TRANSKRİPT 🔗
Jacob Luetzow: What's up? I'm Jacob Litzo, your Elixir Mentor, and welcome to another exciting episode of the Elixir Mentor Podcast. This is where we discuss everything related to Elixir from interviews with enthusi enthusiasts and pioneers in the community to innovative projects and libraries shaping Elixir's future. Today's guest is not our typical, he's not an Elixir guy, but he's bootstrapped one company to millions in profit and sold it, then founded another that grew into a unicorn, serving over 200,000 customers. And he's been writing about how to build software businesses for nearly two decades. And he just published his first book on growth, hidden multipliers. Jason Cohen, welcome to the show. welcome, Jason. It's awesome to meet you. we were just talking before the before we went live that, you know, I've been reading hidden multipliers and we were kind of joking about, you know, the engineer that procrastinates and just builds features instead of focusing on marketing. And I think that's just a common thread amongst us software developers.
Jason Cohen: Yeah, course, because it's fun to build features. That's why we became software developers and that's what's fun to do with AI. And we know what to do, or at least we think we do. Since we're not talking
Jacob Luetzow: Yeah.
Jason Cohen: to customers, we're not quite sure what to do, but we think we know because we're the customer. Of course, we're not the customer. We're the person making software that
Jacob Luetzow: Wait, you're supposed to talk to customers?
Jason Cohen: we want other people to buy from us. And so our customer is someone who doesn't do that. So we are exactly not like our customers, but we think we are. and then we build stuff because it's fun. And then you forget that you need to do marketing and sales and get attention and make something someone actually wants to buy, not just use. And you just ignore all that because it's hard and sucks. And it's also a bunch of people telling you no. you say, Hey, you want to
Jacob Luetzow: Mm.
Jason Cohen: see this homepage? And 99% say no. That's what a 1% conversion rate means. And then
Jacob Luetzow: Dude.
Jason Cohen: they get to the homepage and say, You want to buy this? And 99% say no. That's what a one percent conversion rate from homepage to purchase is, which is a pretty typical number for a successful company. And actually most of them, you know, aren't successful. So it's pretty brutal. And so yeah, it's more fun to open up Visual Studio or go back to Cloud Code and type. It's way more fun. Unfortunately, that
Jacob Luetzow: It is.
Jason Cohen: doesn't make things that people want to buy. So
Jacob Luetzow: I recently started digging into cold outreach emails and things, learning all about that. And yeah, I think one percent positive reply rate is about what I've been getting. So I get like 600 emails out. So many pissed off people just reply, wanting to be removed from the list. And then I get one like, hey, I am interested. This seems cool.
Jason Cohen: Yeah, cold outbound is is especially brutal.
Jacob Luetzow: Yes. Yeah. Yep. But like I'm trying to grow like okay, before we dive too far into this, for everyone that doesn't know you, which I I think would kind of be crazy, but who is Jason Cohen? And like what what brought you like before WP Engine or even Smart Bear, what kind of brought you to those spots in your life?
Jason Cohen: Well, geez, I've been doing this for twenty five years, so who can remember? I made a start for the
Jacob Luetzow: It is a long time.
Jason Cohen: same reason everybody does, I think, which is you just want to. You're compelled to do it. You don't want to be told what to do by other people. You hate being an employee. You're like, screw it, I gotta do this. if I fail, at least it's mine. If it's bad, at least I did it. And if it's only a little bit successful, at least again it's mine. That's a pretty good reason. So that's what I did. Also was hoping to make money. That turned out to work out, fortunately, but it only took twenty five years, so it was really easy. and so I don't know. I d it's I'm I'm probably the story of most of the people listening to this. I like the code when I was a kid, although that meant something very different than some of the people listening to this. yeah.
Jacob Luetzow: Times have changed.
Jason Cohen: 10 print Jason, 20, go to 10. That's how I got started.
Jacob Luetzow: Ha ha ha.
Jason Cohen: Look at that. and but who cares? Because it was a long time ago and it doesn't doesn't affect anybody else. And we all have our own journeys to get here. I do think we have some things in common. If you're in the arena, you got at least that in common. At least you're doing it. That's pretty cool.
Jacob Luetzow: Yeah.
Jason Cohen: That's so for me that's that's a pretty good gate right there.
Jacob Luetzow: Yeah. All right. So I'm curious. how many products or companies did you start and fail at before you had like your first first success? Was your first success Smart Bear? It was like the PR review.
Jason Cohen: Well, I didn't really do that because, it's it's only in the modern era that people have this ridiculous idea that they'll start ten things and one might work, like a VC.
Jacob Luetzow: Yeah, yeah.
Jason Cohen: and then they point to one of six people who have successfully done that and they say, see, they did it, not realizing that, okay, so six out of how many thousands or tens of thousands of people have tried to do that? So yeah, it's possible, but isn't that terrible odd? So isn't that probably not how to do it? So the way I like to say it is, look, I planted planted all these seeds, but nothing's coming up. Well, d Are you watering the seeds? No, I'm too busy planting more seeds. Okay. Cause that's not how companies go. So now back in the day you couldn't do that. You couldn't just launch 10 companies. Like taking a credit card was almost impossible.
Jacob Luetzow: Mm-hmm.
Jason Cohen:
Load full transcript
Like you needed a merchant account and all this paperwork and personal guarantees. You can't be launching a company every month. The internet barely exists. You can't be doing that. So that's
Jacob Luetzow: No.
Jason Cohen: not how that's not how it went. so The first company was a consulting company, was okay. Then the bubble burst, so forget that. Then then the company called IT Watchdogs, which was okay again, it was a hardware company with embedded software on a little chip, which in in like 2000 was pretty cool and and difficult, and it was inexpensive, which was also something at that time to have something small and embedded and internet enabled. So that was server room climate monitors. So is it hot? What's the humidity level? It measures power. Especially in power strips where we'd have special devices. Yes. And
Jacob Luetzow: It's kind like before IoT was a thing. That's awesome.
Jason Cohen: so it was really it was really useful, really cool. but it was also a bootstrapped and a bootstrapped hardware company is really, really difficult because you have all these upfront costs and inventory. Returns are super expensive, things break. We used to joke that there was a lot of customers where it would have been cheaper to wrap hundred dollar bills around the devices, send it to the customer and say, Never call us. Just don't send anything back, don't call us, just Keep it all. Keep the device. Keep the money. Just don't cut,
Jacob Luetzow: Ha ha ha ha.
Jason Cohen: so that's that's rough. and that was okay. We ended up selling that. I I computed that I made about the same amount of money as I as I was making at the job that I quit to do it. And of course, that's a really good outcome. It's
Jacob Luetzow: I was gonna say that's pretty good.
Jason Cohen: great because it's like, well, it wasn't, you know, because I had this incredible experience and it was fun and you know, all this kind of stuff. then I started Smart Bear and it initially product didn't work well the product worked fine but it wasn't that good of an idea the original product was called code historian and my insight I'm putting that in in in air quotes is version control is is you write to it and you never read it other than coordinating files but there's all this data that could be useful for example if I have not changed this core function that many things depend on for five years and now I change the code are you scared?
Jacob Luetzow: Yeah, terrified.
Jason Cohen: Right. But if if if like I changed it yesterday and the day before three times and I change it, it's like, yeah, that's just under development. It's not scary. We probably don't even need to look at it for a while. So that's useful information that the that the version control system has if you could see it. So I had put visualizations into things like Visual Studio. and nobody cared. I mean, I say nobody, a couple of people bought it, but wasn't that interesting. But what people did is they abused it to do code
review. So I'd get these feature requests like, Hey, can I can you print out these diffs and have someone write on it and then give it back to me? what are you talking about? So it turned out people wanted to do code reviews. They wanted to show people diffs, get comments, and and talk about the code. And so that I ended up building that product, and that was a success. And again, Git doesn't exist, GitHub doesn't exist. Current now we all code review things and code reprs. That was not the case then. There was no tools for that. So
Jacob Luetzow: Yeah.
Jason Cohen: I built this tool that was the most successful tool for that by far. We had these huge case studies like at Cisco with 2,500 code reviews. So we knew with data, what's efficient and not efficient ways to do code reviews, when to do them, when not to do them. So I wrote a book about that as well with all this data and stuff. So we were certainly the world experts in code review and had this tool. So are my customers for things like Microsoft, Intuit, Qualcomm, EA Sports, this kind of stuff. And so that was now again, nowadays you're like, wait, that doesn't make any sense. I know I'm old. What do you want me? What do you want from me? But that's what happened.
Jacob Luetzow: How did you how did you do how did you do code review? Was that like a manual process with engineers or did you actually write code to
Jason Cohen: No, we didn't do it. We made a tool so they could do it.
Jacob Luetzow: I see. Okay. Okay. I thought you were they were paying to get their code reviewed. Yeah, yeah. Okay.
Jason Cohen: No, that's a consulting company and also it's weird. I think that does work when it's a highly targeted expertise that you want the review of. A security
Jacob Luetzow: Mm-hmm.
Jason Cohen: review makes sense. a governance review of some s of some specific governance sort makes sense. Umadays you could say, well, how much of that could be AI? you know, et cetera, et cetera. Yeah, I know. But anyway, th those kind of reviews can make sense. I mean, with an external person.
Jacob Luetzow: Right.
Jason Cohen: maybe a database type of thing of review our database architecture and queries and indexes and give us some advice. I think that that can make sense. So there's th deep expertise with a specific thing. Then an external person could could actually be quite helpful. but the daily stuff that doesn't make sense to me.
Jacob Luetzow: Mm-hmm. I'm curious, what are your thoughts around code review now? How have things shifted with AI? you're starting to see the trends of like, in my opinion, I've seen some very smart people claiming that they don't review code anymore, that they they trust their agent loops and yeah.
Jason Cohen: That's because they don't. Yeah, they don't. And and you get what you pay for, So I think the the usual argument there goes like this. the bottleneck used to be writing code,
Jacob Luetzow: Mm-hmm.
Jason Cohen: and now the bottleneck is reviewing code because AI writes it faster than we can review it. And so human beings should review code. Now, sometimes the argument stops there. And then I would say, true, human beings can review the code. And if you do, you will see why you need to.
Jacob Luetzow: Yes.
Jason Cohen: Because people saying that I don't need to look at it, it's like, well, if you look at it, you'll realize that either you're not a very good developer and so you think that's fine. Or you're a senior developer and you're like, but I just don't want to do it, which I understand, but that doesn't mean it's okay. It's
Jacob Luetzow: Yeah. Yeah.
Jason Cohen: it's not good. What happens, right? Now, if you're vibe coding something for your personal, you know, personal management stuff, who cares? You don't need the code to be good. Or throwaway
Jacob Luetzow: Right. I agree.
Jason Cohen: code. Throwaway code doesn't need to be good. Don't, don't review it. Just do it, you know, like not everything needs it. But if you have production code that's that matters and then it does matter if if the code's not good and or poorly architected or wrong in certain ways and so on. So on the other hand, what do you do? Take all your best talent and make them review code all day? Cause I would quit that job. So that's so
Jacob Luetzow: That's what jobs have turned into.
Jason Cohen: so so that's a that's a that's a that's a bad thing. So
Jacob Luetzow: Yeah.
Jason Cohen: then you move on and you say, well, I should have agents do that. that code review for both of those reasons, both because it's still a human it is a bottleneck, it's still a human bottleneck, and humans
Jacob Luetzow: Mm-hmm.
Jason Cohen: don't want to do it anyways. So we should have agents do that again. And maybe I'll do something like, well, the agent will be specially tasked, maybe I'll use a different vendor, maybe it'll be a panel of experts. there's all kinds of ways you could try to say the reviews will be better. And of course, the reviews are better than nothing and probably can scale and probably can help remove that bottleneck. So then what happens is I think there's a secret bottleneck of quality. So in other words, because no one's looking at it, I'm telling you the quality's not very good. Now you could say, yes, but we'll notice
Jacob Luetzow: Yeah, probably.
Jason Cohen: it, we'll tell it to fix it. Yeah, maybe. I yeah, maybe that maybe that is the answer, but I think there is this hidden
Jacob Luetzow: Yeah.
Jason Cohen: bottleneck of quality. Anytime I look at it, maybe I can find the quality issues. But when I don't look at it, I don't find them or I find them late or whatever. And that's obvious with things like security issues. We know there's that AI is constantly finding security issues. We know that there's a quality there's quality gaps, especially in vibe coded stuff, it's a it's a disaster, right? So and again, that could be fine depending on what your vi what you're doing with the vi cutting cut, right? But if
it's not fine
Jacob Luetzow: Yeah, yeah. Well, for example, I just vibe coded, I reverse engineered the APK app to my mesh router system because it was so buggy. And
Jason Cohen: Mm-hmm. Mm-hmm.
Jacob Luetzow: I have this local host monitoring system. I didn't read any of the code, but it works and it's amazing. And I use it every day. Parental
Jason Cohen: Yeah.
Jacob Luetzow: controls on all the kids, you know?
Jason Cohen: Yeah. So so now you could also argue, wait, it's not like human beings were perfect to quality and security in the first place. If they were, AI
Jacob Luetzow: I was just gonna say,
Jason Cohen: wouldn't be finding a extra security holes in all the software that we've been using if if we were good. So so just saying a human being should review it is again, maybe not even the answer again. and so and so what is the answer? I think we're figuring it out. So I'm obviously I don't have some magic answer that's that's a perfect way to do coding at scale with high quality. you certainly are making a trade-off of cost to quality, right? If you want to have code agents review it and in all these different ways and so on, there's certainly a cost. Now, I I would suppose that the total cost is way cheaper than the number of human beings needed for a similar quality bar. But I do think it'd be it's interesting to ask stuff like, what do you mean by quality? How would we know? And of course these things can't just be one number. And maybe they're not even
Jacob Luetzow: Mm-hmm.
Jason Cohen: objective. I agree with all that. Nevertheless, we should try a bit because when we can make it a bit more objective and a bit more like numbers, then we can make systems that are controlled to that. And that would be good. So again, I don't want to overindex on things like metrics and these things because I I pre-agree that the world is more subtle than that. But often we have none. Often I see don't you see people that have nothing at all? Like you ask what's the quality? They're like, I don't know, it seems fine, and that's all they can say. Well, that surely
Jacob Luetzow: Yeah.
Jason Cohen: isn't good enough. So surely there's something better than that we could say. So for example, one of the things I like to do is force all the agents to be TD TDD and 100% code coverage. So they always write the test first that fails, then and on and 100% coverage, which means mock objects and stuff. Another thing that means
Jacob Luetzow: Mm-hmm.
Jason Cohen: is you have to you can't just take any code and work that way. You can't take any code and have 100% coverage because you need mocks and all kinds and other, you know, you need to just architect
Jacob Luetzow: Right.
Jason Cohen: the code so that that's even possible. So by forcing it to do it. You're also forcing the design to be better and architected better in the first place. That's mockable. That probably means it's more modular. That probably means it's more observable and loggable. That probably means you could switch things out later if you needed to. and so actually the code's probably architected better, and that's going to help the quality in the first place. And even if the quality is bad or you know, something needs to change, you're kind of set up for a little better than you would have been if you hadn't said any of that. So I do think so that's an example of what I do, but I'm not claiming I have all the answers. I'm just telling you, here's an idea that I do. It's okay. It's not like solving the whole problem. But there are things we could do. And I think we I think just saying I'll have another agent review it is lazy and avoiding the problem. Whereas
Jacob Luetzow: It is.
