often look like. So just because the answer was simple. That's actually a good thing. That means you distilled all kinds of complexity into something executable. So that's that's a that's a positive thing. and we figured out there were these things. And so, okay, in the in the book, I go through like here's exactly how we decided how to interview customers. Like, how do you know what to ask? And then what do we do with that information? So I go into detail, but the information about how did I interview, that is a blog post. So you can go look that up. It's called iterative development. And look, you can just have that. And then For the items, we found things like this. People would say things like, Listen, I don't mind spending $99 a month for my main site. That's fine. But I have this one site that for my mom, and I have this other one for my brother's wedding, and it gets like two hits. You know, and are you really making me get an account somewhere else? Or can you just let me run those little nothing things in my $99 account? At the time, we and all our competitors, it had our pricing was you buy a site for a certain amount of money. You have two sites, you buy two. seats, you know, and at whatever price. But what I learned is, people who buy a higher plan, they have little pissant sites they want to bring on and can they just have it? And so we created stuff like, okay, the middle plan for a hundred dollars, you could actually have up to 10 sites and it's agr it will have aggregate usage like page views or whatever, it doesn't matter. And that's what the pricing will be limited to. So that eventually effectively we could say yes to that. now again You have to be careful when people ask for stuff. Are they dragging you away from your strategy, your vision, your this or that? the way we knew that those customers were in fact perfect, great customers for us, and therefore we should listen to them is that think about what this is. It's a person who says, I have a medium or large site. So already that's kind of interesting. And I build other sites for other people too. So that sounds like someone who's really invested, really engaged, wanting to like. put a lot of stuff on you, that sounds better and better. Like that sounds like a really solid forever customer. Furthermore, they were often freelancers or agencies where they built actually lots of sites for lots of clients. Okay, wait a minute. So if I, if they love us for whatever reason that they love us, maybe it's our personality or, you know, like what if they love us, not only could we get all their clients, but every client they add. And so this is an account that could grow and grow on this one relationship.
Brian Bell (09:30.092)Yeah.
Jason Cohen (09:39.384)So it's not about this one site for $99 or this mother's website. If we just make this a great relationship, could we be in business together for 10 years while you just put more and more of your clients on us and you just grow? Well, that is a fantastic, awesome customer. So once you put that together, you realize yes, of course, that's what we should do. So that's the decision process that went to yeah. Because of course we had lots of requests for other things that were distracting us. I'll give you example. a lot of times enterprises, and we do sell to large enterprises, At the time, only a little bit. Later, we put together a whole motion for tens of millions of dollars. We invested in all of it. So like we didn't lose our focus. We invested in an adjacent expansion, which is not the same thing. and but at the beginning it was like they would sign up sometimes and we'd say, Okay, that's great. But they would ask for weird stuff like can we have our own custom infrastructure because we need to do this and we need to bora a VPN hole into this other thing because reasons. And we thought about it at first because they could give us a lot of money. Lot of money is good. That's a good reason to do it. But it was so hard to manage it, maintain it, also with things like security. If we're supposed to if we're promising to have good security, but you can literally do anything you want on the infrastructure and and break the VPN, can we really promise security? Like can we in fact deliver on the product? And does it and can it scale even with a lot of money, even if it's a high dollar thing? And so we ended up saying, no, we actually that would be against our ability to scale who we are.
Brian Bell (10:39.573)Okay.
Jason Cohen (11:05.675)what the the things we're promising to do, it would be too hard to fulfill our own promises if we did that. And so we would say no to those things. Now you could argue whether we're right or wrong, but the point is it was a good decision process to say this is why, because then it's not us and it's not our our prom our product promise. We can't fulfill it as opposed to this other thing where, man, this is a growing customer for a long time. And just having some more sites on the same plane doesn't violate any product ideas that we have or any kind of strategy or vision about the product. That's a pricing and packaging thing only. Okay, that is good. So that's how we sifted these kind of things that we heard into like, so, but should we do it? Now, this all presumes you know you have a strategy and you know who your target customer is. Of course, I have other chapters in the book and the things that are just online that you don't have to look in the book to do that. I should also add I'm St. Sam just on its monologue here. there's also AI skills that go with the book. So that's at skills.hidden multipliers.com. Those are also not only free. And open, like you don't even have to put in an email address or anything. They're just there. Okay. And there's a GitHub there too. Like you just have it. And it includes things like the skills for how do I develop these customer interview questions? How do I find my perfect customer? Of course, I write about in the book. Obviously, that's the idea. But they're all there. So you could just do it and see if you like it. Like hit it. Yeah.
Brian Bell (12:10.956)Pretty bold. I love that. Yeah. that's a good idea for a book. You know, just kind of stand up a GPT kind of thing in front of it.
Jason Cohen (12:30.007)Well what's so great I'm not a I don't have to make any money on the book. I'm not a consultant, I don't sell courses, I don't you know, there's no products coming. I'm this is for the love of the game. So I don't mind giving away all the stuff. I just want people to have the information and a and a dead and dead trees cost money. So that's the only reason why the book costs money, right? So
Brian Bell (12:47.629)It sounds like there was two kind of what you call hidden multipliers there in that story. You know, one was the pricing, right? Your price was too low. and it was too high. Okay, tell us more. What is a so how do you define a hidden multiplier and yeah.
Jason Cohen (12:55.799)Too high. Yeah, because they would have to, because they'd have to buy, because they'd have because in order to have these five sites that are all little, they'd have to pay like 30 or 50 bucks per site. And they're like, no, no, no, I'm already giving you a hundred bucks and I'm happy. Let me just fold it under there. No, our pricing was too high. But in specific ways, the answer wasn't just reduce all the price on the website, right? That's not what they were saying. It was how is it that we can achieve we can keep our higher price for a lot of service, but in what other ways can we work with people? Right.
Brian Bell (13:12.232)interesting. Yeah. So you uncovered a hidden multiplier that what you might call price segmentation, bundling seg segmentation, right? So you were able to kind of say, these agencies that run 10 sites, they shouldn't pay $300 a month. That's like too much. but $100 a month allows us to get a lot more revenue than we would have otherwise had. And you only found that out by just talking to your customers.
Jason Cohen (13:34.411)Yes. With a structured interview. That's the key. Like if you just talk, like who knows? It may be at random you'll find some stuff, but that's not very good use of your time. Or your customer's time for that matter. So you it needs to be a structured, thoughtful way that you're gonna get the information that you need without, let's say, leading the witness into saying what you hope that they say. So that's that is what this this particular process I use and I've used many times. that's that's the idea behind it.
Brian Bell (13:54.753)Right. Yeah, so walk us through Yeah.
