Jason Cohen - Why AI is Different, the Paradox of Niching Down, and Writing as Legacy

Jonathan Stark (00:01.492)
Hello and welcome to Ditching Hourly. I'm Jonathan Stark. Today I am joined by a very special guest, Jason Cohen, the founder and CIO of WP Engine. Jason, welcome to the show.

Jason Cohen (00:12.526)
Thank you. My pleasure.

Jonathan Stark (00:15.06)
So what we are going to talk about primarily today is an article you recently wrote over at your blog. Do you call it a blog or an email list over at a smart beer.

Jason Cohen (00:23.534)
It's a blog. I started it 17 years ago when blogging was cool. Maybe it's cool again, like like jeans from the 90s. I don't know.

Jonathan Stark (00:30.612)
I don't know if I agree. But yes, that is fantastic. So maybe we'll even get around a little bit to how you're writing figures into your business and rest of your life. But first we want to talk about this AI article for folks who are listening, especially software developers who are listening, who might have an idea for a startup. Maybe it's something they're going to try and bootstrap or maybe going to try and get a little bit of funding. But I've talked, I have several people in my...

audience who I know have dipped their toes into this water. And I thought this was a really good, I thought the article was great at maybe setting expectations for what might be in store for someone who does that. So the title of the article, and I'll link to it in the show notes, is AI startups require new strategies. This time it's actually different and talks about the typical dynamics between startups and incumbents.

So maybe we could start off by defining what incumbents means in this context.

Jason Cohen (01:33.646)
Sure, so normally, so you see AI and you say, oh boy, a bunch of opportunities. That means as a person trying to build a new company, now is the time for stuff that has never existed before. So why not me? Especially because those old fuddy duddy large companies, they don't innovate, they're slow. And so this is a place where I can win. And of course there are gonna be a lot of AI startups that do win.

And also, of course, there's way too many of them, which means most of them will fail because that's how that goes. So by incumbent, what I mean is an existing large company. And so typically startups. So if you're a big, large company, you have a lot of advantages over a small company. In fact, when you start listing them, it starts to get real scary because they seem to have everything. They obviously have a ton more money. They have brand recognition. They have customers. And just saying they have customers, I mean, they can.

They can pull them. They can see the behavior of their customers. At WP Engine, my company, for example, we have 200 ,000 customers. So when we ask the customer base something, we can get thousands of responses and get statistically significant information. So just saying they have customers, I'm trying to emphasize this really is an advantage. Big companies have a sales team, they have a marketing team. In other words, if they make something, they can sell it really well. They have more engineering teams.

I mean, you just go on and on, right? Like big companies actually, it's scary how much advantages they have. And yet, of course, startups thrive because there are advantages startups have. So these are things like startups will take risks that larger companies will not take. Cause larger companies say things like, well, if we do anything, it's going to cost us many millions of dollars and we need to return on that. And so the thing needs to generate tens of millions of dollars or hundreds of millions of dollars for it to even matter in our revenue.

and for it to be worth our attention. So what data can we get about what markets that could be that big? And no startup thinks like that. And you don't need to think like that because you don't need to make the hundred million dollars, right? So you can take risks that larger companies won't take. You can address a market that doesn't exist yet. Whereas a large company says, where's the data? And there isn't any. So they don't do it. You can do things that don't scale, like be very personal about stuff that a large company can't do.

Jason Cohen (03:58.286)
You can go into a niche that's very profitable. That's too small for the large company or simply there. It would be too complex for them to have to attack many individual niches. So they don't. One of those things is leveraging new tech. So brand new tech comes out like AI and because it's new, it's automatically risky. So like as a large company, it's true. You shouldn't go look at the latest thing and then deploy it to your 200 ,000 customers. It's not going to be very robust.

It's going to have a lot of bugs. Maybe it won't even be supported in five years. Maybe it'll be hard to hire people that know that system because it's new. Like the big companies are correct and not leveraging new brand new technology for all those reasons. They have different needs, but for a startup, it's like, whatever. Yes, that's a risk, but I'll take it in order to have a product that's unique. So this isn't even all the things that, uh, that startups have in their advantage. In fact, another article I have lists even more reasons, more ways that startups have advantages.