Jason Cohen: what I'm saying is at least trying to tackle it, acknowledging the faults it still has, and just trying to say, so what could we do?
Jacob Luetzow: I think the the good part about test driven development too is you're forcing yourself to think through the problem and then you're solving it right. And then you also all you like out of the box you have test coverage. And I think it's wild to not take advantage of using an LLM to write your tests now. It's almost it's silly to me. But I've also now seen the argument floating around on X where people aren't writing tests anymore that, you know, problems will reveal themselves at some point. And I I I don't know. I I agree.
Jason Cohen: No, that's stupid because no, that's it's dumb because even the LLMs do way better with tests because they'll make
Jacob Luetzow: Yes.
Jason Cohen: a change and then the test will fail and it will then know objectively whatever it did was wrong.
Jacob Luetzow: It iterates on it, yeah.
Jason Cohen: And so it's a check on itself. And so if you're not even gonna do that, again, see here I think now you're just not now you're not being sensible about what these systems are.
Jacob Luetzow: And it's like senior engineers saying stuff like this. And I don't know if it's just just throwing in the towel, giving up, don't want to use your brain anymore. But I like personally, LLMs have just unlocked me to do a lot more by myself. And I love that. I still am engineering. N am I writing code? No, but I'm like the orchestrator of code and I'm still reading it all. And
Jason Cohen: Yeah, that that's none of that's none of that's in and that none that's under debate, right? Of course. So
Jacob Luetzow: Right.
Jason Cohen: the question that what I'm saying is I think that we have a bottleneck on quality. We try to solve that with human
Jacob Luetzow: Yes.
Jason Cohen: code review. It's obvious why that's not a great answer, but it's an answer. And sometimes it's appropriate. And then you could say, well, I'm gonna code review with other agents. And here's where I'm saying, that's probably is better, but I still think we're we're not fully like there's more to say, like I just did. I just gave examples of
Jacob Luetzow: Yeah.
Jason Cohen: things to do. So I think I think acknowledging
that this exists. And saying what are some practical things, because I'm still an optimist. It's not like so we shouldn't do it. I'm not saying that, But what should we
Jacob Luetzow: I agree.
Jason Cohen: do? What should we add in? Another funny thing is we keep we add it, all of the even the stuff I'm saying are things that human beings have figured out to do that we could impose on the LLMs. But it's not, there's no reason to believe that whatever we've thought of before is the best way to do something like have quality at scale, quality in the way that I've been defining it. Maybe there's other ways. So That wouldn't work with a human, but would work with LLMs. So maybe we're not really thinking about that even broadly enough. I don't know how to think that way. maybe I'm not smart enough for that. But I think that's an interesting question. Is like, so at the minimum, we should take the best practices that we do know and make it do that. shouldn't that be the minimum? But that's probably
Jacob Luetzow: Mm-hmm.
Jason Cohen: not enough or right. So, for example, a lot of people have these things like okay, I built a you use paperclip or something to build a An org chart, like a personal org chart. And you have your head chief of staff and you have your, different departments and things, and and and and you they organize work and everything. That's fine. But what that does is take this human organization, which by the way, those same people hate. people who build that, do you think they like being part of an organization? Do you think they like being, a line manager and or
Jacob Luetzow: No, it's the worst part.
Jason Cohen: they hate that. Okay, but nevertheless, they've recreated it and said, here's a good way to organize work, right? To send work in, to do work, to report back, to have goals. This is a good way. Is it? Because last year you were telling me how you hate the work because it's so inefficient and dumb. So you're taking an organization that's designed for human beings with their foibles to work together, which LLMs have different foibles. And anyway, you hated it, even for that purpose, you hated it. And we're like, we're like, let's make LLMs do that. It's like, well, that's wrong on multiple fronts. What is the right way for LLMs to do work? Well, At least we can s understand what the LMs are doing and I think that's why we do it. It's all we know. But again, there's no reason to believe a human an org chart is the right way for LMs to organize work.
Jacob Luetzow: Yeah, no, that's a fair point. I feel like we're we're figuring it all out right now and we're trying to make it work in ways we understand.
Jason Cohen: Sure. Right.
Jacob Luetzow: We're just forcing forcing something. Yeah, I don't know what it means. but I absolutely agree with T D D for for writing code and at least having some sort of structure to your process.
Jason Cohen: Now, w back in back in my smart bear days when I was teaching people how to do code review with data to prove it, I had a I always had this nice thing I thought was good and especially a good way to get people introduced to it if they didn't do code review before and they didn't see the value in it. Is I said, why don't you just code review the unit tests, not the code? And the reason is if the unit tests are good, and we can define good, right? They're they're relatively complete. They do the corner cases, they're not over specified so that we can't refactor code in a reasonable way. If the unit tests are good, whatever that means, in a funny way, that what I'm about to say is not true exactly, but it doesn't matter what the code is. Now that's not quite true, but it kind of doesn't matter
Jacob Luetzow: Yeah. But no, I get what you're saying though.
Jason Cohen: because you can fix it safely. Like whatever's wrong with it, like it's too slow, it does a stupid algorithm. Okay, but now it's easy to fix because we had good tests.
Jacob Luetzow: Mm-hmm.
Jason Cohen: Whereas if we didn't have good tests, it's not e c clear, it's not safe and easy to fix, because you fix it and then you're scared. And you're like, Ugh, ugh, I hope this doesn't
Jacob Luetzow: Or if you blindly only focus on code coverage, but then you actually look at the tests and they're just written to pass, code quality matters. Yeah. Yeah.
Jason Cohen: Right. Right. So if you knew that the code coverage was 100% and you review just the tests, you could conclude the code is pretty good. good enough, easy to fix later if it isn't. So that would be a way to do less code review, but but be thinking about quality and maintainability and and so on. That's not a bad trade-off. And so again, that doesn't mean okay, well, that's the rule for everything. I'm just throwing it out there as these are these are ways to ask the correct question, which is the following. When is it a you good use of our time to review code? That is the question. The answer could be never. Okay. Probably it's not never and probably it's not always, right? So the better we can answer that, we are using our time wisely. So
Jacob Luetzow: Yeah.
Jason Cohen: reviewing just unit tests, that's again probably too simplistic, but it's maybe pointing at some other ideas for when what is the answer to that question? And again, for vibe-coded stuff, the answer is definitely never. Like don't don't worry about it. But for stuff that's in production, we have to care. We probably need a more sophisticated view about what is the best use of our time.
Jacob Luetzow: Yeah, a hundred percent. I do I do think that focusing on tests and whenever I'm doing PR reviews, I always focus on the tests because it usually points to a larger problem pretty quickly if you find edge cases or gaps in the coverage. Yeah, no, that's cool. all right. I think that's enough about tests, but hit so hidden multipliers, what made you want to write this book?
Jason Cohen: Well, I've been writing for 20 years, so that's what I do is I
write things.
Jacob Luetzow: Yeah.
Jason Cohen: usually they're articles. I left WPN Gen in May of twenty twenty-four after a mere 15 years. So yeah. Yeah, the overnight success and I I
Jacob Luetzow: Yeah, overnight success.
Jason Cohen: started and got the hell out of there as soon as I could. and and I you know, I was decompressing for a while. And one of the I was asking myself is like, geez, I've been writing for so long on all kinds of topics and startups and and companies, and not actually engineering and writing tests, by the way, not that. And I was I was I almost felt like what do I even think? Like, does this add up to anything, or is it just a whole bunch of random thoughts? Like, is there is there any concepts in here? And so there's a couple of themes that I thought, ooh, that that that's pretty interesting. And so one of the what one of the things I was hearing when as I talked to founders, which I do all the time in conferences and online, of course, and at at and in in investing, which I don't do very much of, but still people want to talk about it. And I'm happy to talk because it's fun and I like to help. And what people, everyone was saying as AI was coming up was growth is slowing. I don't know why. And often with a real company, right? They're at 100K a year, a million a year, two two two million a year, 10, 20, and is everything's been fine and then grow growth has slowed or even stopped or reversed and they're like, What what what's going on? And of course you could say AI, but there's plenty of companies w that are growing also. So that's not the that's not the answer. Especially when I started writing this book in twenty twenty four, the answer was not at that time AI. so I realized I had so many thoughts on this. And so some of the things I've written about in the past are in the book, and of course there's a lot of new material too, but it's answering this thing. How do I grow with my current team and budget by doing something smarter than I'm doing now. Ideally, something where if I make a relatively minor change, it might be hard, but minor, could it have a big impact? And so you could say 80 20. It's like, well, y yes, of course, of course, if I'm gonna restart growth with my existing budget and team, it has to be something where I have some leverage, yes, but that's that's that's sort of not the end of the story, is it? It it needs to exploit some kind of mechanistic truth about how my business works that I can understand so that I can exploit it and and and iterate on it. So I need So I need to understand what those mechanisms are that might apply to me so I can and see what other people have done to exploit it so I can maybe do that. So that's what the book is. As you know, every chapter is one of these hidden multipliers.
Jacob Luetzow: Yeah.
Jason Cohen: and so of course the idea is that maybe some of it applies to any given person and it's useful. the feedback on it so far is yes, that's what it does. It gives
me lots of ideas.
Jacob Luetzow: Yeah.
Jason Cohen: I try lots of things. Then the other thing is there's skills, which are free and open, they're on GitHub. You can go to skills. Dot hidden multipliers.com. You don't even have to put in an email address. They're just freaking there, right? And the skills facilitate everything in the book and even some of my articles too. So it's all these things where the AI does not give you answers because I don't believe that's a good idea. AI can't set your strategy or tell you who your ideal customer is or tell you who you are or or you know, this is not things AI can do. But I do think AI it can it can't do it. And so what it can
Jacob Luetzow: Well, the second you say you don't agree with it, it goes, You're absolutely right. Yeah.
Jason Cohen: do is facilitate. So, if you say, My product is a CRM for small business, it can say, small business is not an I an ICP. That's like 800 markets j unioned together. You need to get more focus.
Jacob Luetzow: Mm-hmm.
Jason Cohen: What is a good and bad? You know, and it can force you to be more specific, go through the things that I say in the book. It can force you through some of that stuff and kind of hold your feet to the fire. In that way, maybe it helps you have better ideas and analysis and and what to do next and so forth in that sense. It's still Can't be right all the time, right? But if it can force you to be more articulate, go through certain processes that I think are useful, okay, maybe that's helpful. Probably not as good as a human being doing it, but we don't have human beings doing it. So this is better than sitting alone
Jacob Luetzow: Yeah. Yeah.
Jason Cohen: trying to come up with your ideal customer by yourself, even with a book with chapters in it.
Jacob Luetzow: I have my team of grokbots these days, And
Jason Cohen: Right.
Jacob Luetzow: it times have sh have shifted quite a bit. so I'm like mostly through the book. I have a couple chapters left, and I've been really enjoying it. one of my like little snippets that it's one of these books I told you before we went live that like I'm gonna reread this thing multiple times, I think, just to pull out the little nuggets of information because you're always just focused on your current problem. So sometimes
Jason Cohen: Yes.
Jacob Luetzow: You don't hear all the other solutions because you're
Jason Cohen: Uh-huh.
Jacob Luetzow: so like one of my issues was I kept having trials sign up right on Kin Switch and they weren't using the app during their trial. So I was like, okay. And then your onboarding stuff, you have to have a really good onboarding flow. And I went through and like added a whole onboarding flow to my web app. And now hopefully, like I literally just did that like a couple days ago. So I hope I start to see. people actually start using the app in their trial. So because when people use it, th they don't they don't drop off. They usually, you know, continue on after my trial time.
So and and it's like just little gold nuggets like that that you really appreciate as you're going through things.
Jason Cohen: Yeah, and and what people sometimes don't realize is how outsized the effect is on the onboarding end.
Jacob Luetzow: Mm.
Jason Cohen: and I give an example, I think this is in the book. Well, anyway, it doesn't matter. I give an I show my YouTube on YouTube it shows you how many people are still watching after how much time. YouTube
Jacob Luetzow: Yeah, yeah.
Jason Cohen: people know this, other other people may not know. So of course there's this massive drop-off after 30 seconds or a minute, and then it just like the slow curve as people
Jacob Luetzow: I know it's horrible. It's sad.
Jason Cohen: drop off later. So I sh so if if you just look at that curve, which is also true of software a lot of times, right? you have huge cancellation in the first month or even day, or they just never activated, so it's effectively now. But then you're then it levels off into whatever your long-term cancellation rate is. That's so this curve is pretty common in in many things. In YouTube, it's nice because it's there's a lot of data and you can see it right away.
Jacob Luetzow: Mm-hmm.
Jason Cohen: so what I point out is if I if I go out to this tail like around an hour, I'm like my podcasts are long, they're like one to two hours. If you go out to like 45 minutes, By the time you get there, you're probably listening to the rest, right? So if I optimize right there for people staying, which is optimizing for your long-term cancellation rate, I could pull that curve up a little, but at that point almost everyone's gone anyway. So while I could pull it up a little, they were pretty sticky. Pulling up a little, it just doesn't affect how many people not in any material way that will it affect what happens next. I I it's already over in a certain sense. So on the other hand, if I go to the front where 50% leave in the first 30 seconds. Even a boost of 10%, getting it from 50 to 55%, that's
Jacob Luetzow: Mm-hmm.
Jason Cohen: enormous because that 5% might flow to the end where eight people, 8% are left. Maybe I could go from 8 to 10 or 12, but that's a 50% boost. So a 10% boost where everyone's leaving could be a 50 or 100% boost at the end. And in software,
Jacob Luetzow: Mm-hmm.
Jason Cohen: this is everything. In software, the only way you make money is by people staying for years and paying and paying. So
Jacob Luetzow: Right.
Jason Cohen: a little boost in onboarding could result in. double the number of people that last for three years. Once you understand
Jacob Luetzow: Yeah, that's what
Jason Cohen: that math, once you because this is math now, I haven't said anything about the product and the industry. It's math. If this is a number on here, if you shift it, then the number over there shifts by a lot relative to what it was. So this this is, and this is what I mean when I say the hidden multiplier has to be mechanistic. It has
to be like, what is the reason why this generally happens and I can rely on it and therefore invest in it.
Jacob Luetzow: Mm-hmm.
Jason Cohen: And you know, so onboarding is like this. A lot of people leave, but you already spent all the time and money to get them here. You had ads or you made v videos and you talked to them over tech support, perhaps. All this stuff happened and they just leave. Like you incurred all the costs
Jacob Luetzow: I know. It's
Jason Cohen: and got none of the benefit. If you can move that even by a little, like 10%, that could have a massive actual impact on the revenue and profit over the next year.
Jacob Luetzow: Yeah, that's huge.
Jason Cohen: so so so it's worth extra effort. To optimize the onboarding. you just did that. Great. You shouldn't stop doing that for
Jacob Luetzow: Right.
Jason Cohen: six months or a year. It's more important to do that than almost anything else.