Jason Cohen (14:17.061)But I wouldn't I wouldn't just say it's it's lower price though. It's also the price structure. So in other words, remember what they said at the $99 site, I'm fine with that. So just saying you should pay less isn't isn't what they were even asking for and not what we should do. It was a structure around but in what configuration do they want to pay more or not? So like when they think about their big sites, they're fine paying more for every site. They weren't even asking for a deal. It was another kind of thing that they wanted to structure in there to So in this case it's a pricing structure of like in what when is it that I want to spend more with you? And I'm fine with that. I think it's fair. And when do I think it's not or and not necessarily it's not fair, but like this would be great if that's a structural question, not just like what is the price on the on the on the on the website.
Brian Bell (15:05.421)So what are some of those interview process processes and techniques that you use that you describe in the book?
Jason Cohen (15:08.685)Sure. Sure. And again, here there's an article that you can read and the skills, so you don't have to get the book for this. so briefly it goes, I'll just here's the brief outline. You start with what are the goals of the interview? So if it's a brand new company, for example, you might have goals like what are the titles, like the job titles of the people that would buy this? How do their budgets work? What makes them say You know what? I need to move my WordPress site because I'm not happy. Because that's not usually what people say. Usually they don't think about it at all. So, like, what caused them to do that? So, I call that an inciting event. What made them say, Today's the day I'm gonna buy something? Because you need to find people in that location. So, you need to understand that because that's where your like ads are gonna target, and you know, right? Like, like you're gonna need to know that. Another one I really like is what language do my customers use to talk about the concepts? So, in other words, A a great example, I actually don't give this example in the book, but a great example is Discord, which of course is chat rooms, right? When they were starting, they were trying to sell to people who used gaming chatrooms. Like they would be playing a game and it was a chat room for that. and so they would say things like buy hey, get Discord, you can have a free gaming chat room or whatever language they used, because that's what it is. And people weren't getting it, not even for free. What they figured out when they were talk the customers is the way the customers thought is we need a Minecraft server. Because they were thinking we got to host Minecraft and then chat there. So their language is I need a Minecraft server. And so when so Discord changed their text to be get a server. Now you and I know it's a chat room. Like, what do you mean server? It's not even really a server. It's not literally, it's not a server, it's a chat room. And there isn't a physical, right? But when they changed it to get a server, people got it all the time. Right. So that's what I mean by what language do my customers use? Not what I would call it. What do they say? Because again, that's the language I need in my ads, my homepage, my pricing, you know, maybe keywords. Not what I think it is, what my customers think it is. I don't even care if they're right or wrong technically. That's that's irrelevant, right? But what is that? Okay. So there's more, but you get the idea. These goals, what am I trying to learn? So there's some interesting things, right? Then how would I test? Well, sorry. Ne next what you do is you say, What are my hypotheses about the answer? Like, I think they'll call that a chat room. Well, that would have been wrong, but that's that's a good hypothesis, right? or in my case, an example I do give is I think people are worried about security online, like that they would so then they might want to pay to have a more secure website. Not everyone, but like I think a lot of people who would care. That's a hypothesis. Okay, so you write all these hypotheses down. Which is weird because the whole point of the interview is to find out the answer. So, what are you doing guessing? Who cares what you think? Who cares what your guesses are? And the answer is there's all these studies that show that only if you write down your predictions first will you learn and update your knowledge. Otherwise, you just retcon yourself. I thought that all along. No, you did it. And now you're not learning, right? So there's a lot of science behind that, but the point is you write it down and then you find out. Now you have to ask questions.
Brian Bell (18:16.693)Mm. Yeah.
Jason Cohen (18:28.811)You can now line up questions with hypotheses. It could be one-to-one, it usually is, but it doesn't have to be. I mean, this is you come on, let's not that formal. You do what you want, but okay. Now the key here is not to lead the witness. So, like, I think security is important. So, here would be a bad question for that hypothesis. Hey, the internet's a scary place, right? Did you see like every day there's another news story about AI hacking, something important? Like it's bad out there, right? Do you wanna like have a secure site? Okay. So, of course, everyone would say yes to that, and you haven't learned a damn thing because that's not what's in their mind necessarily when they come to your homepage. Right? So, a better question that would not lead the witness is something like, How do you think about website security, if at all? Or what do you do today about website security, if anything? How did you come to think that? Like, what was your thought process of what you needed to do or not need to do, or did you not think about it? See, I'm not suggesting the answer. I'm just trying to find out how they think about it. Right? So I did that, and guess what? I was wrong about the security. They don't think that. Here's what here's what everybody said. I say everybody. Here's what 80% said. I'm just a little website. No one's gonna hack me. And they're totally wrong, right? Now, what they are correct about is a nation state won't try to hack your website because that's you're right. They don't care about you. But that's not what almost all hacking is. It's just people running scripts against every domain that in existence. Just trying to find a machine they can take a hold of so that they can do whatever they really want it to do, like bounce through there to hide their IP address, or send spam email, or send fraudulent clicks on a website. Like there's other things they want to do. They just want to control your site. They just want to serve it to that they can hack. They don't know who you are, but nobody knows but you know, almost no one knows that. You have to be kind of a technical insider person to know that. So I was totally wrong. My hypothesis was wrong. I mean, I am right about how the world works, but Totally wrong if that's what customers think. Because I asked open-ended questions, I was able to discover that, right? So what you generally find, well, I don't want to put really numbers on it, but let's just say even a third of your hypotheses might be right, but even so, you'll get more detail that like makes them even more precise. So great. And then a third of them will be wrong, just like me. And you'll find out what is right. And of course, with open-ended questions like that, hey, how do you think about security? Right. It's not a yes or no question. So with open-ended questions, they'll generate more stuff for you. And the shit you didn't even think about that now, wait, that makes me think of a new hypothesis or a new, I wonder if other people think that. You know, it's okay. So you're now you're learning, exploring new ideas, which may or may not turn out to be true. Could be a one-off thing, who knows, right? But now you're getting that. So that's what happens. And so I did so many of these interviews, and so my hypotheses changed. Again, you you don't have one interview and change your mind about everything. That's not quite right, but you know, you morph over time as you hear stuff. And so Your hypotheses and maybe even the questions can change too as you as you interview because you're trying to learn. Now, it if things aren't going well, what happens is you never converge on any truth. It's just everyone you talk to says something different. You're like, I'm just not seeing the patterns here. It's just there's nothing I can hold on to. That doesn't mean there's it's impossible to make a company here. It's just unclear what it is. But sometimes like WP Engine, they'll converge on like, my God. 80% of the time they say this, or man, I've discovered there's two segments. There's little small businesses and there's people that work at big companies. And the small businesses will answer these questions this way and the big company this way. And so we're going to tackle this one. And, you know, and it just gets to be boring almost to do the interviews because you just know what they're going to say. And that's when you stop interviewing. Because if you're not if you're bored, you're not learning. If you're not learning, there's this is a waste of your time. So I always say you should do at least five of these. If you do five, you'll realize you're not done yet and probably do 10. You'd I who knows. I know people have done over a hundred. I think that's probably delaying. I think that's probably a procrastination of getting stuff done. but I think not talking to your customers, not doing this, is dumb because of course you'll build the wrong thing. So that's a waste of time. On the other hand, doing this too much, of course, that's also waste of time. So all right, you have to you have to like do it. This technique can help you learn some things, and so that's what you should do, right? Anyway, so that's what I did. Before WP Engine even started, that's what I did again to test pricing and over and over again with different products we would build and different other things. I used the same thing.