Jonathan Stark (04:48.116)
What's, if I could just, yeah.

Jason Cohen (04:57.23)
And that's useful because you want to leverage those, right? You want to use those. That's those are your, those are the ways you can, you can win. So the, uh, so I know I spent a little bit of time on that, but hopefully it's useful to, to enumerate that because you're like, Oh, I should do that more. I should take advantage of that. I should make that part of my strategy because a big company can't stop me from doing that. That's nice. Anyway, the thing with AI is a lot of these things are not true. They're normally true, but in AI, they're not true. So.

Jonathan Stark (05:16.5)
Right. So off.

Jason Cohen (05:26.574)
For example, you would expect AI being a new tech, that means big existing companies, which I call incumbents, the people that are already in the market, that they won't use it because it's brand new technology. So they stay away. Oh boy, is that not happening with AI. They're all rushing into do AI. Their shareholders are forcing them to rush in and do stuff. Now I'm not saying it's good. I'm not saying they're making great products out of AI. Some are.

Jonathan Stark (05:35.54)
but not open AI, not open AI. You mean like...

Jason Cohen (05:53.294)
Like there's some instances Adobe's products by all measure seems like they're pretty good. Um, intercoms products looks like maybe they're pretty good. So some of them are pretty good. And a lot of them are just putting AI on something and it's not really that good. Right. But then again, a lot of the startups are destroying it AI stuff. And it's also not that good yet. Right. So whatever. The point being, um, incumbents are not seeding the AI stuff to startups. They're, they're rushing in. They're embracing it fast.

even though it's an uncertain market, even though no one knows how it'll unfold, even though it's brand new technology, they're like, yeah, I know I'm doing it anyways. And they're spending like historically large amounts of money on it. Like they're not just doing it as a side project. They're saying we're pivoting the whole company to AI, or at least they're spending tens of millions, hundreds of millions of dollars on it. And so as a startup, you go, wait a minute, hold on, that's what you're not supposed to do.

If that's the kind of effort they put in, plus they have the advantages like brand and customers and existing sales channels, that makes it a lot harder to compete against them. In fact, so yeah, just on the AI. So AI is not the same as previous kinds of tech innovations, whether that simple things like headless development or more of these C changes, like the rise of mobile and before that, the rise of the internet even.

Jonathan Stark (07:02.804)
based just on the AI. So like, AI is not a differentiator.

Jason Cohen (07:17.646)
That took a really long time for large companies to kind of embrace and know what to do with. There were decades, there was a decade in there with the internet, maybe two you could argue, that the incumbents didn't know what to do and there was huge space for startups and they existed. This is not happening with AI. Now again, and there's more things than just this, I'm just giving one example, but it's not that AI is terrible or something and there's no way to do it, it's not that pessimistic. But I would say,

Jonathan Stark (07:22.196)

Jonathan Stark (07:26.196)
Mm -hmm.

Jason Cohen (07:45.55)
Typical strategies that a startup would use May not work because the typical assumptions about what the advantage ours are for a startup are not true in AI like they usually are So it's not that you can't it's that the strategy may not be the same as usual You can't just say it's new. So I have a shot. It's actually not true this time

Jonathan Stark (08:06.676)
Do you have any, any, any guess as to why AI is being adopted so rapidly by compared to say something like blockchain? Why is, why does it seem like every fortune 500 companies jumping on AI, but they didn't jump on VR or something.

Jason Cohen (08:24.558)
Well, neither blockchain nor VR actually made it to the general population. Whereas OpenAI got, you know, insane mouths like right away, got to a billion dollars in revenue right away. And it seems very exciting and like it can change things. And there's certainly things now that it is changing. And there's no doubt that in the coming years, it'll change everything more. That's not the kind of sea change you can ignore.

With blockchain, I know there's blockchain enthusiasts that thought this will change everything. It actually hasn't changed anything. It hasn't changed financial markets at all. It hasn't changed currency markets at all. None of them have become interesting currencies. And no interesting software has been built on top of it that is in any way disruptive to existing software. There's a lot of projects, no examples of stuff that's actually disrupting. So.