Jacob Luetzow: Well, what's wild is before I listen to your book, which by the way, I love that you read it. I I think that that's a huge bonus. but just listening to it, I never thought about that. It was one of those things I was like, yeah, that's a nice extra, but do I really need it? And it completely makes sense because just looking at my data, which I don't have a huge user base, I'm growing, but you know, every month is more users than the month before. So I'm happy. But just watching how many trials weren't using. the product. It's like, well, why are you even signing up? And I like, so I think it makes it makes a lot of sense. Coach them through your product, teach them how to use it and make it a pleasant experience.
Jason Cohen: Yeah, so the men the mindset that I talk about is something called the happy path, which is a general concept been used a long time. We also use it at WP Engine, so I'm used to the words. But the idea is when a customer is successful, what does that look like? And you can usually map it out pretty well. Well, they sign up and then they add this data or they input this thing or they they confirm their email address within an hour. there's all these little and
Jacob Luetzow: Mm-hmm.
Jason Cohen: then they upload this data or they connect this service within another 15 minutes. And there's this thing that happens when a cusp when a person is successful. And obviously I'm I'm simplifying because people are different, but you there's there's certain patterns that you see. They're like, yeah, this is what it looks like when someone's successful. So you call that the happy path, and you measure it. Every one of these things I'm talking about, you log, you have it logged somewhere so that you can measure the times in real time and all this kind of stuff. With AI, you can do this, you can set all this up. So So you set up this happy path and then you say, okay, so whenever someone signs up, I always got to know where they are along the happy path. And at some point, most of them will fall off the happy path, maybe in the very first step, maybe on step four, At some point they fall off. And so you do something. What is that? I don't know. It could be you send them an email, it could be you try to chat with them. It could be, you know, it depends on the product, you know. And then if they don't answer that, then what? And so you can imagine this tree of stuff that you do
Jacob Luetzow: Yeah.
Jason Cohen: so that wherever they are. At that moment when they fall off, and also the context of what else have they done and stuff, what can you do to revive them? In the old days, this would just be very basic decision trees with simple measures. Nowadays with AI, it could be completely personalized with their name and where they came from. And you could use their email address to find their company. And so when you write that email to revive them, I mean, it could be amazingly specific, which is probably more effective. You could know what time zone they're in potentially and and then send them an email when it's their morning before they get into their day. Like with AI, you could probably get really, really good at this compared to what we used to be able to do on the happy path. But the idea is the same. There's a happy path when they fall off the happy path. Do something, which means you have to be measuring
Jacob Luetzow: Yeah.
Jason Cohen: it. So you can be doing automatically and all this stuff. With AI, you can absolutely build a happy path tool. Now, if it weren't true that all that was so impactful to the business, you might say it's not worth all that time, right? But my contention is that it's so unbelievably valuable to the business. There's so much leverage in each person that you get through the happy path. That in fact it is worth everything I just said. And so you could do another three features and still almost everyone that sign up for a trial with not even try product. If they're not even trying the product, yet they don't even know the features are there. So I don't
Jacob Luetzow: They don't even know the features are there. Yeah.
Jason Cohen: believe you that one more feature will help because they've never even looked. So
Jacob Luetzow: I think this is the toxic behavior that software engineers do to themselves, right? you just procrastinate. You don't want to deal with the truth and the hard part of onboarding users and marketing. So you distract yourself with being with busy work essentially. Like of course you can write a new feature. You're a software developer.
Jason Cohen: But it's more than busy work. It's fun and probably the reason you're doing this at all. So
Jacob Luetzow: For sure. I agree.
Jason Cohen: it's fulfilling. It's fulfilling and enjoyable
Jacob Luetzow: Yeah.
Jason Cohen: to ignore building a business and instead make your software. So I get it. And if you just want
Jacob Luetzow: Yeah.
Jason Cohen: to make a project, then just make a project on GitHub. Don't worry about it. But if you state that your goal is to have
customers paying you for stuff, I said, okay, well, then you're not making a project for you. So you stop acting like you're making
Jacob Luetzow: Right.
Jason Cohen: a project for you.
Jacob Luetzow: And you kinda have to step out of your engineering hat a lot more than you want to too. Right.
Jason Cohen: No not kind of. A business has many hats. If you're the only person there, they're all on your head. Or at least you're
Jacob Luetzow: Yeah.
Jason Cohen: swapping them out during the day. And if you're like, no, I like my little engineering cap, it's like, okay, then make a project. It's it's totally fine.
Jacob Luetzow: Sometimes you just want to go in your little engineering cave and be like, I don't like just build,
Jason Cohen: Yes, it's totally fine. What you can do is say, Well, I'll engineerify these things. And I would say that half works.
Jacob Luetzow: Yeah.
Jason Cohen: you could engineer up this happy path stuff. Good. Isn't that writing code and fun and AI? It is. Well, then fulfill
Jacob Luetzow: Mm-hmm.
Jason Cohen: yourself with that because that's actually going to help. But ultimately
Jacob Luetzow: Yeah.
Jason Cohen: you have to know what customers actually want and actually want to buy and what they want to see in that happy path and why did they come up in the first place? And AI is not going to tell you that. And writing code will not tell you that. You have to find that on your own. So I do think you can do some of this. Some of this is I could apply my engineering hat to or my engineering behavior to some of these things. Okay, halfway true.
Jacob Luetzow: Mm-hmm.
Jason Cohen: But when you abdogate responsibility for saying what is what does someone want to buy and who is that someone and where would I find them? Okay, then they're not then then you will lose.
Jacob Luetzow: Yeah. The thing that has helped me, and I never used to do this because I I hated it. Well, it's just like customer outreach, right? I actually really enjoy talking to the customers that cancel my trial and they are like kind, you know, just feeling the waters, but they cancel immediately. and I've been reaching out and I use AI, like a my my GrockBots monitor like my stripe activity and things, and they'll be like, Hey, this user. Signed up yesterday and canceled an hour later. and then I'll open up an email conversation with them. And honestly, I've gotten some of the best product feedback from those people. And I've changed a lot of things about my product because of the people that don't want to continue using it. And now it's like the most beneficial because you have happy customers, they'll never give you bad feedback if they really like your product.
Jason Cohen: Right. And of course there's a there's a there's a paragraph or there's a maybe a page about this in the book, right? Because there's a the whole chap two chapters about cancellation and what to do. because it's such such again a financial impact to the business. And also it means people didn't want it. Don't you want people to want it? So it's you know
Jacob Luetzow: Mm-hmm. Ideally, yeah.
Jason Cohen: so yeah, cancellations are are good if you're talking to your ICP, if you're talking to the customers that you want to have. The danger is that you talk to everyone. Indiscriminately and take all that feedback. But if you built the product
Jacob Luetzow: That's hard.
Jason Cohen: for person A and person B signs up anyway and cancels, which they absolutely do,
Jacob Luetzow: Mm-hmm.
Jason Cohen: no matter what you do, that's going to happen. And then you listen to B's feedback and do it, then you're making the product worse for your customers. And you're violating
Jacob Luetzow: Yeah. I actually
Jason Cohen: your own strategy. And so step question one is: do you know? Who your target customer is, and the answer is usually no. So that's why there's other chapters in the book about that and skills. So you can just use the skills and not
Jacob Luetzow: I was just gonna say that's the hardest part. Yeah. Yeah.
Jason Cohen: buy the book. That's fine. Just do that, right? And so that's why it's so important to know. I mean, one of many reasons why it's so important to know who it is, who that is. Because not only is all your marketing going to be for the ideal customer and the homepage is, and the pricing has to match what they want, and the features have to be what they want. And then also when you get feedback, whether it's from cancellations or not, the only feedback you can listen to is from those ideal customers. You have to ignore the rest.
Jacob Luetzow: Mm-hmm. Well, and I think that's the hardest part because when you build a product, at least this is my perspective. I'm like, well, it's for everyone, but you can't market to everyone, can you? And
Jason Cohen: Well forget marketing. You can't build a product that's good for everybody.
Jacob Luetzow: Correct. Yeah. You're always gonna alienate someone or someone won't like something. And like
Jason Cohen: Unless it's total sh unless everyone hates it.
Jacob Luetzow: That's true. what is that? Jira? I don't know. Yeah. That data people love Jira, right? Yeah.
Jason Cohen: No, some people love Jira. That's the point. Man, some a lot of managers love Jira because they can they they have lots of fields. They have lots of fields and they think that that gives them a lot of control and visibility. And maybe it does somewhat. It also has other side effects. And then
Jacob Luetzow: Yeah. Yeah.
Jason Cohen: other people love linear, and then managers don't like linear because they can't get any of that information out of it. So yeah,
Jacob Luetzow: Yeah.
Jason Cohen: this it's good examples of how you're right. You cannot build for everyone even if you try. You cannot. And and and
Jacob Luetzow: No. Yeah, yeah, you can't.
Jason Cohen: you should not.
Jacob Luetzow: So my question for you though is okay, like in your book it's Carol, right? Carol's your ideal customer.
Jason Cohen:
Yeah, I named that her named the ideal person Carol, right? Because I did
Jacob Luetzow: I like it.
Jason Cohen: that fifteen years ago, so I kept that. But it's really great
Jacob Luetzow: That's funny. But
Jason Cohen: now because of Pluribus. So now people people have made little videos and stuff. It's like, we we just want you to be happy, Carol.
Jacob Luetzow: That's hilarious.
Jason Cohen: And it's like, my god, it's perfect. But I made this up fifteen years
Jacob Luetzow: That's hilarious.
Jason Cohen: ago, so so it's just a fun coincidence.
Jacob Luetzow: Yeah, that is funny. how how do you go about finding your Carol? Like I like, for example, we can take any product. It doesn't really matter. I think it's it's just like it's very difficult. Like Kin switch, for example, is, you know, it's I'm kind of aiming it at it's a dead man switch associated to your digital files, but for family legacy, like I can pass it on to beneficiaries
Jason Cohen: Mm.
Jacob Luetzow: if I stop checking in.
Jason Cohen: I need that I'm I I'm doing that right now with one pass, yeah.
Jacob Luetzow: Yeah. So I think it's when I look at this product, it's anyone with a family. You need this product, but you you obviously can't think that way. Yeah.
Jason Cohen: No, no, it's not. No, it isn't. Because everyone with a family doesn't want to buy it. See, notice I've always said that. It's not
Jacob Luetzow: Yeah, yeah.
Jason Cohen: do people want it, and it's not if they have the problem. It's also that.
Jacob Luetzow: Yeah, yeah.
Jason Cohen: It's what do they want to buy it? So I'll answer like let's let's let's take yours as an example. Now everything I'm saying in here, yes, it's in the book, but it's also in those free skills. So you don't have to buy the book. And anyway, I'm gonna tell you it right now. So it's no worries. Okay. So, okay, so first of all, since you have an existing product, we can also use it. The process is similar if you don't, but you don't have information. So you have to fill
Jacob Luetzow: Mm-hmm.
Jason Cohen: in in certain ways. But okay, let's just stick with yours and try to keep it simpler. So, so first of all, let's go to what I'm gonna skip ahead a little bit. You're right that everyone theoretically has the problem, has the pain, whatever. Now, in any given day. What normal people, and let's even just r l'et's even just laser focus on the United States. There's no reason to do that other than we we're both there. So
Jacob Luetzow: Right.
Jason Cohen: most people in the United States do not wake up in the morning and say, today's the day I'm gonna buy a kill switch software for my passwords so that my k kids can access my bank accounts. That is not what people wake up in the morning and think on most days, right? Okay.
Jacob Luetzow: That's true. Yeah.
Jason Cohen: However, at any given moment. There are people who think exactly that. I'm going to buy a solution for this today.
Jacob Luetzow: Mm-hmm.
Jason Cohen: So why is that? So I call this an inciting event. Something has happened that's causing someone to say, today's the day I'm buying one of these things. Maybe it's yours and maybe it's a competitor's. Fine. Maybe it's
Jacob Luetzow: Mm-hmm. Yeah.
Jason Cohen: some other system that's not any of you. They're going to write it down on a piece of paper and put it in a vault. It's an alternative is what we would say in economics, right? But one way or another, they're gonna do something today. So I call that inciting event. So my question is, what are some inciting events that happen that causes someone to say that? And so what's the answer?
Jacob Luetzow: It could be anything, right?
Jason Cohen: No, what is the answer?
Jacob Luetzow: could be what the inciting event.
Jason Cohen: It's it cannot be anything. We're out of milk. We're out of milk. We're out of milk. So I decided to buy no, it's not anything. What are the kinds of things?
Jacob Luetzow: It could be a death in the family. It could be fair. Some tragic event made them think about their demise or something,
Jason Cohen: Like what?
Jacob Luetzow: one of their friends could have passed away, and a car accident or anything. Anything that puts them thinking about the fragiless of life.
Jason Cohen: Yeah. So that's a start. I like it. You could end there, but you could probably get even more personal than that. Your own brother, parent, somebody super close to you dies, and you can't access their bank account.
Jacob Luetzow: Yeah. Yeah.
Jason Cohen: And so right now you're like, my God, that that's what would happen to my spouse, to my kids, to my brother.
Jacob Luetzow: Mm-hmm.
Jason Cohen: It's not even theoretical and hypothetical. I'm not even vaguely thinking about death. Although I agree, that's a good, that's not a bad answer. But this is a better answer. I'm not vaguely thinking about death. I can't get into my mom's bank account right now.
Jacob Luetzow: Yeah. Yeah.
Jason Cohen: And I realize, my God, I can't do this to my family. That's an even better one. Now,
Jacob Luetzow: Right. Yeah.
Jason Cohen: look how much easier. Now, having even just said that, let's say we wrote that down that's saying that's an inciting event, right? Look how much easier it already is to do things like make ads and keywords. What is the kind of ad and key and keyword would you write for someone in that predicament? How do I get into one password from my mom? Mom died. How do I, right? I don't know because I'm
Jacob Luetzow: Yeah, yeah.
Jason Cohen: not an expert in the area, but I'm guessing you could you could come up with a lot of very specific things of what that person would be typing in. And that
Jacob Luetzow: Mm-hmm.
Jason Cohen: would tell you, what do I need my ads to come up with? If I am doing SEO and you haven't given up on that, what does that need to say? Right? Whereas if you just said, I saw a plane crash on the news, it's still pretty good answer, but not quite as evocative of what I would type in
Jacob
Luetzow: Yeah, fair. You need
Jason Cohen: Google.
Jacob Luetzow: yeah, you need the the pain point of a specific event.
Jason Cohen: Mm-hmm. So again, it's still pretty good. I still could think of ads for that. it's not bad.
Jacob Luetzow: Yeah, yeah. Right.
Jason Cohen: But but notice how when I get super hyper crazy perfect specific, like the perfect case, it suddenly becomes extremely easy. Now, of course, as you said, there are many reasons why someone might decide to search for this. But
Jacob Luetzow: Mm-hmm.
Jason Cohen: we have created a scenario so that it's easier for you to write the ads and keywords because we were specific. So it was worth us getting more specific than the reality actually is, because it made our lives easier. It was so clear what to
Jacob Luetzow: Mm-hmm.
Jason Cohen: do. Then what would you put on the homepage? It's easier to think of what to put on the homepage or a landing page.
Jacob Luetzow: Right.
Jason Cohen: It's much easier to think of things that would make sense and go test than if we just said,
Jacob Luetzow: 'Cause you write your message to that per to Carol instead of to the broad scope.