Brian Bell (22:44.781)There's lots of founders, you know, probably in the camp A that we were just describing, like should I build a business around this? Right. which I think you just described pretty well, like how to, you know, go do that customer discovery and try to figure out if there's a business there. Yeah, that's one way. Okay, let's let's talk about some of the other pieces. Yeah, because lots of entrepreneurs listen to this podcast.
Jason Cohen (23:00.205)Well that's one way. One piece. Well, yeah. So to me, a there's a lot of ways to think about this, and they're all probably about right. One way to think of it is I think of it like a Drake equation, you know, the Drake equation with aliens. So for people that don't know, it's and even for people who do know, the idea is you say, what's the chance that there's aliens? And you're like, well, let's let's let's let's make a big equation with variables. And so you're like, well, there's this many galaxies, and every galaxy has this many stars.
Brian Bell (23:18.284)Yeah.
Jason Cohen (23:33.632)And every star has on average this many planets, and on average this many are in habitals habitable zones as if we know what that is, you know, for other kinds of life. And so on and so and what's the chance that life develops and what's the chance that okay, you know, you do all this stuff, right? And so you have all these variables and you say, Well, that's the chance of life. And of course, part of the problem is we don't even know from a power of 10 what those are supposed to be. And so it's, you know. But his point of making it was to say, actually to try to explain why don't we see any? Because when you do that, you kind of
Brian Bell (23:39.542)Right.
Jason Cohen (24:03.553)You often compute that we should that there should be billions of civilizations. And so why aren't we seeing any? And maybe which of these variables are right or wrong in our equation. And I'm not saying that's even the right equation, but that's what I mean by the Drake equation. So I think of a startup like a Drake equation, where a ton of stuff has to go right, or at least right enough, or else it dies. And that explains why startups usually don't work. Because that's not enough to say, like, I have a good product. Yes, but also you have to get attention for it. And there has to be enough people who want it. And you have to be able to get to them. And that has to be not too expensive. So it's just unprofitable to get to them. And they have to have a budget. And they have to want to buy today. They, if they're locked into a three-year contract, it still doesn't matter. And they have to want to buy from you and not a competitor. So you need some way to be differentiated. And you have to be able to build the thing in the first place. And you know, it's just go down the line of like all these things have to be true, or at least true enough, or not fatal, let's say. and when you multiply a lot of probabilities like that.
Brian Bell (24:50.667)Yeah.
Jason Cohen (24:58.059)The total probability goes down, down, down, down, down, because for all the and it's a bunch of ands, right? It's very hard for all that to succeed. So if you like that mental model of how what why startup is low probability, it suggests answers to well, how would I increase the chance of success? so for example, you'd say, well, just from my natural strengths and assets, meaning I have these capabilities, you know, I know how to do marketing or I know how to do engineering. So just from my strengths and a s some of these are actually a higher percentage chance of working because I'm good at it. So that's not going to be the reason I fail. So if I'm an engineering, an engineer, it's very unlikely that the reason my company fails is because I'm unable to build it. So what what does that mean though? But let's let's just say marketing, because it's often the case, is my weakest suit. I've never done it, I hate it, I don't know what to do. SEO's broken now, so I can't hide behind my screen and just make articles with AI, that's not gonna work. my gosh, I'm gonna have to actually learn marketing maybe. I don't know how. So that's a very high risk. That's one of the reasons why you're gonna die. Now, as an engineer, if you know that there's a project and some aspects of the project are high risk and some are not. Which do you tackle first? We all know the answer. The high risk ones. We have to tackle the unknowns first, because if we do the stuff we know how to do, of course we'll just be a few months, we'll be a few months down the road. We'll have accomplished exactly what we knew we always could. So we haven't learned a thing. And all those unknowns are still ahead of us. And if we're not able to solve them, we'll have wasted this entire time. We could have found out in the beginning that this was impossible or we couldn't surmount the things we didn't know. And saved all that engineering time that we already knew we could do. So we know this from engineering projects. So if you're looking at your Drake equation, you say, what are my lowest categories that are probably going to kill me? And I should do those first. But that's not what we do. We say, but that's uncomfortable. I'm going to do the comfortable fun thing, which is to make features. And you're like, all right. And then you're going to, it's not going to work. Right. So that's one thing you can do. With this equation is realize what is this math telling you to do? Right? also you can ask things like, Do I need a co-founder or someone else to help me with some of these things so that it's overall more likely? Because a lot of times what happens is if you have a co-founder, it's like two engineers. Well, that didn't help. Right? Part of what you can do with more than one person is cover more of this equation. Right? And of course I'm being, you I don't take it too literally, but you know.
Brian Bell (27:08.631)Right. Yeah.
Jason Cohen (27:32.332)Like this is a help, this is a helpful way to think about how do I de-risk this business. Another thing you could think is could I pick, since I'm bad at one of these things or I'm scared of one of these things, could I pick a different strategy? Meaning, who am I selling to, at what price? Why are they picking me? What are the attributes of my product? What does it do or not do? What are those trade-offs? That's what I by the strategy, these major decisions about what is it? What am I doing? Could I pick a strategy? That sidesteps some of those weaknesses. So, like, I hate sales. Okay, well, then I need a product-led growth strategy. Now that may or may not end up working, but the sales one definitely wouldn't because it's running right into a huge weakness of mine. Yeah, or a sales co-founder, and then it should be, then I shouldn't use product led growth because then that doesn't exploit that person's strength. And that's what we need to do. Every good strategy exploits strengths that you have, of course. So you could think that what is there a different strategy that's still logical.
Brian Bell (28:13.517)I need a sales co founder. Yeah. Yeah.
Jason Cohen (28:29.441)cons you know, cohesive, coherent with itself, that would avoid some of those things I don't have so that I don't have to fix it. Or at least they're not they're not as likely to kill a whole company because they're not that important anymore. Because I've picked a path that doesn't require me to do that. So that's another thing you could do with that. So I think while of course it's just a mental model, it's quite actionable. I mean I just told you three different things that it could that you can do with it just with that idea. so I do think that's useful. And so that's why Should you interview customers? Yes, because you're you're you're figuring out some of this stuff, but there's all these other pieces too. So interviewing customers doesn't maybe doesn't tell you some things like how many of them are there. it might tell you something about your competition, maybe. That's a little harder to tell. I I I do think there's a lot of that you can learn by interviews. And actually mapping that out might help you know what are my goals. Well, my goals are to figure out the answers to a lot of these things, you know. So that might be helpful.