Wait and see seems like a good thing. You know, if if if next year blockchain breaks out, that's great. But if actually if you blockchain is something like 15 or 20 years old now, but Bitcoin is you were you'd be right to not engage with it other than I guess you should have speculatively bought some just for the just for the just for the financial value. But if you built your

Jonathan Stark (09:20.436)

Jason Cohen (09:42.478)
company on blockchain, you made a mistake. So it was correct to stand by and wait and see if that thing unfolded. Same with VR. Perhaps VR is finally having a moment. Although when I look at people, they say everyone says they get the Apple thing and it's not that good. So again, not a bad idea to step back and wait and see whether that thing will work. And if so, when, whereas with AI, it's simply not true. There are lots of people out of work now.

because AI makes some people makes you could say everybody, but some people more effective and it's not necessary to have people and you know, that's going to accelerate as AI becomes better. And so you can't afford to not be there. Now that's the logical argument, but you could say it's simply how people feel. People seem to think AI will change everything. People seem to think blockchain won't, and it could be feelings. It could be.

Of course I don't know, but it's not too hard to paint a picture of one is actually affecting the world and the other one is not.

Jonathan Stark (10:43.636)
Okay, so if I'm a, say a solo startup or small startup bootstrapping, I'm probably going to employ AI in some way, but you're just saying it's not gonna give me any particular kind of leverage against the incumbents. So I would look at the other list of things that you had, being able to take bigger risks, being able to go after a smaller niche, not needing to make $100 million.

Jason Cohen (11:12.27)
Mm -hmm.

Jonathan Stark (11:12.692)
before I get out of bed, all those sorts of things.

Jason Cohen (11:16.014)
Yes, and there's more. You can have a personality that's more human and specific. Large companies are usually pretty bland. Some people really respond to that and the big company cannot really respond to that. It's never authentic at a big company. So that can matter. There's things like social media, having a presence like that, which a big company cannot stop you from. Big companies can do things like buy up all the advertisement or bid up all the advertisements that it's very expensive.

and therefore difficult for you to use that to grow. But they can't stop you from having a newsletter that's successful or a social media presence that people like. So that kind of stuff is good. You also might be able to use low cost. Now in general, low cost is among the least interesting strategic moves because of course it means you have less money and less profit in general. However, if you have a really low

Jonathan Stark (12:09.3)
Yeah, right, right.

Jason Cohen (12:13.262)
cost basis, then you could be low price and yet still profitable, especially again, if you pick a niche and so forth that where there's a reason why that works out. And a lot of times large company, in fact, even especially now in 2024, where companies are optimizing not just for growth, but also for profitability, in some cases, maybe even more so for profit profitability. If you have, if you're a lower price option,

Number one, other people trying to save money are interested in that more than they might otherwise. And two, the incumbent cannot match you because they have to keep the prices high to keep their profits up. So that becomes something they can't defend against. So low cost is another one. So there are still various advantages that you can have. But having AI is not one because of course everyone has AI. Another good way to think of it is this.

Jonathan Stark (12:56.052)

Jason Cohen (13:10.702)
People say stuff like, I'm going to make an AI thing, but AI is not a thing people want. It's not, AI is not a problem that people want to have solved. AI is a solution, not a problem. So in other words, people's problems are the same. Marketers still want high quality leads. Sales still wants to close more deals. Data teams want to analyze data. Like the things that people do are the same.

Jonathan Stark (13:14.548)
Mm -hmm.

Jonathan Stark (13:36.02)
Mm -hmm.

Jason Cohen (13:39.214)
The question is, can AI do that better? Maybe that means more efficiently, more effectively, with fewer people, with higher quality, faster. I mean, there's all kinds of ways what better could mean, right? But the question is, can AI solve the problem, or I guess take advantage of an opportunity, better than the, without AI? So AI is a means to an end. It's the solution side of the equation, not the problem.

Jonathan Stark (13:50.644)

Jason Cohen (14:07.566)
Nobody's saying like, I want an AI marketing thing. What they, what they might say is, can I write more content with the same number of people or the same amount of content with fewer people? And if AI says it gives me the, if AI means the answer is yes, then I'm interested in it, but only because it's achieving something I want it to achieve anyway. So rather than say an AI startup, it's exactly as you said, I, if I'm making a startup, it's because.