Jason Cohen: Yes. Yes. Now the truth is a there's a market a hundred times bigger than just Carol.
Jacob Luetzow: Yeah.
Jason Cohen: Who will react to that, type in those things, see the homepage and think, this is great. That's the truth. Good. I know. That's that I know. That's why
Jacob Luetzow: Yeah.
Jason Cohen: this works. Because you think if I'm hyper specific like that, I'll only sign up those people. That's what you're afraid of. But what actually
Jacob Luetzow: Yes, yep.
Jason Cohen: happens is your specificity is compelling. And it's compelling to the whole target market. Your target market, which is not everyone who buys software or anything, but it's your whatever your your actual good market is. By being specific about the bullseye, you get the whole target.
Jacob Luetzow: Mm-hmm.
Jason Cohen: And so you'll get all these people actually sign up. That's what actually happens when you do that. So rather than just saying it's a kill switch for the nobody knows what that even means. that's fun for me. And I actually like a kill switch is
Jacob Luetzow: R right.
Jason Cohen: a really funny term. So I actually really like that for me, but that's not what someone's typing to Google, obviously. So when we're when we're really clear about what Carol is with things like inciting events. It just it just sh it it almost like the right thing to do just falls falls out. Isn't that great? that's part of why it's so good.
Jacob Luetzow: Yeah. Well, y you know what's funny, actually, I I rebranded a few weeks ago now, probably like a month, because I had so many people say they love the name Kill Switch, but they don't want a loved one getting an email from Kill Switch.
Jason Cohen: Mm.
Jacob Luetzow: So I changed it to Kin Switch. And
Jason Cohen: Yeah, yeah, yeah.
Jacob Luetzow: some the people that really loved it were like ups like a little sad, but they understood the shift. And if you do care about SEO, it Does matter a little bit on search ranking, at least for me. And it was a little easier to climb it back up.
Jason Cohen: Now, now just to be clear, inciting event isn't the only thing that defines Carol. I just went deep there to show how useful it is.
Jacob Luetzow: Right. No, that makes sense.
Jason Cohen: So other things, and again, it's in the skills, it's in the book, other things that are part of what Carol is. Here's actually the the the the thing I walk through. First, you ask, what are the strengths and weaknesses of my product and company today? This is again because you have a product, because there's already ways in which you're better than the competition or worse, or
Jacob Luetzow: Right.
Jason Cohen: better than alternative or worse. And some of that might be features and stuff, but it also can be personal shit. I love sales, so I don't mind going to something like a an estate lawyer and getting them to resell this for me. I like that. Ooh, that's a whole different track. Then the the the the the customers now, an estate attorney, not even the customer. That's a whole nother strategy. Cool one.
Jacob Luetzow: Mm-hmm.
Jason Cohen: But you might say, no, I hate that. I want it to be PLG and I love people and I hate lawyers. And so, no, this is to avoid the lawyers. I don't know, you I'm just making this up, right?
Jacob Luetzow: Yeah, yeah.
Jason Cohen: And so, okay, if that's your attitude, then that would be a then what I just said would be a horrible strategy for you. Okay,
Jacob Luetzow: Yeah, yeah.
Jason Cohen: so you have to plumb when I say strengths and weaknesses, I'm talking about you personally, which is a whole universe. That's chapter 11, by the way, in the book. And and skills. You're almost there.
Jacob Luetzow: I'm almost there. I'm getting there.
Jason Cohen: And skills too. and I have some articles online too that are that are free to talk about this. Anyway, so it's me personally, strengths and weaknesses that I need to apply to my strategy or who my Carol should be that I want to work with.
Jacob Luetzow: Mm-hmm.
Jason Cohen: And then. I ask, what is a kind of it for that for what what is going on? Well, so the inciting event is what happened that caused them to say now I'm buying. But actually, before that, there's a question of like who has this problem, not a little bit, but like a lot. Like,
Jacob Luetzow: Mm-hmm.
Jason Cohen: so for example, I would say, if I have basically no assets, I'm not thinking about this, and honestly, who cares? On the other hand, If I have lots of assets, I have a family office, I have wealth managers, I have things in trusts. There's a lot. And actually,
Jacob Luetzow: Mm-hmm.
Jason Cohen: I need to avoid probate. That's part of why I have trusts. And there's people, there's a lot going on. At that point, I'm already a person who has se said, I care about my estate and about what
happens next. I am spending
Jacob Luetzow: Mm-hmm.
Jason Cohen: hundreds of thousands of dollars, maybe even annually, on this problem already. So I am a person who really invests in this problem.
Jacob Luetzow: Mm-hmm.
Jason Cohen: So that's very different. And to me, that's a carol, obviously. To me, that's an obvious carol. Okay.
Jacob Luetzow: Yeah, yeah.
Jason Cohen: So it doesn't have to be the only thing. You could have other things too. Well, here's another way people do it. That's fine. But this is a carol that you could, you know, right? And so there's many fewer of them, but they are they don't care about spending money. They'll spend whatever
Jacob Luetzow: Right.
Jason Cohen: this, right? Like so, and so and you it's easy to find them, like, like it's you know, there's As usual, pluses and vices. But so there's these things of like who really, really exemplifies this problem and spends money on this problem, you know, these kinds of things. Who is that?
Jacob Luetzow: Yeah.
Jason Cohen: That's this question of who's right. Then there's also inciting events. Even so, they don't necessarily think today's the day, right? Even so, something right.
Jacob Luetzow: Everyone wants to not think about their demise.
Jason Cohen: So it could be, for example, when someone has come into money and is now setting up their wealth management system practices.
Jacob Luetzow: Yes.
Jason Cohen: Now that's an inciting event to get something like this, right? Or someone in the family died. I think that's a very, very good one. And that's an exciting event.
Jacob Luetzow: Yeah.
Jason Cohen: Okay. So again, there's not just one. it could be just one, but usually there's not just one, right? Okay. So they have
Jacob Luetzow: Right. Or for instance, I just had a son and now I'm thinking about how is he
Jason Cohen: Mm-hmm. Yeah.
Jacob Luetzow: going to know about all the things that I have and does he get access too soon or at the right time? When does he get it? And same with like my wife knowing where passwords are or wherever.
Jason Cohen: Exactly. So it could be a yeah. Yeah. Yeah. It could be a birth. Now, of course, at birth they don't they're not going to get the passwords, but you still might think about it now. So I agree. I agree. That might be still
Jacob Luetzow: Right. Exactly.
Jason Cohen: what you're thinking. But you might realize they are, but they're not really serious because it's not for decades that this kid can even do it. But here's another one. Kid goes to college.
Jacob Luetzow: Yeah. Yeah. Yeah.
Jason Cohen: Now you need to set this up. Right? So it could so ironically, as a proxy, a kid going to college might be a time to advertise about this.
Jacob Luetzow: Yeah, it's not bad.
Jason Cohen: And that gives you all kinds of weird ideas. You could go on podcasts about this, hitting people who are in that stage of their life, because now's the time where the parents are a little older, so they're starting to think about mortality more. Kids are actually
Jacob Luetzow: Mm-hmm.
Jason Cohen: old enough that actually I would give it to the kids. Now I'm
Jacob Luetzow: Right.
Jason Cohen: making this up because I don't know. I'm just This is the kind of brainstorming that's that's that's good brainstorming. That's what I'm saying. I'm not claiming I'm right about
Jacob Luetzow: Mm-hmm. Yeah.
Jason Cohen: all these things, but this is a good brainstorming, yeah, there's like this totally different inciting event that totally right. Okay. So so they
Jacob Luetzow: Yeah.
Jason Cohen: have the pain and and and problem. This is an inciting event. Another thing I have people walk through is there are deal breakers where even though all that other good stuff might be exist, they're not gonna buy.
Jacob Luetzow: Mm-hmm.
Jason Cohen: Now, you may not have any of those, or well, yeah, I don't know whether you have those or not. an example, though, that's true of most software that sells to businesses is that let's say you sell software to small business and an enterprise comes over and says, We need you to be SOC 2 compliant. And you say, Well, I'm not SOC2 compliant. So that means they won't buy because they have a SOC 2 requirement and you don't have it. Now, the product might be absolutely ideal for them.
Jacob Luetzow: Mm-hmm.
Jason Cohen: But since you're not SOC 2, they won't buy. So that's what I mean by a deal breaker. So SOC
Jacob Luetzow: Right.
Jason Cohen: 2 becomes a deal breaker. Now you don't have to list out 50 deal breakers, but there might be some segments of the market, like enterprise, where you're like,
Jacob Luetzow: Yeah.
Jason Cohen: I'm just not gonna sell the enterprise. Now, will you sell the enterprise anyway? Sometimes, yes. That's just gonna happen anyway. But it's not your ideal customer. When we talk about your ideal customer, it's not an ideal customer does not want SOC 2.
Jacob Luetzow: Right. Yeah.
Jason Cohen: So we so while sometimes they'll just show up and you're fine. Sure, if they want to buy it, fine. Because remember, we're not talking about that. We're talking about who's ideal. Well, someone who wants governance is not ideal.
Jacob Luetzow: Mm-hmm.
Jason Cohen: But some products are just the opposite. Yeah, we're ISO this and socked that and da-da-da-da-do. okay. Well, then for an enterprise, that's great. We're we're we're all in line here. But for someone that just wants a cheap piece of software that just does something, this actually is a deal breaker because It's all this complicated stuff and security. And you're like, I don't want all that. That actually makes it hard to use, right? Okay. So
Jacob Luetzow: Mm-hmm. Yeah.
Jason Cohen: of course that's in the eye of the beholder, which is the whole point. So there are these deal breakers where even though everything else is lined up, they still won't buy because dot dot dot. And then that cuts away parts of the market of the ideal. It's minus subtracting out of
Jacob Luetzow: Right. Yep.
Jason
Cohen: what's ideal. So these are the things that you do. And then at the very end, how do we does how do we actually write down in words? What this ideal customer is. we've already started talking about this stuff. And some of these bullet points, this inciting event happened, that counts. But it's still nice to say, what is it? And here we're tempted to say things like companies with 10 employees and this kind of demographic stuff. But that's rarely correct. It
Jacob Luetzow: I also always get stuck making it me, like solo founder. I
Jason Cohen: Mm-hmm.
Jacob Luetzow: have like all these servers no one knows how to deploy. So I need like my handoffs to go to someone capable of continuing my companies if something happens to me. But I don't know if I'm necessarily the ideal customer of my own product. I don't know.
Jason Cohen: I know you're not.
Jacob Luetzow: Yeah. Okay.
Jason Cohen: Yeah. Because a company wouldn't do it this way. A company wouldn't talk
Jacob Luetzow: Yeah, but it's Yeah.
Jason Cohen: about kill and it wouldn't be like that. A company will have some kind of shared thing because a company has more than one person that does anything.
Jacob Luetzow: Yeah. Except for today.
Jason Cohen: And so the
Jacob Luetzow: Today, we got a lot of solo founders popping up. How
Jason Cohen: No, not except for today. No.
Jacob Luetzow: do you figure?
Jason Cohen: Those th almost none of those are real businesses that are going concerns with real customers and everything. There's a lot of vibe coded stuff. There's very few. There's some. But if I talk about businesses, what percentage is one person? And the answer is, I don't know, point one percent. I don't know.
Jacob Luetzow: I bet it's growing pretty big.
Jason Cohen: Okay, if I'm selling two businesses today, what percentage of them
Jacob Luetzow: Mm.
Jason Cohen: is a w is one person by themselves?
Jacob Luetzow: I I don't know the statistic, but I would think it's
Jason Cohen: Is it ten is it is ten percent one person by themselves of software companies?
Jacob Luetzow: maybe. I don't know. I would just be making things
Jason Cohen: Really? You think ten percent of of software companies that you could sell software to is one person? That's actually what you think?
Jacob Luetzow: No, I don't think I'm just like I'm looking at the perspective of
Jason Cohen: Okay, what percentage it just power of ten? Is it ten percent, one percent, point one percent? What power of ten?
Jacob Luetzow: let's just say one percent or less, I guess if we wanna
Jason Cohen: You think it's one percent.
Jacob Luetzow: I'm just throwing bullshit numbers out. I have no idea. Yeah.
Jason Cohen: It's not. It's not. It's not.
Jacob Luetzow: But you're you're telling me you don't think that there's more solo founders right now making a good living off of one product that they're
Jason Cohen: That's not what I said. I didn't say that. You you're you're
Jacob Luetzow: You're a
Jason Cohen: you're changing you're changing
the topic in order to make a point. What I'm saying is businesses
Jacob Luetzow: I wasn't trying to. Maybe we just were talking about different things.
Jason Cohen: that are going concerns have a team of people. And so they need
Jacob Luetzow: Mm-hmm.
Jason Cohen: a team based password management, not someone died and now this one other person can do it, who pr who prior to today never had access to the servers and doesn't know how to do those deploys. That is not how a business operates.
Jacob Luetzow: W I think that
Jason Cohen: And when I say a business, I mean ninety nine point nine percent of businesses. That's what I mean.
Jacob Luetzow: Okay. All right, fair. I'll give you that. But I'm I'm looking at my perspective though. But
Jason Cohen: Well that's okay. Thank you.
Jacob Luetzow: like I I have businesses that make money and I'm in that exact problem, right?
Jason Cohen: I know. And you're not going to buy this software for it.
Jacob Luetzow: No.
Jason Cohen: No. Do you have people lined up who are pre-trained on how to do all those deploys right now?
Jacob Luetzow: No.
Jason Cohen: So you're making my point for me. A business that is a real going concern that cares about anything like this does not
Jacob Luetzow: Mm-hmm.
Jason Cohen: do it in the way that your software does it.
Jacob Luetzow: Mm-hmm.
Jason Cohen: Because you you can't operate a business like that.
Jacob Luetzow: Right.
Jason Cohen: Right? That's a fragile thing where if something happens to one person, the whole thing blows up. That is not a business that's a Going concern. Okay.
Jacob Luetzow: Fair, yeah.
Jason Cohen: Now look, I've operated businesses like that. I'm not putting it down. I that's me. And you know, no,
Jacob Luetzow: Yeah. Well you had to start somewhere. Like you you always yeah.
Jason Cohen: may never want anyone else at the business. I love that.
Jacob Luetzow: Yeah.
Jason Cohen: I have a podcast where I help founders. A lot of
Jacob Luetzow: Yeah.
Jason Cohen: them are like this. I love it. This is not a value judgment. I love it. And it and I've done it. I'm just saying if I'm selling software to software companies about server with that involve server credentials.
Jacob Luetzow: Mm-hmm.
Jason Cohen: there's the market of the solo founder who what if in case they die they want their spouse who's never used SSH before to do what exactly? Is that a good market? No, it's a really stupid market. That's not even a market. That doesn't exist. The
Jacob Luetzow: That's fair. Well, one of my really, just a really good software friend would inherit a company overnight. But yeah, there's
Jason Cohen: Yeah, that's that's a terrible, that's a terrible plan. Now a better plan would be, hey, now that I'm at this point in the company's life, I need to start thinking about things like sustainability, durability, predictability. I want to start thinking
Jacob Luetzow: Mm-hmm.