Brian Bell (29:27.339)One thing you mentioned is you throughout shipping new features and new, you know, product lines, you kept talking to customers. How do you know how much and how often to keep doing that as you kind of scale?
Jason Cohen (29:40.471)Keep doing a new feature.
Brian Bell (29:42.571)Yeah, or like if like I'm expanding the product or yeah, I'm gonna I'm trying to prioritize my roadmap or how do how do founders and founding teams know when to go talk to customers and how often how much.
Jason Cohen (29:53.803)All the time. If you're not talking to him a few times a week, like I don't know what you're doing.
Brian Bell (29:58.872)So all the time. Got it. Yep. One thing in the notes is you shared data that showing negative reviews can lift sales and lower returns. Walk us through that. What is that?
Jason Cohen (30:10.283)Yeah. So this is I one of my soapboxes for a couple of decades is that honesty works in business. Especially with so many businesses not being honest. It's refreshing. And I have lots and lots of examples all over the place of in of in what ways does it work. And this is a one of the really fun ways what you're touching on. So the date there's data about this from a company in Austin called Bizarre Voice. Bizarre Voice, 20 years ago, crew powered all the user generated reviews on websites like Walmart and eBay and Best Buy and like all these kind of companies. At that time there were no the again, the internet's brand new, so we don't have stuff like that. And nobody knew how to implement anything. So they implemented like a a a a module so all these companies could have user generated reviews on all the products. And the retailers liked it, but so did the Sort of the manufacturers, like Rubbermaid, wants to know what people think about the Rubbermaid product so they can improve it and all this kind of stuff, right? So really interesting, useful data. And now a lot of there's lots of ways to get it. Back then that was new. So they learn things about what do you show in review? So here's what you would think, or I would think most people would think I think is you want to show the five-star reviews for the thing. People saying, this is great because reasons, and that will help people buy the thing. But that's not actually what increased sales. When they Tested just the five star reviews against five star reviews in column in the left column and all the bad reviews in the right column at the same time though. So the person was shown both. Sales went up and returns went down. And why did sales go up? Now you can make up your own reason if you want, but right away the but the bottom line is if you're honest, you make more sales. And by returns going down, you might say, and your customers are happier with the product. You might you might conclude.
Brian Bell (32:02.081)And they also know why people returned it were unsatisfied, right? yeah.
Jason Cohen (32:05.025)Well, but so here's here's I think why this happened. I could be wrong about the why, but the conclusion's the same. when you know what the trade-offs are, which is what you're seeing, it's good at this, but it's bad at that. That's called the trade-off. I know the product has trade offs. I know it's not a perfect product, but I don't know what they are, so it makes it hard for me to buy it. Once you tell me what they are, I can decide if I care or if this is a trade off I like. Many times the good stuff is like. Ooh, that's it. See, that's exact that was my number one thing. It's that thing that it's good at. This thing it's bad at, I don't even care. Or yeah, that would be nice, but like I care less that it's got that problem than I like that it has the my number one thing. So I'll buy it. It's a trade-off that I want, that I'm I accept. And so if another product has a different trade-off, I want this one. And if I don't know what the trade-off is of this other product, I'm I'm concerned about buying, and I'd rather buy something I understand that is a reasonable fit to what I want. No matter how analytical or not analytical I'm I'm consciously being, that's the case. That's why I think sales go up. Now, what about the returns? This one I think is more obvious. If I already know what the problem is and I get the product and it has that problem, I don't return it. I expected that. It's not a surprise. It's a it's when it's a surprise and you're like, my God, and you return it. And you might be right, because you might have gotten a bad trade off or whatever, right? Now, for manufacturers, returns are Horrifically expensive because manufacturers may only have a margin of five or 10%. So a return, especially if it's gets lost in the mail, it's damaged or you know, it might take them 10 or 20 sales worth of profits to replace the one loss from that one thing. Insanely, insanely expensive. It's even expensive for software companies because you still spent money on things like getting the person in, maybe tech support. So like it still costs money, it costs them a lot of money. Anyway, the bottom line is reducing returns. Is actually an enormous financial benefit. Like if you change it by even 5%, it's enormous for a Walmart. So anyway, that's that's the story of why being honest and showing the negative reviews is actually quite valuable. And I do think you can extrapolate the more general lesson of being honest works to many other aspects of the business.
Brian Bell (34:23.123)so distribution is the new moat, very popular right now. What's your take on that?
Jason Cohen (34:31.105)So I have a more nuanced take on that. First of all, what do you mean new? I mean, if 10 years ago I had an unassailable way to get sales for customers that stay forever, is that not a moat? So I first of all disagree that it's a new idea. Now I do agree that with AI, we can make stuff faster than we could before. But so can you so can your competitors. So if we're both making stuff faster, both making stuff slower, it's still a tie. So It was always easier to make a new feature than it was to get people to buy the damn thing. Always. So again, is that new? And I would say yes and no. It's always been harder to get distribution than to make the product. And it's always been true that if you're really good at distribution, you can win even with a mediocre product. People have been writing about that for 20, 30, 40 years. So that's not new. I do agree that it's so fast to develop features that makes that you could say that makes the balance even greater. But the general concept is not new. I have articles myself going back 15 years that say that. So it's not new, even if it is more exaggerated. And I agree that it's so much more exaggerated that's different. And I agree with that, but let's not get carried away. The other thing is, is it a moat? So a moat has a definition. You know, Warren Buffett coined the term, it has a definition. It's a competitive advantage that even a well funded, intelligent competitor cannot take away from you. Like no matter what they do, you still have this advantage. That's what a moat is. And you might say long term. Some people say permanent. And then you could say nothing's permanent. Okay. Let's say in the software world, at least 10 years. Call it that, right? so is distribution a moat? And my answer is some is and some isn't. Five years ago, people would tell you that SEO is a moat because once you're locked in on Google, you can stay ahead and forever and ever you'll get that. And Google has now proven it's not a moat, it's gone. You like that? You like that moat? So is that a moat? And I would say, who cares? I don't care to I don't care to label it a moat or not. I would just take each marketing channel by itself and ask, what are its attributes, what are its risks? Like the risk with that is platform risk that they change it on you. Are you okay with that? Early on, you have to be okay with platform risk because you know you can't you can only do so much. So you just pick something that you're good at, that your customers are also there, and whatever's in that intersection in the van diagram of what you can execute well.