Jonathan Stark (14:11.412)

Jonathan Stark (14:28.788)

Jason Cohen (14:34.094)
I see a problem where I can create value for a customer. I see a special way to do it. It goes, it's in line with my personal strengths and abilities so that I can, I am the right person to make this thing and I will make it especially well. Here's the shape of the market that means people will actually buy from me and so forth, the usual stuff. And AI just becomes one of the things that drives the features that makes this a compelling product.

And it's exactly because the incumbents will also have AI. That just means, yes, of course you need to also, otherwise your product will just be worse. You don't want to do that. So sure. But that doesn't even really make it an AI company, quote unquote. Really the only AI companies are things like OpenAI or Google doing Bard or XAI. Those are AI companies. Those are in fact competing in what is clearly going to be an enormous market. But that's a totally different thing than saying I'm making a marketing tool.

Jonathan Stark (15:19.348)
Mm -hmm. Right.

Jason Cohen (15:30.446)
with AI.

Jonathan Stark (15:32.02)
Yeah, it's like the difference between being Oracle and having a relational database in your software. Yeah, you're not, you're not in the database market. So let's talk about that. What's your, what's your definition of a market? Because that is a poorly understood term in my audience.

Jason Cohen (15:36.974)
Right. Right.

Jason Cohen (15:47.95)
So I like the Jeffrey Moore version. This is now decades old. Jeffrey Moore is the guy who wrote Crossing the Chasm and in there he defines a market.

Jason Cohen (15:59.438)
he, if I'm paraphrasing, defines it as a group of customers who generally buy the same type of thing for the same types of reasons and talk to each other. So like if you go to a conference about something, those people are probably in a market together. So I mean, that's a little bit rough to say it that way. But so if I, so for example, if I, if

Jonathan Stark (16:11.092)
Okay, that's perfect.

Jonathan Stark (16:18.388)

Jonathan Stark (16:26.452)
And there's some desire to buy a solution.

Jason Cohen (16:30.19)
Yeah, so for example,

When Oracle decides what security tools to buy, they're looking at a certain set of vendors because they have all kinds of multinational, really sophisticated, high governance requirements. When a dentist office is looking to get quote unquote a secure website, they're looking at a completely different vendors for that. They also care about security as a topic.

but their requirements and their needs are different. Maybe they care about HIPAA and maybe Oracle doesn't, but maybe they also have very small requirements for the one website versus Oracle. So in that sense, although they're both buying security software, they're not one market. They're two different markets. The dentist might talk to, certainly other dentists, but they might talk to other medical people because they have similar kinds of needs for security for their website.

Jonathan Stark (17:17.908)
Mm -hmm.

Jason Cohen (17:29.742)
They may not talk to someone who has a thing on Shopify because that's sufficiently different that they don't really have the same needs and they don't look at the same products. So therefore they're in different markets. Now, of course, like this kind of makes sense, but obviously it's, it's easy to come up with cases that are fuzzy or on the line. Fine. That's, that's, it's not really the point to draw the line. So, so crisply, but to just say there's a, there's a set of customers who act similarly.

Jonathan Stark (17:38.9)

Jason Cohen (17:58.99)
in that they have not exactly the same needs, but let's say similar compatible needs, therefore looking at the same set of vendors or options. And they talk to each other like it, like they may go to the same conference or read the same blogs and so forth. They're in the same, they're kind of in the same social space, you might even say. And so, um, that defines a market. And then you can start asking things like how much money do people in that market usually budget? What kinds of needs do they have?

Where can I find them in marketing? Are they willing to adopt new tech if it helps their company succeed? Or are they fearful of that and they don't do that? They wait till the tech is very old and almost stodgy before they agree to adopt it. This is, of course, what I just said is the main topic of Crossing the Chasm in particular. But just one example of questions you would ask of a market that's useful and interesting. If I'm a startup using brand new tech, for example,

Jonathan Stark (18:30.324)
Mm -hmm.

Jonathan Stark (18:42.1)

Jonathan Stark (18:46.484)
Yeah, great book.

Jason Cohen (18:54.67)
then I want to target a market who does want to adopt new tech as well. I would not want to sell a brand new product with a brand new company, brand new tech to a stodgy old company that only buy stodgy old stuff because they're fearful. That would be the bad market for me. They're a good market for someone else like Oracle. So you can't say a market by itself is good or bad. Rather, does it match what I have to offer, what I can build?