Jason Cohen: about that. And I'm not saying you have to, I'm just saying you could. It's
very logical to do that. I want to protect what I've got. I want my customers
Jacob Luetzow: Mm-hmm.
Jason Cohen: to have continuity if I die. I'm starting, I, you know what? I want to start prioritizing that. You could decide.
Jacob Luetzow: And obviously that's important, right?
Jason Cohen: You could decide that. Now, if I go to Twitter and find all the solo founders, how many have decided that?
Jacob Luetzow: Probably not many. Yeah.
Jason Cohen: Almost zero. And there already are almost none of them. So this is a very, very, very tiny little market. But okay, you could say that that's my market. But I think we're getting into pretty weird territory. Even so, what would my DR even so what would my DR plan be? Would it
Jacob Luetzow: No, I think we agree we're I'm not the right market. Yeah.
Jason Cohen: be to designate someone to get all my API keys? But I'm never gonna let them log in now. Nope. Only if I
Jacob Luetzow: Yeah, fair.
Jason Cohen: die. They won't even have tried it once because they don't have access. Because that's what your software does.
Jacob Luetzow: Mm-hmm.
Jason Cohen: So even if I'm a solar founder, who has decided to have a continuity policy, a DR policy of myself? I still wouldn't use your software for that. It's
Jacob Luetzow: Yeah.
Jason Cohen: still the wrong software. So my point is this is why I'm just adamant, like, no, it's not. Stop it. Stop it.
Jacob Luetzow: No, you're right. You're right. And I don't market towards me, but that's what originally started the idea, right?
Jason Cohen: Of course you don't, because it doesn't make sense. Now, now it's good that we had this conversation though, because it's not obvious unless you talk through it like we just did. It's not obvious whether that's
Jacob Luetzow: Right. Yeah.
Jason Cohen: the case or not. So we need to talk through it. And so for businesses, maybe them. And then I would say no, because businesses, you know, the 99.9% of software businesses who are a going concern who would ever think about things like DR, they have a team of people. Even if it's
Jacob Luetzow: Mm-hmm.
Jason Cohen: a team of two or three, it's not one. And so that team Does things like deploys, or it's in GitHub already, and that has the secrets to deploy. it's not one person with their laptop. That's not what's happening.
Jacob Luetzow: Mm-hmm.
Jason Cohen: They have CI C D deploys and another person, and And so they need a system to manage that, whatever that means. There's lots of systems that do that, but that's what they need. Not one person dies and now one other person suddenly has access. That is not how you run a sustainable business.
Jacob Luetzow: No, of course not. And obviously it's not ideal, but it
Jason Cohen: It's less it's not it's worse than not ideal. It you it's a horrible, terrible thing that no one would want to do on purpose. No one would do that.
Jacob Luetzow: I don't think you wanna be in that position for at least for a very long time. I'm obviously it's always on the top of my mind, right? Because Kin switch needs to be something that people rely on and trust and it's always going to work when they need it most, right?
Jason Cohen: Obviously, but you didn't build the software that I'm describing, and nor should you, because that already exists.
Jacob Luetzow: Right. Right. Right. Yeah.
Jason Cohen: Right? so so the my point is simply that as you think through this, you realize stuff like okay, so a company with their people that deploy stuff, because as you brought up, this idea of of deployment keys is absolutely not Carol. Right? And so we won't
Jacob Luetzow: I agree. I agree completely. And I'm actually glad that we actually even dove into it more because it made it even more
Jason Cohen: Meet y it
Jacob Luetzow: apparent.
Jason Cohen: Yeah, exactly. In fact it's a deal breaker that they're at a company and they're thinking about it
Jacob Luetzow: Mm-hmm.
Jason Cohen: for a company. Now you could argue it's not a deal breaker, and that's fine. on the fringes, I
Jacob Luetzow: Yeah.
Jason Cohen: don't actually know. I'm just talking, right? As an operator myself. But you could you could find out that lots of people want it. I don't know. great,
Jacob Luetzow: Yeah.
Jason Cohen: great. They found a market. I didn't know existed. That's terrific. You should do things like that, you know?
Jacob Luetzow: Well, I it's like the typical, it's the bus factor, right? Bus factor
Jason Cohen: Yes.
Jacob Luetzow: of one is a big problem and that's commonly spoken about, at least in like the startup world. And that's not a good spot to be in.
Jason Cohen: Right. Right. And so your software, your software doesn't fix that. You just change who's
Jacob Luetzow: Right, it does not.
Jason Cohen: the one person. Now for a family, that's beautiful that's perfect. But
Jacob Luetzow: Mm-hmm.
Jason Cohen: for but for, you know, WP Engine, where we have 250 engineers and we have and like a third of them are in infrastructure, that would be bad. Right? That's not how we should operate. That would be
Jacob Luetzow: That would be bad.
Jason Cohen: bad. Okay. So anyway, so so it's nice talking through all this because you start Go, okay, so that's not my Carol for sure. And that's why I have these deal breakers that subtract, because it's good to subtract.
Jacob Luetzow: Yeah.
Jason Cohen: Okay, that's not it. That's not it. Right. Saying what it's not helps you. It doesn't say what it is, but it does help. it does help you get to yeah, yeah. So it
Jacob Luetzow: It does, yeah. It gets a l a lot of the noise out of the way.
Jason Cohen: gets well, yeah, and things that it's not. So anyway, so this is just a very long because because I didn't want to have people have to buy the book to to get all of the information. So this is.
Jacob Luetzow: The link will be in the description of the podcast. You guys should all go and buy the book. It's great.
Jason Cohen: Yes, you should anyway. But yeah. So yeah, it's it's it's so
and as usual with these things, half the value is in the process of having to go through it and decide
Jacob Luetzow: Mm-hmm.
Jason Cohen: make all these decisions that you haven't been making all along. And look
Jacob Luetzow: Yeah.
Jason Cohen: at look what happens when you get down to like it's wealthy families where there's a death in the family and they set up this thing, it or whatever, right? You da-da-da-da. And then the ads just and ideas just start flowing. And what do I write about in SEO? Just starts flowing. And what kinds of things would I A B test on the homepage? And by the way, what can they afford? Well, now that you've painted the
Jacob Luetzow: Mm-hmm.
Jason Cohen: picture so clearly, I can understand that better. And what is their alternatives? And who does they sell through? are you selling direct or are you selling through something like in your case, lawyers or in other cases, agencies that build stuff? Like right now in AI, agencies are building a lot of the AI stuff, in fact. So it might be better
Jacob Luetzow: Mm-hmm.
Jason Cohen: to sell your AI tool through an agency who's developing stuff than it is to sell direct sometimes. And so, like that's interesting. So, okay, so it just starts suggesting all this stuff. And then what are you building features for? Now you've got this crisp idea. So whatever that person wants, that's what I'm building features for. If anyone
Jacob Luetzow: Mm-hmm.
Jason Cohen: else likes it too, they're welcome. But if they don't like it, I don't care. It just crisps
Jacob Luetzow: Yeah.
Jason Cohen: up your prioritization of features, bugs, and where this all came from, cancellations. Because now, if this ideal person cancels, you're like, no, it was built for you, and you said no. It was built for you. You saw my homepage in pricing, you said yes. And then you actually started using it and you said, wait, no, I have got to find out why. Yes,
Jacob Luetzow: Yeah. Yeah.
Jason Cohen: hell yes, that's worth a lot of your time to find out why, because this is the whole point. On the other hand,
Jacob Luetzow: Yeah.
Jason Cohen: if somebody who's a a solopreneur with their deploy keys cancels, you're like, don't even I don't even want to talk to them. It doesn't matter why it didn't work out for them. I don't care.
Jacob Luetzow: Yeah, fair. Yeah.
Jason Cohen: It just doesn't matter. You were never right for me. I was happy you were here, but I'm not crying that you left. Right? I don't I it's
Jacob Luetzow: Yeah. Yep.
Jason Cohen: just distracting and a waste of time to even know what you think. Mm.
Jacob Luetzow: And it's so easy to get distracted by the the
Jason Cohen: Yeah.
Jacob Luetzow: customers that don't matter. Cause you will like your instinct, at least as an engineer, is to well, I can also solve that problem. But then you get distracted by your actual true customer.
Jason Cohen: Exactly. And that's why going through this process to really decide exactly what is it, then you can act
upon it. And without doing that, you can't act. You're like, is that ideal? You're like,
Jacob Luetzow: Yeah, yeah.
Jason Cohen: I don't know. You know, you don't know. So you gotta that by the way, still having made all that definition, how do you know if a certain person who canceled is an ICP or not? So that begs the question of like, yeah, do we have that information in our database? Well, how would we know? Does it depend on what they were doing? Did we ask them questions when they onboarded so that we can slot them in? What questions would we
Jacob Luetzow: Mm.
Jason Cohen: ask that they would actually answer and not, stack up the onboarding? Should we let them onboard first and then ask? That way, if they ignore us, at least they caught using the software and we didn't make it a barrier.
Jacob Luetzow: Yeah.
Jason Cohen: That's an interesting question. could we ask, could we drip questions out as long as they keep answer answering? Maybe we can drip stuff out and collect data a little bit when we can. In other words, can we do our best? That's all we can do. This is not gonna be perfect. Can we do our best?
Jacob Luetzow: I now I I'd be interested in your take on this. I find onboarding questions very hard to do well. It's like hard to know timing if people are annoyed by them. It's hard.
Jason Cohen: Yeah. I think so I I I'm not an expert on this, but some thoughts are it when in doubt, don't bother them because it's better that they do
Jacob Luetzow: Mm-hmm.
Jason Cohen: they get the value, they get time to value quick. That's the most important thing, period. The end. So when
Jacob Luetzow: Yeah.
Jason Cohen: in doubt, ask them later. And yes, that means most won't answer.
Jacob Luetzow: Yeah, yeah, that makes sense.
Jason Cohen: Fine, at least they're using the damn product. You know, like because as much
Jacob Luetzow: Yeah, that's that's true.
Jason Cohen: as as much as I do believe everything I just said, let's be real. We want them to be customers, okay? Giving
Jacob Luetzow: Yeah.
Jason Cohen: a break, right? Okay. number two. If the only is there any benefit to the customer to answer this question? So if it's a selfish question, it only benefits you and doesn't benefit them, then you are interrupting them and you are annoying them.
Jacob Luetzow: Mm-hmm.
Jason Cohen: Now maybe you could do a survey later, maybe some want to answer, okay, but like it's you know, that's what you're doing. Let's be honest. But sometimes there's questions which not only can help you categorize and qualify them, but also Can set up the software for them so it's more likely that they will be successful with it. So for example, we're using Riverside right now. Some people on Riverside are people that want to make videos. Some people using Riverside are are guests on a podcast who have never used it before. Some are producers that do random things while the show's going on. So Riverside wants to know who you are so that it can configure the UI to do whatever you're going to want to do.
Jacob Luetzow: Mm.
Jason Cohen: And it can say that. It can say, hey, Riverside can do many things. We would like it to be best for you. So if you can tell us just a little bit about what you're doing, we'll make sure it's configured perfectly for you on day on minute one. Say that. Like literally say that. Now
Jacob Luetzow: Mm-hmm. Yeah.
Jason Cohen: it's benefit to me, the customer. Now I'm gonna get a better UX and more likely to be succeed in this. I answer these questions. Now it's configuration. Now it's just customization, personalization, configuration. That's good for me, the customer. But at the same time, you're learning what's up, right? Okay. So don't be selfish. That's
Jacob Luetzow: I like that. That's actually really smart. Yeah.
Jason Cohen: a so that I would scat, I would title that don't be selfish. Get
Jacob Luetzow: Yeah, that's good.
Jason Cohen: what you need, yes, but do it in a way that helps them. It can help them just a little tiny bit. Like you and I know Riverside is 99.98% the same for everybody, actually. Right.
Jacob Luetzow: Yeah, yeah.
Jason Cohen: So okay, you doesn't have to be much. All right. Doesn't have to be much, just a little bit. Okay.
Jacob Luetzow: Yeah, it might be a button size changes, right? So it's like more apparent that you want to go click on that thing because that's what you're using it for and yeah.
Jason Cohen: Yeah. It could be the default of what you might want to do first. It could be
Jacob Luetzow: Yeah.
Jason Cohen: whether this thing goes into an advanced menu somewhere or it's right right in front of you, right? it could be very minor, but
Jacob Luetzow: Mm-hmm. Yeah.
Jason Cohen: at least you're saying that it's for their benefit. And it hopefully it's the dots are connected a little bit as you figure this out. So in the case of something
Jacob Luetzow: Yeah. No, I like that.
Jason Cohen: like a WP engine where we host WordPress sites, you a very illogical thing would be like, hey, is this like an e-commerce site? Is it a media site? Is it a personal blog? Is it a one page homepage? Is it a landing page? you could ask these things. Well, so that we can set up a site that's like that.
Jacob Luetzow: Right. you need to know if you're stream
Jason Cohen: Right. it's just so obvious, right?
Jacob Luetzow: if you're streaming video versus just displaying products. There's a big difference to how you serve that.
Jason Cohen: Right. And and configure it and what the default website might look like. Of course there would be. In
Jacob Luetzow: Mm-hmm.
Jason Cohen: fact, you're helping me by setting this up right. That's good. I don't have to set up whatever that is, I don't have
Jacob Luetzow: Yeah, yeah.
Jason Cohen: to know what all these knobs are. at least not on day one. The knobs will be okay, reasonable, and I can still mess with the knobs later. That's ideal, actually. You're help you are helping
Jacob Luetzow: Mm.
Jason Cohen: me. So obviously
all this is going to be very, very product and marketing customer specific, what you actually ask and do, but you get the idea here. Don't be selfish.
Jacob Luetzow: Yeah. I like that. That's a good tip. And honestly, it's given me a lot of ideas to ask questions in my onboarding process now too. Instead of like just being a tour of the product and helping them, you know, make their first couple actions, you can actually get some information.
Jason Cohen: Yeah, again, I think a product tour is perfectly fine. So of course I'm not of course I'm not criticizing that.
Jacob Luetzow: Yeah.
Jason Cohen: But if we wanted a nitpick, I would say, do they want a tour of the whole product? Do they even
Jacob Luetzow: Maybe not.
Jason Cohen: do they even want the tour of one, two, three? Is that what they want?
Jacob Luetzow: Yeah. That's true. Well, ideally you don't need a tour of a product 'cause it's so intuitive and the your your UX is great and they just know what
Jason Cohen: Or
Jacob Luetzow: to do.
Jason Cohen: or the tour g actually sets them up with something. So in other words, well there you go. Yeah.
Jacob Luetzow: Right. That's what my current tour does. It actually gets them creating their first document and creating their first dead man switch and all of that.
Jason Cohen: Right. Right. So maybe that's more of a tutorial or handheld tutorials handheld setup.
Jacob Luetzow: Yes, yes.
Jason Cohen: Yeah. So to me that's awesome. But I've of course we've all used products where it's like on the left here you'll find your files. I'm like, I I can see that. I've seen I've done this before,
Jacob Luetzow: yeah, it says file. Like, yeah, that yeah. And you always get annoyed, especially if there's not a skip or like just close the whole tutorial. Like, stop it.