Brian Bell (36:29.79)-huh.
Jason Cohen (36:58.241)And where your customers actually are, you just do it. That's what you do. Now, later on, if you're WP Engine and you have 200,000 customers and you've and your you know monthly paid marketing budget is many millions of dollars a month. Of course, then you can say, We should be on, we should diversify, we should have many channels because you never know what's going to change. And we could have different teams or people who specialize in them to mine each one as much as we can, and we have a portfolio, and that becomes correct when you can afford it, but not at first, right? So Those are the ways you kind of deal with that, I think. So I do think you could say some of those are more permanent or more moaty than others. Again, I just don't care for the to use the term because I think it's distracting from what's really going on. I just don't think it's that useful to argue about the moat part. so for example, I could I would I do say if I'm just buying ads, of course it's not a moat. someone else can run the same ad and I don't know, like it's also a platform.
Brian Bell (37:54.19)Right. Yeah, unless you can make the ads, you know, I don't know, twice as good as your competition and that's sustainable somehow. Yeah. Yeah. Right, right.
Jason Cohen (37:58.786)That's not a moat, they can do it too. They can literally copy what you say. So it's not a moat, right? But brand is it is more moaty. Apple's Apple. They come up with a new i, i, i, iPad mini, it's way too expensive and they're gonna sell out. Why? That's whatever that answer is a moat. Because it's not like no other companies aren't spending billions of dollars a year trying to compete against that, right? That's a moat. Okay, so that's a moat. so brand can be a moat, still, right? So okay.
Brian Bell (38:21.013)Those same products, right? Yeah. Mm-hmm.
Jason Cohen (38:29.227)I also think that having a personal marketing campaign. So like when you're a YouTube personality, of course, there could be infinite other YouTube personalities, but they're not taking away from the fact that people like to watch you. Right? Like just because there's someone else there doesn't mean n but now I don't want to watch this person. What? You know, that's not true. Right. It's not you're not I mean, you're sort of competitive competing for attention, but like they can't they can't take it away from you.
Brian Bell (38:45.153)Right. Yeah. Yeah.
Jason Cohen (38:58.475)You know, like we're the only ones with this feature. Well, now 10 people have that feature. You c they can just eliminate the idea that you're the only one with the feature. But a ch a YouTube channel that has a million views, of course, I know that's a lot, but someone else can't just remove that from you. Now there's platform risk, I suppose, from Google, but if you have a lot of if you have a million subscribers, which again, that's that's a lot, but if you do, like, yes, but come on, that's pretty good in this in this day and age. Like, wouldn't you call that moat-ish? You know, right? It's close enough, like without arguing about it, the terminology too much. That's pretty good.
Brian Bell (39:11.841)Right. Yeah. Yeah.
Jason Cohen (39:28.023)Better than ads, you know. All right. Like, so that's how I would do it is I would I would stop saying moat. I would say, like, which of these is it hard for other people to do anything about? Which of these do I think I can take for the long term? But there's downsides. Like, if it is a personality YouTube thing, it's tied to you. That means you can never sell the company. Or you can, but like this doesn't come with it. So there's a downside. So
Brian Bell (39:48.673)Yeah. That's why I didn't call my V C firm Bell Capital. I called it Team Ignite. That's exactly why.
Jason Cohen (39:51.882)Right. Right. So it's so that but that's a trade off. So does it is it a moat? Well, I don't you know, a moat you could if a true moat, I could sell the whole company, it would still be there, right? I I think. I mean again, you could argue with me. And again, and they and the answer is I don't care. Like that's just how I think of it. I I I don't get to decide, you know. But that's what I would do is ask these kinds of questions. What kinds of risks are they? Am I okay with that? When do I want to diversify? And what does that mean? you know, it If I do have so for example, if I have what's not diversification is saying I'll do SEO on Google and then SEO on Bing. That's not diversified. But SEO on Google and then paid ads and then affiliates. Okay, those are all very different things. One of those could explode and it wouldn't change the other ones at all. Might even help them. Like if search traffic goes down, affiliates with big newsletter subscribers might actually increase in their power because you no longer can get it from Google. Maybe the newsletters are more likely where you get it from. I just made that up. I don't know if that's true. I'm just saying when it's diversified, just like markets, sometimes when one goes down, another one might go up. Anyway, you're doing that for yourself. So you could ask what's uncorrelated or different. Because if my goal, if my goal is to diversify marketing, to make myself more sustainable and predictable, well, then I can be intentional about that, right? and again, we're we're talking roughly, but nevertheless, that's that's smart. So which of those is a moat? I don't know.
Brian Bell (41:17.069)A lot of good thoughts there on Motes. so you've written about the myth of exponential hypergrowth and the elephant curve. I've never heard the term elephant curve. Maybe you could help the audience understand that.
Jason Cohen (41:17.358)Yeah. That's because I made it up. I made up the word. That's why you haven't heard it. I've made up many things and I've never named it after myself. So yeah. So this also is an article, so people can just click and see it. yeah, so this was an article that took me over a decade to write because I had this notion forever ago, and I didn't know if I was right or not. And I always wanted to like research whether it was right. But I was building companies, I don't have time to, you know, research stuff.
Brian Bell (41:34.647)That's good.
Jason Cohen (41:56.726)And so every time I came across another like example of my theory, I was like, see, there it is again. And I would capture it, throw it in a note. So it was 2022. And I I had a one month sabbatical from WP Engine. And with that little break, my mind sort of went back into writing and thinking, right? And I I came by my note, I'm like, I think I'm in the headspace where I could do this. And I did. So I have this full theory and all this data. And so the resulting the result of all this is it's sort of a magnum opus of data of mine, I suppose you could say, on this concept, which is people say companies grow exponentially in some cases, like the Facebooks of the world, the Slacks of the World, you know, the answer, they say exponential. And maybe with AI, it actually could be true. But even with Facebook and Slack and these type of things, it's not true. And if you look at their actual growth of things like users or revenue.
Brian Bell (42:41.333)Anthropic, you know. Yeah.
Jason Cohen (42:55.509)It's never exponential. Now, the reason I said Facebook and Slack is because in 2022, Facebook was the fastest growing consumer tech company ever. And Slack at the time was the fastest growing B2B company ever. If they're not exponential and they're the fastest ever, ain't nobody exponential. Or at least not in a way that's like important to or like, right? Why not? And why do we say that? So that's what the whole thing breaks down. And I break it down into pieces and I guess the I guess the bros would say first principles I don't, because I think unless you're talking about whether the integral of the wave function from negative infinity to positive infinity of the wave function times its complement is one, you're not talking about first principles. You're talking about you're just making that up to sound smart, right? Yes I if we're talking about the axiom of choice, then we're at first principles. Otherwise, yeah, I don't think so, you know. so
Brian Bell (43:44.781)You're just borrowing some term. Yeah.