Jason Cohen: Yeah, and then I'll say then I'll say something like, now over here, when you're ready, this is how you publish. Well, I'm not ready, so am I even gonna remember that? Is that what I wanted to know right now? Or should you pop
Jacob Luetzow: Yeah.
Jason Cohen: that up later when I've got something that maybe I could publish? Now
Jacob Luetzow: Yeah. Yep.
Jason Cohen: I'm nitpicking. I'm admitting I'm nit nitpicking. This is not the most important thing. Like you probably have more important things to do today than to worry about that.
Jacob Luetzow: No, no, no. It's good. This is good. But it's good to think about flow. I I personally think onboarding is a very difficult thing and it's very product specific. And I don't there's no right answer for everyone's product, right?
Jason Cohen: Yeah. I I guess what what it's doing is it's it is highlighting what I think is wise and and helpful, which is am I always thinking what the what would in fact be best for the customer right now? What would either be something they need, even if they know it or not, or that would delight them, whether they know it or not? Or just whatever they actually want, whatever that is, whether they whether
that's a good idea or not, it's just what they freaking want right now. Can I deliver that?
Jacob Luetzow: Yeah.
Jason Cohen: This mentality of like what is actually what they want. Right now. That's the right mentality. Not what do I want to do? I want them to learn my product. So I showed them a tour. I want to know something from my database. So I asked them a question. Well, there's that too, but let's call that a Venn diagram circle. And there's this other circle of what does the customer actually want? And maybe we should ensure that that second circle. Let's let's over optimize on that. And let's see
Jacob Luetzow: Yeah.
Jason Cohen: if we can fit your circle into that because the intersection is definitely better. Right? So let's let's do that.
Jacob Luetzow: Right. Yeah.
Jason Cohen: But let's one of those is more important than the other and let's try to scoot the other one in, that's a good mentality. So I think that rule of thumb for everything in onboarding really everything, period, but okay, onboarding especially, this would be a good rule of thumb.
Jacob Luetzow: Yeah. Yeah, for sure. Yeah, now I have a lot to think about. I'm curious, with like WP engine, onboarding is probably was like probably one of your most critical things, right? Because you're setting up kind of like done for you hosting, right? You don't have to think about it and
Jason Cohen: I it's hard to think of a company where onboarding is not one of their most critical things.
Jacob Luetzow: Yeah, fair. Okay. Because you want them to use your product immediately and they have to know what they're doing.
Jason Cohen: Mean in product we c we say things like time to value.
Jacob Luetzow: Yeah.
Jason Cohen: you can look at any you can look at the history of any s big successful company and there's always stories about at Netflix, when we first did CDs, we discovered that a person that put three or more titles in their queue in the first four days was nine times more likely to be still be a customer six months later. So we, you know, would suggest and nudge
Jacob Luetzow: Yeah.
Jason Cohen: them to, you know, okay. Or we just filled it with titles and they were like, well, I don't like that, but it would cause them to go change it out, right? Like,
Jacob Luetzow: interesting.
Jason Cohen: You look at any
Jacob Luetzow: Yeah.
Jason Cohen: company that's successful, you will see these little patterns of like, Yeah, the onboarding's really important.
Jacob Luetzow: Yeah. Mine is if they confirm their email, create at least one upload a document or create a a note and then create a switch, they'll they'll usually always convert.
Jason Cohen: Love it.
Jacob Luetzow: Yeah. And if they don't do that, I kinda know they're gonna cancel and then I start, you know, emailing them and try to get information out of them if I can.
Jason Cohen: Mm-hmm.
Jacob Luetzow: But I would say you're lucky if I don't know, ten percent's probably a big number. I don't even think ten percent respond, but the little golden nuggets that do, it's very beneficial.
Jason Cohen: So at some now this gets to the difference between the vision statement of what we would like to have happen and what happens. Of course, right? Like you want that, and and you
Jacob Luetzow: You want them all to respond.
Jason Cohen: know that the responses are a heavily biased sample.
Jacob Luetzow: Yes.
Jason Cohen: It is not a random sample of people who respond, right?
Jacob Luetzow: That is true.
Jason Cohen: How is it biased? I don't even know. I don't even claim to know. I just know it is, right?
Jacob Luetzow: Yeah.
Jason Cohen: So, well, if we were designing an experiment, it wouldn't work like this. I know. But welcome to the real world where that's not happening. So, of course, when I when I say what would be good, I of course I'm ex I'm saying a vision statement of what would what we're going for, because it's useful to say what we're going for. Because reality
Jacob Luetzow: Mm-hmm.
Jason Cohen: is harsh and we can't get there in reality. Therefore, it's nice to say what we would like. And then we get we do as well as we can. Of course. That's what happens in real life. It's as good as we can. It's better than nothing. Like you said, 10% respond, and maybe only half of those are are even actionable. But then I get nuggets and that's enough.
Jacob Luetzow: Right. Yep.
Jason Cohen: Yeah.
Jacob Luetzow: You could have one customer give you a great response like once a year, once a quarter, and it's probably worth a thousand emails to everyone, you know?
Jason Cohen: Yeah, and we've got to intersect this stuff with real life and get real. I've said things
Jacob Luetzow: Yeah.
Jason Cohen: too, like the onboarding, you should check it and there should be all these objective measures of where are they, but is all that really an objective measure? Not usually. Aren't there aren't
Jacob Luetzow: Yeah.
Jason Cohen: there ranges on these numbers at least? Yes. Is it always linear? Do the people hop around? yeah, okay. Right? Like, of course reality is more complicated, but still, it's hard for me to convey the idea without s making it simplistic.
Jacob Luetzow: Yeah. Yeah.
Jason Cohen: And it's useful even for yourself to keep the simplistic goal in mind because at least that's the North Star. That's the idea. It's like, okay, but that's what we're trying to do though. That's helpful.
Jacob Luetzow: Mm-hmm.
Jason Cohen: Let's not get confused though with the the the helpful the helpful goal. The help with well, but what can we do? Aha, yes. Yes, there's that. And all but
Jacob Luetzow: Yeah.
Jason Cohen: also let's not let ourselves off the hook too fast. Like, is 10% a good response rate? Well, I know I even have an example in the book. It's okay, but you can definitely get to 20. You probably can't get to 30 though. So
Jacob Luetzow: Yeah.
Jason Cohen: should you beat yourself up? And is that the next thing you should try to do? Probably not. But if I were at
1% response rate, I would say, you're just not getting value out here. And your customers are not responsive. Your customers are not engaging you. You're you probably need to fix that because you're just not getting the information at all. And so if it was
Jacob Luetzow: Yeah.
Jason Cohen: useful to get any information, that's not happening. So you're probably not done here. Even then, of course, you could argue with everything I just said, but you know, that's that's yeah.
Jacob Luetzow: I would take a one percent response, my landing page probably isn't conveying the proper messaging and they think they're signing up for something different. I don't know though.
Jason Cohen: They're just so disengaged at that point. Something's wrong. So for
Jacob Luetzow: Yeah.
Jason Cohen: example, one one failure mode I see a lot is people will say, my NPS is really high. And then you ask, like, what happened? how did you do that? And they're like, Well, I emailed my customers and eight percent responded, and I got an NPS of 65. Okay, so of the people who responded. So first of all, you didn't ask the people who canceled. What do you think their MPS would be?
Jacob Luetzow: Yeah.
Jason Cohen: Many times more people have canceled than are still there in the history of the company. And yet you only you didn't ask them. So first of all, it's not 65%. It's 65% of the people who weren't upset with you, which is which might be the minority. Second of all
Jacob Luetzow: Right. And normally you don't hear from happy customers. They'd like I don't like to fill out surveys or respond. Yeah.
Jason Cohen: I don't know. Second of all, you said second of all, you said 8% responded. That means 92% were so apathetic towards you that they they they they care about you so little that they wouldn't they can't they they can't even be bothered to write on to like click a number in an email. So
Jacob Luetzow: Yeah.
Jason Cohen: are you sure it's 65? Are you sure people who couldn't be bothered, you're saying are what, just like the people who could be bothered? You're saying they're what, a 10? Is that what you're saying? So Apathy is a huge signal. And so if you don't if you're looking at the responses and not the response rate, you're not measuring apathy. And apathy is bad.
Jacob Luetzow: Yeah.
Jason Cohen: Now, can you are you gonna get an eighty percent response rate? No. Come on. Come on. But when you when your response rate is low, that's a signal.
Jacob Luetzow: Yeah. Yeah, that's interesting. I never even thought about that. Because it's
Jason Cohen: Let me give you a counterexample.
Jacob Luetzow: sometimes it's even hard to get responses like just, happy customers. And if they're not replying, what does that mean about your product or yeah.
Jason Cohen: That's right. It means they don't care. Yeah, that means they don't care. They don't care about your future. They don't even they want they don't want to give you feature requests. They just don't even care. That's bad.
Now, again,
Jacob Luetzow: Yeah. That is.
Jason Cohen: that may not be the next thing you should action today, but
Jacob Luetzow: Mm-hmm.
Jason Cohen: let's let let's not let's be real about what this is and not just ignore it. So for
Jacob Luetzow: Mm-hmm.
Jason Cohen: let me give you a counterexample. At WP Engine, our customers are very engaged and it's very good, including with things that they want to be better, right? But they're engaged.
Jacob Luetzow: Yeah.
Jason Cohen: So we wanted to launch, we were thinking about launching, this is years ago. We're thinking about launching a new product in the headless space, headless WordPress. Is that a good product idea? Let's talk to some of our agency partners who build websites and see if they like to build this way. Do they build this way now? Would they like to build this way? What should we do? So in inside of one week, we had a hundred calls scheduled. Scheduled, like it's on the calendar.
Jacob Luetzow: Mm-hmm.
Jason Cohen: That's an engaged customer base.
Jacob Luetzow: Yeah.
Jason Cohen: Now it It it doesn't have to be that way, right? But if it's a product that maybe people are might be interested in for real. Now that doesn't prove we should build it, but it's like, okay, that's an it's already a good signal. If they're if it's a product they're interested in and the customers give a crap about you, they will engage. So it's a signal
Jacob Luetzow: Yeah.
Jason Cohen: when they don't.
Jacob Luetzow: Mm hmm. Yeah. Yeah. All these painful data points that you swallow as a a founder.
Jason Cohen: We ignore them because they're painful.
Jacob Luetzow: Yeah, for sure.
Jason Cohen: Or maybe we just don't know. We didn't think of it. That's fine.
Jacob Luetzow: Yeah. I think that's the at least the biggest problem from my learning curve of building a product, listening to customers, and getting out of my getting the engineer of me out of the way.
Jason Cohen: Yeah. Well, the truth hurts. And also, we may not have just thought of that before. You
Jacob Luetzow: Yeah. Yeah.
Jason Cohen: just like like I think onboarding's like that. The math is clear. Once you see it, you're like, Jesus. Okay. Duh.
Jacob Luetzow: Yeah.
Jason Cohen: But okay, what happened before that? I don't know. I just hadn't heard that before. That's all. Like there's there's no
Jacob Luetzow: Mm-hmm.
Jason Cohen: there's no shame in that. Today's the day. Today was the inciting event. Today you had an inciting event, which is podcast.
Jacob Luetzow: Well, you don't know what you don't know until you learn, right? Yeah.
Jason Cohen: Well you learned this. Okay. Well you got the book
Jacob Luetzow: That's right.
Jason Cohen: because of the the book is full of stuff like this. So maybe today's the day you got that. That's why the book's helpful, because it's saying stuff that in retrospect
Jacob Luetzow: Yeah. It is.
Jason Cohen: you're like, yeah, right. Okay. But like maybe you didn't think of that yet. So good. That's why there's a book.
Jacob Luetzow: And that's what it like the the funny thing is, every single chapter, it's like, well, duh, that makes sense, but it's something you don't really think about.
Jason Cohen: Yeah, hopefully you say that, right? Otherwise
Jacob Luetzow: Yeah.
Jason Cohen: Yeah. Yeah.
Jacob Luetzow: Otherwise you got more work to do. I'm gonna come back to the booklet. No, I don't know. I just like I I think it's just always good to hear perspective, especially from someone that's lived through building successful products. I think that's because I spent like 10 years of my life like, you know, grinding hard in startups where you never really grow past a few thousand users. And so you don't really I don't know, you you don't learn that much. You don't learn what it like the pains of scaling a scaling your servers, having thousands of requests a second. Like you don't ever hit those pain points and you don't you don't learn those things until you learn those things.
Jason Cohen: You also don't learn how to make a company that grows past that. people say, Well, I
Jacob Luetzow: Right. Yeah.
Jason Cohen: failed, but I learned. And and and you probably did learn some things, but not as much as you think. Because what you did not learn
Jacob Luetzow: To right.
Jason Cohen: is how to make a company that works. That's not what you learned.
Jacob Luetzow: Yeah. Yeah. Yep. And it's hard. It's hard. You learn a lot of new things.
Jason Cohen: I think what's funny about the learning is I actually think you learn less from either success or failure than people give it credit for. In other words, people like to assuage their themselves by saying, Well, it's a failure, but I learned. And of course you did learn some things, but again, like saying so, for example, let's say I
Jacob Luetzow: It makes it hurt last saying that,
Jason Cohen: said, well, I tried Google Ads and it didn't work. But I sure learned a lot. Well, I don't I don't know whether you learned something or not, but here's what you definitely did not learn. How to make a Google ad that works. And so did you learn anything useful or very much that's useful? Was it that
Jacob Luetzow: Yeah.
Jason Cohen: great of a learning experience? Or did or you still just don't know how to do that? Right. so now still sometimes it's like, well, yeah, sometimes sometimes like you might have learned if you build it, will they come? And the answer is no, and now you learn that. So I agree, like that maybe, maybe you needed to see that.
Jacob Luetzow: That is sad. I believed that ten years ago. That was my, you build it well. It's a product that people want. They'll find it.
Jason Cohen: Right. So that's not true. And and maybe that is an example where maybe through failure you could learn that lesson. So fair
Jacob Luetzow: Yes.
Jason Cohen: enough. So so again, sometimes you can, but I think we learn a lot less than we say we do. But here's the other thing. When you look at a successful company, they
also had the same failures and problems that you had when you failed. Like
Jacob Luetzow: Mm-hmm.
Jason Cohen: you can look at a successful company who also failed to do AdWords very well, but they were
Jacob Luetzow: Yeah.
Jason Cohen: successful. So you're sure you learn the difference between success and failure because they were full of crap too. They fell full of garbage in their company too.
Jacob Luetzow: Mm-hmm.
Jason Cohen: But somehow they succeeded anyway. So are you sure which things were the difference between success and failure? Because just because something on your side didn't work out, that does not mean that's why it failed. You don't know that.
Jacob Luetzow: Right.
Jason Cohen: Neither do I. Nobody knows that. That's the point. It's not an experiment we can run again. So nobody knows. But I would say the same of success. look, W pinchon's such a success, which it is. So what now I know what? I know how to do Facebook ads for everyone? Are you sure?
Jacob Luetzow: Yeah.
Jason Cohen: Or did it or was it successful despite our terrible Facebook ads? Uh-huh. Yeah, that's probably true. That's probably right. Yeah, that's really the answer. Right? So did so did I learn anything? Yes. But what? probably not probably unc less clear than we'd hope it is, is what I would say.
Jacob Luetzow: Right. Yeah, I would say there's skill and knowledge, but there's also a bit of luck with growing any business, right?
Jason Cohen: Hmm? A bit?
Jacob Luetzow: probably more luck than you wanna give it credit for. Yeah. Yeah.
Jason Cohen: Here we go again. I get I agree. Right? How much? Nobody knows.
Jacob Luetzow: Yeah.
Jason Cohen: Nobody knows. And also, even in retrospect, no one still knows. In retrospect,
Jacob Luetzow: Right. Yeah.
Jason Cohen: how much of W Pinchin's success is luck versus work? I don't know.
Jacob Luetzow: Yeah.
Jason Cohen: So if I can't even know in retrospect when I have all the information, all the all the consequences that came next, and I still can't know, well, I certainly couldn't know in the moment, could I?
Jacob Luetzow: No, but there must be something
Jason Cohen: Okay.
Jacob Luetzow: in your process in your approach to solving problems that does work, right? And
Jason Cohen: Yeah, it's clearly clearly successes have to be a mixture of skill and luck. And we don't know in what proportion and we don't know which
Jacob Luetzow: Mm-hmm.
Jason Cohen: of those of the skill part is, was really the thing that that that really pushed it over the top. So all we can do is at least what all I think all I can do is what I do, which is I think some of these ideas are pretty good. That does not mean it'll work for everyone else. And I can't prove that it's like the critical idea for W pinch, but it seems like it's pretty good. So I'm gonna try to write them up as best I can. And I'll give them out. And then you decide if you think it's a g idea worth taking on board or not. And that's it. The end of story.
Jacob Luetzow: Well, if if anything, it gives you a new it can shift your perspective and that can help you just approach a problem differently, right?
Jason Cohen: Yeah. So I so I think if the idea is plausible and you like it, that's already a pretty good start. Cause at least you'll take at least you'll give it a good shot because you liked it. So, so
Jacob Luetzow: Right.
Jason Cohen: you'll you'll have legitimately given it good shot. And look, there's probably twenty good ways to do goal setting and twenty good ways to do metrics and twenty good ways to prioritize tickets for pro they're all fine. So if
Jacob Luetzow: Yeah.
Jason Cohen: they're all fine, then why don't you pick one That really seems to resonate with you personally, because there's a better chance that you can get that to work. That's what I think. So
Jacob Luetzow: Yep, I agree. Yeah.
Jason Cohen: everything that I write, everything I put in the book, that's how that's my attitude toward it. This is probably a pretty decent idea. That does and all it means is that. Now, may it may any one of them may or may not resonate with you where you're like, ooh, that. If it does, then try it. And if it doesn't, don't. Like it's that simple.
Jacob Luetzow: Yeah. yeah. Yeah. No, and I think that's just life, right? You pick and choose strategies
Jason Cohen: Yeah. Mm-hmm.
Jacob Luetzow: and workflows that work for you. And
Jason Cohen: Mm-hmm.
Jacob Luetzow: like even w how fast AI is shifting and how we write code. I'll hear of how some of my buddies are doing it, how other people are doing it. And
Jason Cohen: Yeah.
Jacob Luetzow: some of them I'm like, I don't want like that's so complicated. I don't want that. And I don't adapt. I and I'm sure they work great, but it it Doesn't fit my workflow.
Jason Cohen: Exactly. So that's that's that's the approach. So my analogy is it's you're trying on clothes. So you see an idea, whatever, it's a process, it's a workflow, it's AI, and and you're like, and and your attitude rather than that's how to do it, I'll do it, like scrum or something, should be,
Jacob Luetzow: Mm-hmm.
Jason Cohen: well, let's try this on like clothes, because if we're lucky, it comes off the rack and just fits. We're like, damn, all right.
Jacob Luetzow: Mm-hmm.
Jason Cohen: And so good. Check plus, move on with your life. But that's usually not what happens. Usually you take it off, it's like, it's a shirt, but it
Jacob Luetzow: Yeah.
Jason Cohen: needs to be altered. So you take this idea and you don't expect it to just work on day one. You expect to have to alter it a bit to match you. And then sometimes you take it off rack and no amount of alteration is gonna fix this thing for you, for you. It's fine for somebody else, not for you. So
Jacob Luetzow: Right. Yep.
Jason Cohen: you put it back and try something else. No problem.
Jacob Luetzow: Yeah, no, I like that.
Jason Cohen: Now, when you have that attitude, then you have this like experimental attitude. Let's be critical
about it. Let's not have be dogmatic. There's only one way to do scrum. no, there's a way to start with and then we'll try it on and go from there. So when you have that attitude, you can you can
Jacob Luetzow: We all know those people though.
Jason Cohen: try to iterate your way to something that does work. So that to me is the right mentality for this question of like workflows frameworks, et cetera.
Jacob Luetzow: Yeah. Yeah, I think that's a good a good way to approach it.
Jason Cohen: What what's bad is what is and we've all experienced this where the CEO of the company goes, All right, everyone, now we're using OKRs.
Jacob Luetzow: Yeah.
Jason Cohen: And the chance that now everyone we're using OKRs is gonna work is low. Right? Whereas
Jacob Luetzow: Yeah. Yep. Yeah.
Jason Cohen: if you did what I said, you said, let's take a team, let's have them try OKRs. Then they'll iterate, they'll report out.
Jacob Luetzow: And go from yeah.
Jason Cohen: Then maybe as maybe another team or that department tries. Maybe, Maybe we give up. Maybe it works for some people, not others. Now we're cooking. Now we're trying stuff because we gotta we wanna, the presumption is we we wanna get better at something like goal setting. Okay, hopefully we agree we want to get better. But on the
Jacob Luetzow: Mm-hmm.
Jason Cohen: how we've got to take this. You could say it's experimentation, but I'm trying to be more specific than just experiment and test. You you have the specific idea in mind that I'm trying it on, and then I'll need to iterate. Because then we need to come back every week or two and speci and make a meeting. How's it going? Do we like this still?
Jacob Luetzow: Mm-hmm.
Jason Cohen: What do we want to change or keep? And by the way, even if you end up, after a few of these, rejecting it wholesale, you will have now a list of we liked think, dink, dink, dink. We didn't like think, think, dink, dink. And now when you go to find the next framework to try, you now have all this criteria
Jacob Luetzow: Mm-hmm.
Jason Cohen: that's that is real, empirical criteria that you didn't have before. So now the chance that the next one is a much closer fit goes way up.
Jacob Luetzow: Yeah. Yep.
Jason Cohen: So even if this quote unquote fails, you now have material you can use to dramatically increase the chance that the next thing will succeed.
Jacob Luetzow: Yeah.
Jason Cohen: So this is more than just saying test, right? This is a specific thing that is testing. It is, but like the specific mentality and thing to do to try to find a framework or process or whatever that's right for your. size for your culture and the size of company and your goal, right, for whatever that means, right?
Jacob Luetzow: In every aspect of life and building a business or building a product, it's all just different different forms of A B testing.
Jason Cohen: I I don't agree. No.
Jacob Luetzow: No. I'm curious why or w what don't you agree with?
Jason Cohen: A-B testing is good when you already know something like this is the right message for my customer. And now
Jacob Luetzow: Yeah.
Jason Cohen: what is the right way to say it to that really resonates with them and gets the click? But
Jacob Luetzow: Mm-hmm.
Jason Cohen: if you haven't picked the right ICP, A B testing that ad is not going to fix that problem. And so
Jacob Luetzow: Right. Well, 'cause both probably both your tests are irrelevant. Yeah. Yeah.
Jason Cohen: It's the wrong customer. So who cares? So when you it's it's so when you know what you're what to do, then A-B testing is a good way to optimize that. But if
Jacob Luetzow: Yeah, fair. Yeah.
Jason Cohen: you don't know what to do, A B testing is not how you figure it out.
Jacob Luetzow: Yeah.
Jason Cohen: And so I call these two, there's two modalities of working. There's not just two. I just have this particular, particular view that's helpful, I think. This particular dichotomy is explore versus execute.
Jacob Luetzow: Mm-hmm.
Jason Cohen: I have an article on this, you can go see it. the reason I care about this is because my observation is in these modalities, everything is different. So if I'm if I don't know what my product is, I need to explore. I'm reading Reddit. I'm seeing if this market's any good. I'm trying to see what people don't like about the competitors. when I interview potential customers, I'm trying to figure out like, how do you what's your life like? What's the pain? How do you solve this today? What language do you use for that? Why doesn't this competitor work? Why do you love this competitor? This is the questions I use when I'm exploring. What metrics
Jacob Luetzow: Mm-hmm.
Jason Cohen: would you give a team if if we were to pretend like we're at scale? What metrics would I give a team who's exploring? It wouldn't be points and tickets and scrum, right?
Jacob Luetzow: Yeah, that's fair. Yeah.
Jason Cohen: Maybe there are no metrics, or maybe it's more like just we have like learnings we've written up. Who even knows? But I don't know. But there's that there's that question. How would we know if we're making progress? Good question. But like, I don't know, there's some answers. And now let's take a different mentality of execute. Like you. My Carol is a person who has just experienced a This is is experiencing this, or they're wealth, they're they're setting up their wealth management systems, and so they're ready to go. So now the question isn't I gotta read lots of Reddit forums. The question is which of these ads is going to get them?
Jacob Luetzow: Mm-hmm.
Jason Cohen: That's A B testing. When they get to the home
Jacob Luetzow: Yeah.
Jason Cohen: page, what should I be saying? Should I use this word or that word? Or maybe for this ad I connected to this landing page and this ad, this land. Correct. That's all execute.
Jacob Luetzow: Mm-hmm.
Jason Cohen: We're we're we're we're we're A B testing the details, the tactics even. The strategy is set. You don't A B test a
strategy. You A B test the details and tactics of execution because you already know what's going on. What metrics would you use? Conversion rates, market sizes, retention rates. it's the it's the metrics of operations. That's what you use for execute. And
Jacob Luetzow: Right. Right.
Jason Cohen: maybe we do use scrum for some teams, or maybe not. Maybe we but somehow are we making progress as something like delivery of things, quality? I don't know, but like there's all this stuff. Some kinds of things that we would use to ask, are we doing good? You know, and the ex at the explore people, that's completely irrelevant stuff. So there's even more categories than that. But the point being, like another one is the mentality of the human beings doing the work. Now, if it's just a solo, this still actually can kind of make sense because that your own mentality can change depending on what you're doing. But let's again go back to saying, I'm a bigger company, how how would this look? Because it's just easier to visualize it.
Jacob Luetzow: Mm-hmm.
Jason Cohen: What kind of person likes working on a product where we don't even know what it is? We don't even know who it's
Jacob Luetzow: Yeah.
Jason Cohen: for. In fact, we might decide it's a bad idea and throw away the whole thing and work on a different product. There are people, yeah.
Jacob Luetzow: I think just listening to you, I think so many more people are stuck in experimental category than actually having a a end goal and a a real product with a with a focus.
Jason Cohen: There are people like founders who love the uncertainty and the and the discovery and of the exploration.
Jacob Luetzow: Yeah.
Jason Cohen: In fact, they're usually bad at execution and that starts failing once they get at some size of company. They really need other people to help.
Jacob Luetzow: Yeah.
Jason Cohen: Right? They tr have trouble switching into this other mode where it's bad to just do things and launch features and talk to and it's like very
Jacob Luetzow: Yeah.
Jason Cohen: disruptive at that point, right? but most people are the opposite. Most people say, I want to know what good looks like so that I can do it. Then I can get a pat on the head, maybe a raise. That's most people.
Jacob Luetzow: Yeah.
Jason Cohen: And by the way, I don't care about this company that much. So I want to do a good job just for like for my own pride. And then I go home. So, what do I need to do? And that's most people. And that's an
Jacob Luetzow: Yeah.
Jason Cohen: execute mindset, not curiosity and throwing yourself in exploring mindset, right? So, what kind of people do you need at the company depending on what modality that's in? Different kinds. That's pretty important
Jacob Luetzow: Mm-hmm. Right.
Jason Cohen: to know. So When I see, man, when it's A, all this stuff, and when it's B, all the stuff's different, then I think it's useful to say, ooh, this is a dichotomy worth calling out. These are words worth like let's put words on it. The actual words don't matter, but like let's just title it in a way that makes some kind of sense. Because since everything that happens, that everything about it, how I think about it, manage it, do it is different, I should know which mode I'm in. Cause what would the only dumb thing would be to mix them. I'll do this. no, don't do this. that can't be right to
Jacob Luetzow: Yeah.
Jason Cohen: do this, So like no. So so that's why it's useful. So so I when I think what is A B testing good for? I'm like, when I'm in execute mode, lots of things can be A B tested. It's what?
Jacob Luetzow: I feel personally attacked right now, by the way. I feel personally attacked. But it's like a good reality check, you know?
Jason Cohen: That's why these are useful. Yeah. So it, you know, when we say A B test, you A B test everything. I would think, yes, in execute mode, when you know the strategy, the ideas,
Jacob Luetzow: Yeah.
Jason Cohen: the important things, the major decisions, our prior major prioritization, then A-B testing is a tool for execution for executing better, for optimizing. Hell yeah,
Jacob Luetzow: Yeah.
Jason Cohen: go do that because none of us can just like write the best copy on day one. Like what? Like
Jacob Luetzow: Right.
Jason Cohen: this is ridiculous. Of course you need A B test. Of course, right? But that's the modality and the topics where you A B test. You don't A B test who you are, what your strengths and weaknesses are, what your strategy should be, what's a A C P. These are not things you A B test.
Jacob Luetzow: Yeah. And I think it's easy to think that you do. And that's probably
Jason Cohen: It is.
Jacob Luetzow: yeah.
Jason Cohen: Especially if you're an engineer and you're like, everything's a number and everything can be solved by engineering, which is total horseshit. Of course it's not. You think all important
Jacob Luetzow: Yeah. Yeah.
Jason Cohen: things about a company is a number? How about your relevance right now in the age of AI? That's not a number, but
Jacob Luetzow: Yeah.
Jason Cohen: might be existentially important, right?
Jacob Luetzow: Yes.
Jason Cohen: come on, this it's it's dumb.
Jacob Luetzow: It is. Good. I feel like I just went to school today.
Jason Cohen: That's the idea.
Jacob Luetzow: No, I love it. yeah, I think we can wrap this up. This has been a really good chat, and I really appreciate your time. And where where can people find your book if they're interested?
Jason Cohen: So my main website is asmartbear.com, Bear Like Animal, because this is two companies ago what I what the company name was. And so A Smart Bear has everything. It's got my podcast where I do this with founders. Founders come in, open their business up, and then me and a co-host,
Jacob Luetzow: Awesome.
Jason Cohen: we dive in and just try to solve something for two hours. And that's the whole podcast. No, no scripts or interviewing, just that.
Jacob Luetzow: That's cool.
Jason Cohen: So the link is there
for that. The link is there for the book. You can buy it from me, or it's all on Amazon with the Kindles and the audiobook and the paperback. And and only articles. I've been writing for 20 years on many topics, some of which we talked about today. And so those
Jacob Luetzow: Yeah.