Jason Cohen (43:53.026)So, so, but I did come I did try to break down my again, it's a theory. So if you like this kind of stuff, then you're gonna love this because it's a fun theory to, you know, you could like it or not, argue with it or not, but it's but at least it's a well thought out data-backed idea. So okay, good, you know. But even if you disagree, it's still fun to disagree with a good argument, right? So I I go and show, like, here's how marketing campaigns typically work, here's the shape of those. And even if I do many of them, you still get this totality shape, which is the same as the shape of one of them. And this is the elephant curve that. Marketing campaigns, when you unlock one, you get this little surge of of growth because you've unlocked whatever. And then maybe in a period of optimization as you kind of get a little bit more out of it. And then often with these campaigns, they they tap out or even start to go down because the channel's diminishing, or people have just seen your ad a bunch of times, or you know, there's various reasons, which I talk about in there. and so it kind of tapers off. So this shape of like high up in the taper, that's like an elephant, is it? I don't know. I put a picture of an elephant in there and I wanted a cute name, like whatever you'd like.
Brian Bell (44:50.389)That kind of yeah. That's like an S shape kind of, yeah.
Jason Cohen (44:52.097)And then it's it yeah, with that sags. You can call it a saggy s, you know. And then if you stack them in a row, like you do when you when you stack on marketing, that and you add it, you still get another elephant curve. It doesn't change what the curve is mathematically. And then I have actual data from WP Engine where we did this. And in fact, the data is exactly this. And our marketing campaigns did exactly this, and we added it, and it's a whole thing. And of course, then I show, yeah, then I show data from all kinds of companies.
Brian Bell (44:56.353)With the sag. This is a sagging S, yeah. Which is an elephant trunk. Yeah, I get it. That's why I got out of marketing fifteen years ago.
Jason Cohen (45:21.015)Facebook, Slack, HubSpot. I show it from bootstrap companies, from VC companies, fast growing, slow growing. I have all this data and they all show these same kinds of effects. And so this is my argument why it's actually a parabola, not or linear even, and not exponential. I, of course, it's all this data-back stuff. Even the growth of the internet is a pr is a very clean parabola. It's like really, really tight parabola, not exponential, not S-curve, even parabola. So I you know, it's kind of however you mobile phone usage, like it's just you look at all kinds of things and it's non-exponential. So
Brian Bell (45:56.344)Nothing can continue forever exponentially, right? That's the there's always limits to growth. Yeah.
Jason Cohen (45:58.828)Yes, cancellations do. Cancellations do. This is this is see, this is part of the application of this knowledge. If you know that marketing is maybe linear, at best quadratic, then you turn around and say, wait a minute, but what about customers leaving if I have a subscription revenue business? What's that like? And the answer is that's exponential. That's why we say 2% per month. The words 2% per month mean exponential. That's what that is, is an exponential curve.
Brian Bell (46:27.821)Well, I mean, but the flip side of that would wouldn't two percent growth per month also be exponential? Hmm.
Jason Cohen (46:28.471)Right. Yes, but that's not what happens. What happens is you have percent growth and that percentage diminishes over time. And there's even data, which I have in the article showing that on average, that percentage goes down 15% every year for public SaaS companies. Well, if the percentage consistently goes down, that's because it's not exponential. And when you plot what that is, if you plot literally, I stay say start at 100% year over year growth, but that goes down 15% every year, which is the actual data. If you plot that, it's a parabola again.
Brian Bell (46:37.739)Right. With that strategy. Yeah. Right. Right. Right. Yeah.
Jason Cohen (47:02.517)It's right back to the elephant curve parabola argument again. Because it's not exponential. We use those language, but it's not. But cancellation is. Two percent of your business leaves every month or whatever. That's true. So because as you grow.
Brian Bell (47:15.991)Really interesting. So you're always you're almost always fighting entropy in the universe of business, you know? Yeah.
Jason Cohen (47:21.993)As you grow, cancellations grow with you exponentially, but marketing doesn't. So cancellations and early on, you don't have any customers. So cancellations are small. So marketing wins and you grow fast. Later on, X the cancellations are more than what marketing can make up for and you stop growing. So I call this the growth ceiling. You can compute what that number is. I have another article about that you could read. and there's lots of things you could do about it. You can try to lower cancels, you can try to have existing customers grow. There's a million things to do.
Brian Bell (47:27.724)Mm. Mm. interesting. New product lines, new segments, categories, markets. Yeah.
Jason Cohen (47:51.766)I devote two entire chapters of the book. Yeah. I I I I devote two entire chapters of the book to this question of how do I either make cancellations better or grow other things that combat cancellations that are also exponential, because this is such a major impact on growth. It's worth all of that conversation about, well, what will I do about it? Like it's worth that much yapping about it because that's how big of a deal it is.
Brian Bell (48:15.671)Yeah. Really yeah, I gotta go read these articles. I think you missed your calling as a professor of business.
Jason Cohen (48:23.103)At WP Engine, my nickname was the professor. And but again, that's why, that's why this stuff is just a labor of love. I just love doing this. I love talking about it. And these ideas, I mean, some of them, like this, like this exponential one, I've been thinking about it for so long. So these are not things I've just dashed off of my top of my head. all the articles, a lot of them are like, there was this idea. We tried them at WP Engine. We, you know, it worked or didn't work, and we and then I have hundreds of things in drafts I never published because it just wasn't good enough. I couldn't.
Brian Bell (48:26.376)awesome. Yeah, see, there you go. Yeah. Yeah. Yeah, you've seen it over and over again in your career. Yeah.
Jason Cohen (48:51.521)really understand what I thought about it or it in retrospect doesn't seem quite right. And so I never published it. So these are only things that again, they're still just opinions. I mean, half the stuff you won't like or you won't agree with or it won't apply to you. I know that. That's okay.
Brian Bell (49:02.893)Well, most of most of knowledge is just an informed opinion or theory based on the data, right? Yeah.
Jason Cohen (49:07.915)Yeah, and contextual. Right. And very rarely do you hear advice or a framework with the context attached that says, and here's when this applies, including my own. I'm guilty of that too. I know. I'm I'm admitting it, right? So I'm just saying, since even I don't even I'm bad at saying that, you you the reader has to decide for herself like what is what if this is applicable to me, what do I like? And yeah, you're on your own. But all I can do is say, look, at least this is at least these are like decent ideas. That's all I can say.
Brian Bell (49:40.109)Speaking of cancellation, you've said that it's too expensive is almost never the real reason customers cancel. What do you think is going on instead?
Jason Cohen (49:48.408)So the background of that is when you ask people why did you cancel, which is a good idea, a lot of times it's a drop-down list or something, right? And too expensive is almost always, no, I would say always in the top three, and often number one. And so what I see people do, including even at WPN, what but teams would want to do, is they say, let's make it cheaper. And my argument is that is never actually the answer, ever.
Brian Bell (49:55.703)Yeah. Wasn't seeing value, like whatever. Yeah. Right.
Jason Cohen (50:18.643)Here's my argument. How did this person sign up in the first place? They came to the website, they read what you promised, they got to the pricing page, they saw the price, and they said, Okay, it wasn't too expensive then. The people that think it's too expensive, they might exist, but they're not your customers. They're the people that didn't buy. Now, is that possible that you are too expensive? Yes, but The way you'd find out is by talking to people who never bought, which is very, very hard. But hey, if you can do it, or maybe you see it in reviews, yeah, right. Like there's ways okay, so that could be, but when someone who's been when someone who's been a customer for a year cancels and says it's too expensive, that cannot be the reason.
Brian Bell (50:57.197)Abandoned carts would be an example of that. Yeah.
Jason Cohen (51:05.441)So what is going on? Well, there was some promise you made or some expectation they had, whether you promise it or not, that's not being met and therefore it's not worth the money. And that promise is the reason, or that expectation is the reason. And so you promise three things, but that third thing, it's not do not working for them. That's why they canceled. So if you were to make the price lower, they would still have signed up because they already did at this price. So of course they would sign up at the blower point, but it's not like they'd still be here because you still promised something and that still didn't work for them. Or again, maybe they're confused. They thought the product would do X. You never said it did, but they thought so. Okay. You could say that's not your fault. Maybe you could be clear, or maybe it's just random. That's okay. But still lowering the price is not the answer. So again, lowering the price is never the answer because that wasn't the cause. At least not if that's where you heard that from.
Brian Bell (51:56.866)Right. Yeah. so through all this, you've actually invested in over sixty startups. what have you learned after so many investments?
Jason Cohen (52:08.267)Well, what I know about investing is deal flow is the most important aspect. And the way to see this is if you see great deals all the time, it doesn't really matter how good your selection process is. If you just invest in a bunch of actually good things, you'll you'll be great. And simultaneously, if you don't get to see the good deals, if the best companies out there don't cross your desk, then no matter how good you are at picking them. You will lose money because you won't have those big winners that of course you need when you're investing. So to me, if you don't have good deal flow, you won't be a good investor, period. The end. You have to invest in that. Whether it's an angel or an institutional investor, that's the deal. So I know I don't have great deal flow and I don't work on it. And I'm not that interested in working on it. That means I shouldn't try to be an investor. So I don't. And instead, what happens is my attitude is: am I having fun? And I'll just use money and I don't care if I make money. That way it's okay if I don't get in the big deals. I don't care because that's not I wasn't getting to get the big deals. I would like to lot not lose money, but I'm not trying to maximize returns. And therefore I don't need to be in the big deals. So as a result, I have actually really good results for an angel because the median investor loses money. And I'm up about four X, and that's much better than losing money. But if you think about four X over, say 12 or 15 years.
Brian Bell (53:11.437)It's like a now it's just a hobby, right? Yeah.
Jason Cohen (53:36.299)And the S P five hundred is up mm eight X over that same period? I could have just bought an ETF.
Brian Bell (53:42.541)it's not up eight X in that same period, but the S P five hundred? I don't I don't think so.
Jason Cohen (53:44.673)Yes it is. False. Yes. Yes, the bull run was seven X. That's right.
Brian Bell (53:53.336)From two thousand nine, like from the bottom after the GFC or Yeah. Yeah. Right.
Jason Cohen (53:56.14)Well, yeah, sure. If you pick it you have to pick it, which is of course you can't pick it. But hey, in companies you're picking companies, right? If we're gonna we're doing like with like. if you had invested in Google and Amazon and stuff, you're definitely doing better. So it's better to have put money in Google and Amazon on the public markets.
Brian Bell (54:08.375)Yeah. Yeah, if you're if you're not three Xing three or four Xing gross on your angel portfolio, you're probably better off just putting it in the Nasdaq or some index fund. Yeah. Over any like random ten year period of time.
Jason Cohen (54:19.317)Over what period of time? Over what period of time? No, 10 years. Yeah. 10 years is it's bad because of the risk. Like, cause you probably won't forex. Like that, that, that's a really, really good result. That no, you have to expect it might be much worse, right? So the bottom line is like, so instead, what I do is I invest in things like Tiny Seed, where someone else who I trust has great deal for and I think good selection. And anyway, I love working with them. That way I can just give them my money, they get they give it to good companies.
Brian Bell (54:29.666)Yeah. Yeah.
Jason Cohen (54:52.503)Then I can just do what I like, which is this. I can just do mentor calls whenever they want. I don't care. Like I don't I don't need to get paid. I don't, you know, I don't I don't need to get involved. Like I've already invested in you indirectly. Right? Like we can just we can just help whenever you want. That end of story. So I can just kind of do the fun part. I don't have to do the paperwork or the taxes or the reporting and or pick them or tell people no all the time, which I hate. But right? And so like, okay, I don't have to do that. So I get to not do all the stuff I don't like and I can just have fun. So
Brian Bell (54:59.692)Yeah. Oof. It's hard. Yeah. The hardest part. Yeah.
Jason Cohen (55:21.505)That to me is like a perfect thing for me. so that's my that's that's like part of things I've learned. By the way, this is this general idea of like the top of you might say the funnel is more important than the rest of the funnel, which is kind of what I said. That's true of every funnel. It's true of the hiring funnel, it's true of the marketing funnel. And for the same reason, if you only see great candidates, it doesn't matter how good your interview process is. If you only see bad candidates, it doesn't matter how good your interview process is. And the same thing with marketing. If I just dump shit traffic in your website, I don't care. The design and the conversion and the it doesn't matter because they weren't, they're it's not good. So once you've learned that, you realize everything that is funnel like, the top of the funnel is far more important, far than everything else in the funnel. And you should be spending an inordinate amount of your time and energy improving the top of the funnel. Whereas we don't, we tend to go, how do we have better interviews for hiring? How do I A B test my homepage better. It doesn't matter. In fact, the more garbage traffic that comes to your website, what are you A B testing in your homepage? What percentage don't bounce? But if most of it's garbage, what you're actually testing is how do I get the garbage to stay? But that's the opposite. You're trying to filter out the garbage. So if you put garbage in, your A B test will be worse. You can't use A-B test. You've broken the A-B test system if you put garbage on the top. So it I cannot stress enough how bad it is to put garbage in the top of any funnel-like thing. It's way another hidden multiple multiplier from the book, right? It's be even if you shift the quality a little bit at the top of the funnel, it's drastically impactful compared to anywhere else in the funnel. That doesn't mean the rest of the funnel is irrelevant. I'm not, of course, I'm not saying it that far, but one of these is.