Jason Cohen: are just there. And there's no ads and there's no like you can put in your email address if you want to get emailed when I post a new article, but that's it. It's not even a newsletter, it's just occasionally I write a new article. Like it's it's it's not annoying. Yeah, it's almost
Jacob Luetzow: I'm on your email list. It's not annoying. And I get annoyed really easily. So it's good.
Jason Cohen: nothing. Cause I I got it's just for the love of the game. I'm not a consultant. I don't selling courses. Okay, it'd like the it'd be nice to sell the book. Whatever. this is for the love of the game. So it's just all out there.
Jacob Luetzow: It's a good it's a good book. It's worth picking up, I think. It's like I I said this to Rob Walling too, like his SAS playbook. It's one of those
Jason Cohen: Mm-hmm.
Jacob Luetzow: books where you you're gonna just re-listen to it and you learn something new every time.
Jason Cohen: I love Rob. I wrote the forward to that book, the Sass Playbook. Yeah. Yeah, it's fantastic. Yeah.
Jacob Luetzow: did you? That's awesome. I I do probably remember that, but it's one of my favorite books. And yeah. And now this is going to be added to my collection because I've been really enjoying
Jason Cohen: So which one's better?
Jacob Luetzow: it. They're both equally amazing. I know I'm teasing too.
Jason Cohen: I I'm just kidding anyway.
Jacob Luetzow: yeah. No, I I think you have little like just nuggets that will change the way you think in both of them. And I think you you need them both. If you're trying to create a SaaS company or some sort of software product, whatever it may be, I think I think you need it. And in the age of AI, we have a lot more technical founders, I think, coming out because it's just easier to build. But
Jason Cohen: I don't think so.
Jacob Luetzow: it's not easier to sell.
Jason Cohen: Yeah. Actually I think we have more less technical foutures because now they can tr they can vibe code something. The the technical people we were we were always writing the code. That was always what we did.
Jacob Luetzow: That actually might make sense. Yeah. But that actually scares me more. That's true. Well, I've always said like from the technical standpoint, building the product was the easy part. And now it's just easier. And
Jason Cohen: Yeah, but but that that part's not different for us for the for engineers. of course it's different,
Jacob Luetzow: That's true. That is true.
Jason Cohen: but what I mean in terms of what what shape of company do we build and what's our challenge?
Jacob Luetzow: Yeah.
Jason Cohen: you know, that that is that is different because in fact, like maybe we could have gotten away with earlier saying, Well, it takes so much time to write code, I do need to spend a lot of time on that. And now it's like, no, you don't know.
Jacob Luetzow: No, no, my it's
Jason Cohen: No, you should be spending even less time on coding than before and even more time on marketing. So if anything it's it makes the Yeah. Yeah.
Jacob Luetzow: I know and I think that's even harder because now you wanna just add more you wanna be a Swiss Army knife of features because it's so easy to get in. you could just dream up an idea. Three
Jason Cohen: Mm-hmm.
Jacob Luetzow: AM, you kick off your prompt and then you do a PR review in the morning.
Jason Cohen: You absolutely can. And so can everyone else. And that's why that's not how you're gonna win or lose. Yeah. And again, if it's your own perfect personal project, never mind. Just build. It's fine.
Jacob Luetzow: Yeah, I agree. You gotta step it up. Yeah. Have at it. Yeah.
Jason Cohen: Have an open source project that nobody looks at, it's fine. I have those. It's fun. Like no judgment. No judgment. It's just if you want
Jacob Luetzow: Yeah. Me too. It is fun. It's a lot of fun. I
Jason Cohen: if you want a company where people pay you for stuff. well,
Jacob Luetzow: Yeah.
Jason Cohen: okay. That that that's what you said. So you've got to do something.
Jacob Luetzow: I know I was trying I was trying to wrap this up, but I'm I'm curious. weird like I'm very optimistic in this, but there's been a lot of negativity around like, you know, the death of SAS and like all this, because everyone can vibe.
Jason Cohen: That's completely not true.
Jacob Luetzow: Yeah, because you're always gonna pay for a product that works well. Like you're not gonna vibe code
Jason Cohen: Well yeah.
Jacob Luetzow: something. You're not gonna vibe code WP engine.
Jason Cohen: Well, so this is a really good question that probably deserves a whole nother episode because it's not a simple thing. We could do that. But I
Jacob Luetzow: Maybe I had to have you come back on. Yeah.
Jason Cohen: think the brief version is some kinds of SaaS will die. That makes sense.
Jacob Luetzow: Yeah, yes.
Jason Cohen: and then there's a lot of kinds which not only are not dying, but are thriving right now and growing faster
Jacob Luetzow: Mm-hmm.
Jason Cohen: than ever, including almost all public SaaS companies right now. Now,
Jacob Luetzow: Yeah. Yeah.
Jason Cohen: people talk about their valuation. Yeah, I understand. Meanwhile, HubSpot's at two billion in revenue growing 30% year over year? Are you out of your mind? And they're and you're saying that's dead? What in the world
Jacob Luetzow: Yeah.
Jason Cohen: are you talking about? So what it is is lazy speaking. It's it's pseudo-intellectual, like, this makes me look smart to say SAS is dead. But if you take one second
Jacob Luetzow: Yeah.
Jason Cohen: to look around, you go, wait, clearly some of it is, and a lot of it isn't. And then the only interesting question is: what kinds of SAS are not only
not dead, but more relevant. Whether it's more relevant or not more relevant, I don't know. Not only not dead, but thriving in the age of AI.
Jacob Luetzow: Mm-hmm.
Jason Cohen: What what is that? That's
Jacob Luetzow: Yeah.
Jason Cohen: the only interesting question. Not SAS is dead. If you say that, then you're
Jacob Luetzow: I agree.
Jason Cohen: not going to answer the question that's the right question, which is why
Jacob Luetzow: Yeah.
Jason Cohen: I call it pseudo-intellectual laziness. So that's the right question. Now there's a lot of interesting stuff on there. Rob Walling actually has a keynote that he did recently in Iceland at Tiny at a Microconf, which you can see on YouTube, which answer some of that and he has a new book. See now I'm now I'm not even now I'm promoting Rob's books. Rob has a new book on this topic how to build AI or sorry, how to build SaaS in the age of AI. And there's a lot of really good stuff that answers this.
Jacob Luetzow: Yeah.
Jason Cohen: I also have some ideas about this, which I actually have not published yet, because I'm like assembling my thoughts. I I write slowly because I try to make it really good and the writing and the thoughts so it's slow. Okay. But hopefully when it happens it's you know hopefully it's it's not drivel. So it's not AI either. but just a sneak peek would be there are certain characteristics of SaaS products where not only are we not vibe coding it, we wouldn't want to. And
Jacob Luetzow: Yeah.
Jason Cohen: so the more of these characteristics your SAS has, the more protected it is, and the less of it, maybe the less. So ad admitting this is half baked, but just throwing out it there anyway. So for example, if it's very important that it be secured, I know if I vibe code it ain't no secure, it's not secured. If I vibe code a
Jacob Luetzow: Right. Yeah.
Jason Cohen: thing and put it on the ed on the net, it's dead, it's dead in two seconds. No doubt. So if it's my accounting software or the key or where I store my keys for anything, like I don't want to vibe code that because the security is important, I need somebody else who's not only developing it with security in mind, but they're responsible for that. I could sue them if they weren't. and they're constantly like, I don't know, blog posting about how they're they're finding their own their own security issues and fixing them while I sleep. And a zero day thing came in and they they fixed that while I sleep. That's what I want to hear. I don't want to vibe code that.
Jacob Luetzow: Mm. Yeah. Yep. Yeah.
Jason Cohen: I think there's a dozen things like that, where the more of these it has, the less vibe coding makes any damn sense. So that's one. Collaboration
Jacob Luetzow: Yeah.
Jason Cohen: is one. Vive coding for myself, that works really well. Vive coding so that 200 people at the company can all use it and they know how to use it and has high uptime. It's easy for them to use and it integrates with everything. Oof. That's it's possible, but way harder. And now if I do that,
Jacob Luetzow: Mm-hmm.
Jason Cohen: except it's people around the world, not at my company, where I can just force them to do things. Now it's really hard. Not impossible.
Jacob Luetzow: Yeah.
Jason Cohen: Again, none of these are impossible. They're just like, that's another little brick that's like, ooh, now the wall's got a little taller. Cause yeah, that that is harder to vibe code than just the thing for me. Yeah.
Jacob Luetzow: Right.
Jason Cohen: And so there's all these bricks you can add. And the more of these bricks, you're like, yeah, I mean, chee whiz. So like right now, is anyone vibe coding?
Jacob Luetzow: Right. Larger the context, larger the problem, harder to vibe code too.
Jason Cohen: Are you vibe coding a new version control system or do you use Git? The answer is you use Git. And I know
Jacob Luetzow: Right, exactly. Yeah.
Jason Cohen: you're gonna tell me Anthropic did it. Okay, that's one and they didn't vibe code it, did they? Right? So it's not vibe coded. So,
Jacob Luetzow: Yeah.
Jason Cohen: and anyway, that's one. And no, everyone's fine with Git. And why? Because it works. The AIs know how use it. It it needs to be fast. It needs to always work
Jacob Luetzow: Mm-hmm.
Jason Cohen: correctly. It needs to integrate with everything else that we have. Et cetera, et cetera, et cetera. All these apps work with it. I mean, just it goes on and on why I want it. So of course I'm not
Jacob Luetzow: Right.
Jason Cohen: vibe coding a a a version control system. Okay. So that's an example, right? Isn't Git being
Jacob Luetzow: Yeah, no, it's a great. Yeah. The downtime of GitHub
Jason Cohen: Used more than ever now? Yes, it's never been used more than now. Okay, so
Jacob Luetzow: has increased tenfold because of their their traffic. Right. Yeah.
Jason Cohen: So that's bad, right? So okay. but but anyway, so so so there's another example. So so there's these bricks. And so you're right. If you have zero or one of those bricks, I agree maybe that sass is dead. But
Jacob Luetzow: Yes.
Jason Cohen: if you have seven of these bricks, I not only I disagree that that sass is dead. And so
Jacob Luetzow: I agree. I could I agree completely. And I'm glad that I feel like that's the overall conversation is people are just realizing if you build a good product that helps people, they're gonna keep paying for it.
Jason Cohen: if it it may be. If it doesn't have
Jacob Luetzow: If it solves a problem, I should say.
Jason Cohen: if it doesn't have those bricks, they might, but also some people will vibe code it anyway.
Jacob Luetzow: And those aren't your customers, right?
Jason Cohen: They're definitely not your customers. I definitely agree with that. I definitely agree that. So
Jacob Luetzow: Yeah.
Jason Cohen: you could take the market argument and say the only people vibe code, it sounds like a lot because you read Twitter, but very, very few people vibe code anything. We are. We're
Jacob Luetzow: And we're in our bubble of software engineers, right?
Jason Cohen: still in the early, early phases in the in the chasm. We're not even in early adopter phase. Early adopters
Jacob Luetzow: Yeah. Yeah.
Jason Cohen: would be 15% of the population of the US is vibe coding their own CRMs. And that's not true. We're still
Jacob Luetzow: Mm-hmm.
Jason Cohen: way in the end. It's a very exciting end. So, in terms of what's happening now, you're right. You could just say, if you're a vibe coder, you're just not my customer. And I would agree, that's right. That's a good strategy right now. Now, if you asked in five or 10 years, should this be my long-term strategy? I would say that depends on the bricks now. And maybe other things.
Jacob Luetzow: Yeah.
Jason Cohen: And maybe you're you don't like my analogy. That's okay. But like it's a different question of in five years, will it still be true? Or in five years will 15% be vibe coding their own person, let's say personal productivity tools. The answer is nobody knows. But I think it's more
Jacob Luetzow: Right.
Jason Cohen: reasonable to say in five years, maybe that's true. And so if you're a personal productivity tool with no bricks like I just defined, you're you're great right now. Are you great in five years? Nobody knows, but it's not a great bet. Right.
Jacob Luetzow: Right. Well, and I think that's common knowledge just in tech in general, right? Just going from like the dot-com boom through today, you know, see just shifts in things and tech companies close up overnight just because they didn't adapt.
Jason Cohen: Yeah, so the question is, i what is it that they did or didn't do? Right. That's the question.
Jacob Luetzow: Yeah, yeah, yeah.
Jason Cohen: also I don't know what's common knowledge and what isn't, to be honest. I know I'm in a bubble. I don't know what I don't know what it is. I've I think if you
Jacob Luetzow: Yeah.
Jason Cohen: if you talk to ten random people in America and ask them some of this stuff, I think you'd be shocked at how not common much knowledge is.
Jacob Luetzow: I know. And that's what I always try to tell myself too, is we're in a bubble of and so the doom and gloom is coming from a perspective of someone already capable of building these products.
Jason Cohen: Or the fear or they're not, but they so they don't the fear of the unknown, which is very valid. I think it's
Jacob Luetzow: Yeah, also true. Yeah.
Jason Cohen: completely rational to be to be fundamentally shaken and fearful of what AI will do, 'cause we don't know. I think
Jacob Luetzow: I think
Jason Cohen: it's a very v that's a very valid p position to hold.
Jacob Luetzow: It is. I think you gotta keep living, right? And keep
Jason Cohen: Yes.
Jacob Luetzow: an eye and know, be prepared for maybe software engineering isn't a job in five years. I don't know. But I
Jason Cohen: Or what it what is it? But the thing is we don't know because but my proof of that is if I rewind the clock two years and asked you what will software development be like in two years, you'd say various things or you'd say, I don't know, but like there's pundits who did say things. They did. So
Jacob Luetzow: Yeah.
Jason Cohen: we can go back and look. And guess what? It's like a coin flip. That's some of it maybe, some not. Like they don't know. We don't know.
Jacob Luetzow: Yeah. Yep.
Jason Cohen: Right? did anyone guess the anthropic would beat open AI in terms of revenue usage and going public first? No, no one thought that.
Jacob Luetzow: No.
Jason Cohen: Okay, this one guy's like I did, okay, right? But then you were wrong about that other stuff, right.
Jacob Luetzow: he was lucky, It's like the crypto millionaires, right?
Jason Cohen: Yeah, that's crypto was also how many people were wrong about crypto? Almost all of them, actually.
Jacob Luetzow: Yeah.
Jason Cohen: It's gonna replace money, no it didn't. It's gonna replace banking systems, no, it didn't. It's gonna get fast enough and and mainstream, no, it didn't. it's all
Jacob Luetzow: Yeah.
Jason Cohen: wrong, all of it. Now, look, I'm not judging, it's okay. we bet we make bets. Sometimes they're not right. whatever. I'm just
Jacob Luetzow: Right, for sure.
Jason Cohen: saying, so I'm not judging. What I'm doing rather is saying this is why. The answer is we don't know. Stop saying I know. You don't know. That's what I'm saying.
Jacob Luetzow: Yeah, we don't. I agree. I agree. All right. Well, Jason, this was awesome. thank you so much for coming on. And everyone, go buy the book. It's worth it. It's
Jason Cohen: Yeah.
Jacob Luetzow: it's well written and the audio book is read by Jason himself. So it's like listening to a podcast practically. It's awesome. Yeah.
Jason Cohen: It is. Thank you.
Jacob Luetzow: Yeah. Thank you so much. And everyone that was tuning in, have a good weekend.