Brian Bell (56:40.885)Really good insights, yeah. Yeah. Mm-hmm. Yeah. Right. Yeah.
Jason Cohen (57:09.909)much more a much bigger lever than the rest. Again, hidden multiplier. Hidden if you didn't know that. Multiplier regardless. Yes. Yeah.
Brian Bell (57:19.285)Really good insights. well, we've seen a lot of changes. obviously, you have in your career. What are you excited about over the next, you know, three to five years?
Jason Cohen (57:29.259) Not having to have a job where I have to worry about AI is pretty exciting. Getting to have missed that. So I can just do stuff like this, which is fun, and not worry about how to have a job as a designer. That's fun. What am excited about? It you it's very it's the reason that it's hard is I actually feel like even people who are supercharged by AI and love that aspect, they say, like, you know, but I'm kind of beat at the end of the day. It's like actually quite energy draining to be context switching and doing all this stuff. And so on the one hand, I did a month's worth of work in a day. That's undeniable, but I'm actually beat to death. And then, you know, you have these people, I don't want to overstate it because you know, some of these things go around.
Brian Bell (57:59.714)Right. Yeah. Right. Yeah.
Jason Cohen (58:11.947)And then you think everyone's doing it, quote unquote, when it's like three people. But there are a lot of people who are like, again, I'm super powered in my as an engineer, as a web designer, as a writer, but I feel hope this help this like, but who cares? Like we've just lost humanity. And like I'm really actually fundamentally sad. Sort of like in COVID, where it wasn't just that we had to stay home and all this. It's like we're losing a part of human our humanity that I think we're still recovering from the shock of that. And or at least many people are, let's say. And
Brian Bell (58:25.517)Yeah. Yeah. Right.
Jason Cohen (58:42.047)AI, I think, even if it's su you're supercharged with it, and I am too I am and I I love that I use it every day and it's really fun. But I also have this feeling of just like, who cares? Like, even stuff like, and don't write prompts, write loops and loops that write their own loops and see it's just doing it itself. And it's like, yes, that's interesting intellectually, and I like playing with it and trying to build it. And then also, what's the point of me anymore? And so I fortunately have like decades of experience and writing to fall back on. So I can be like, well, I can talk about stuff AI cannot talk about.
Brian Bell (59:05.185)Yeah, it's almost like we're we're almost automating ourselves out of yeah.
Jason Cohen (59:11.711)I can write a book, A I cannot write. I can have a podcast, which I do, that A I cannot have. And so that's good. So I guess I'm excited about my little projects, but I'm not excited for humanity about my little projects. So I don't know. Like that's a bad answer because you say, What are you excited about? I'm like, I'm a negative Nelly. And I and again, I'm not saying we shouldn't do it. We should stop. I'm not saying that. I'm just I'm just kind of bemoaning like not all is lost, but rather that's important. What should we do about that? Like I'm not saying stop AI, but I am saying, look, but this other stuff is also true. Can we simply say both are true? Can we just admit that? And they're both important. Can we agree? And so should we spend time, not in lieu of AI, but in addition to AI, can we spend time on this question of humanity and this thing of like, well, people will just make art. it's just not true. It's never been true. And the times when humanity has been able to do things like that, that's not what they've done. And when people d are out of work, that's not what they do. Now I'm not saying okay, you can pick, well, this person did. Yes, I know. I'm talking about 90%. The aristocracy in Europe, when they didn't have to work, that's not what they did, actually. Right. And so even then they they did their version of drugs and drinking and and maybe paid other artists who were starving. So that's not ever what humanity has done with freedom from work. So again, I'm not saying so everyone should work, but.
Brian Bell (01:00:22.557)Through parties and tried on wigs and stuff. Mm.
Jason Cohen (01:00:39.615)Th just saying like everyone will do art what that is abdicating you could say responsibility, abdicating even just the even admitting that it's a problem that is not solved by this kind of flippant statement like that.
Brian Bell (01:00:53.867)Yeah, I wish we could go back and kind of interview nomadic hunter gatherers, right? Who lived off the land and just kind of moved from place to place.
Jason Cohen (01:00:59.947)You can. That's what that's what and be people have been studying them for a hundred years. They have. You can read that. You can read about that.
Brian Bell (01:01:04.951)So yeah, so it's interesting. I think, you know, they only worked, you know, two or three hours a day. And the rest of the time they were just grooming each other and hanging out and, you know, playing games and making love and making art. And so I I think it's gonna be an adjustment because we've been living in this industrialized society where we are required to go trade our time and labor and capital for money.
Jason Cohen (01:01:29.687)We're not. We're not. You you I can show you millions of people in America that do not trade their time for money. But what they're not doing is making art as in general. Mm-hmm.
Brian Bell (01:01:40.983)Right. So what do you what do you think we do in this transition when you know AI is eating labor, obviously, robots are right behind that, right? And that's you know, the twenty thirties is gonna look really different, you know, and that's four years away, three years away.
Jason Cohen (01:01:58.593)Yeah, that again, like this is this is a this is a challenge that we shouldn't tackle, we shouldn't stop AI to tackle it, but on the other hand, we should tackle it and we should expend a lot of energy and time and money to tackle it. I'm not even saying it has to be identical money to AI. You know, like i I'm trying to make this false equivalence, but surely we're not spending enough time and money on it. And of course I don't have the answer offhand, obviously. Yeah.
Brian Bell (01:02:14.935)Mm. Yeah. Right. No, that's why they call it the singularity. It's hard to imagine like what life is like after that, right? really enjoyed the conversation. the with the professor Jason Cohen. where can folks find you online?
Jason Cohen (01:02:35.021)Ha. the nexus is asmartbear.com because I'm a smart bear on Twitter and et cetera. So that's where all these articles are, and then links to these AI skills and links to where you can buy the book. and I have a podcast where a founder goes in the hot seat and opens up their business and with their biggest challenge. And then me and another founder of a related company or who has a related problem, we all just work on that problem together. So it's there's no script, it's not an interview. We don't try to make anybody look good to their dismay sometimes. It's a working set, it's just like a working session. You're a fly on the wall as we as we as we try to tackle something for two hours. And that's the podcast. And so there's a link to that. It's called Smart Bear Live. and that's on YouTube and also all the podcasting thingies. So, yes, there's many things, and all of them are AI proof in the sense that, like, if you want any of that, then you wouldn't want an AI writing that. So, there you go.
Brian Bell (01:03:11.373)It's like a working session that's recorded. Yeah. Yeah. Well, that's cool. Yeah. Right. Well, really enjoyed it. Thanks so much for coming on. Yeah.
Jason Cohen (01:03:37.025)Thanks for having me.