Jonathan Stark (18:58.9)
Mm -hmm.

Jonathan Stark (19:07.7)
Mm -hmm.

Jason Cohen (19:22.478)
what my insights are, what my product is, what I have the capacity to make. I want a market that matches what I can do, then it's a good market for me.

Jonathan Stark (19:31.156)
Great. So this is a perfect segue into a sort of a bizarre situation. Right before we jumped on this call, a past student of mine and I were catching up and he sent me a couple of articles by guess who? Jason Cohen, just total coincidence. And one of them, I haven't even had time to read yet, but he summarized it to me as something that is both right up my wheelhouse, but right in my wheelhouse. But he said, you did a great job of describing.

why it's okay. So with that setup, I would love if you could talk about the article you wrote called Selling to Carol and why getting super specific in who you're marketing to, I don't want to put words in your mouth, but why getting super specific about who you're marketing to doesn't actually limit your opportunities or even your sales in the way that most people who haven't tried it might think.

Jason Cohen (20:24.206)
Yeah, so you've probably heard this advice before that you should identify this hyper specific person that you're selling to that's your perfect customer and then focus everything on them. The features, the language of the homepage, advertisement, the tone and voice of the company, everything. Focus on this one narrow segment and the narrower the better because it'll be so specific that you'll know just what to say.

And the argument, the quick argument for it is if there is a perfect customer for your product, like someone who not has the pain already uses the language that you use instantly understands that this product is for them already has the budget for it is looking to buy it. Um, every feature you have as a feature they absolutely need and every feature you lack as a feature they don't care about, or even they like that it's not there. Cause that would be clutter.

Jonathan Stark (21:20.98)
Mm -hmm. Yes, yes.

Jason Cohen (21:21.934)
As far as they're concerned, this, this, this imaginary person, the persona exists. It's sort of definitionally built around what you are, right? Kind of in reverse. Now, whether there are humans like that, I don't know. Maybe there's a market, maybe there's not, right? But, but this persona exists because you've built it in like exactly for you. Now, what I would argue is for many companies, if that perfect customer who I called Carol in this, in this, um,

Jonathan Stark (21:35.572)

Jason Cohen (21:52.238)
in this article, if that perfect customer hit the homepage, would they know that your product is right for them? Or are you so bad at communicating the product and so non -specific about it that Carol would not even realize that this is perfect? Even Carol wouldn't know. I would argue that's a bad homepage because when Carol does show up, you need to close her for sure because you deserve to and she wants to be closed because this is in fact perfect.

Jonathan Stark (22:09.236)

Jason Cohen (22:20.782)
So if your home page, your pricing page, your website, your ads, don't talk to her so that she doesn't know that you're definitely bad. Cause that means you're not talking to anybody. You're not convincing anybody. So let's suppose you agree. Maybe you don't, but let's just for the sake of argument, let's suppose you agree that sounds powerful. There's a really obvious thing to worry about, which is, okay, let's say I do everything you said. And let's say Carol, the Carol's of the world sign up.

Jonathan Stark (22:31.7)
Mm -hmm.

Jason Cohen (22:48.782)
The problem is that's this tiny little group and I want to sell to more people than that. And I'm afraid that if I just speak to Carol, just use that language, just, you know, characterize them and so forth, then everyone else I'll push away. I'll miss, I'll lose. And so basically my company will, I might have 10 Carols and they're happy, but then, but I need to get a thousand customers and then 10 ,000 not have 10. So this is limiting is the worry.

Jonathan Stark (23:15.124)
Mm -hmm.

Jason Cohen (23:17.23)
And it's a very rational thing to worry about, by the way, like this, this is not a, um, this is not a cognitive bias or something. It's a very rational thing. Um, my experience, however, is it does not work like that. It is not true that just because you've got a, you're shooting at the center of the bullseye that all the other rings of the bullseye don't come along. My, my experience is that they do. And it's true. There's 10 times as many people that are similar to Carol, but not the same. And another 10 times as many that are.

even still again directionally Carol, but different and so forth. Those people exist and you might think of rings in a bullseye. And it turns out they'll buy anyway. Your specificity when you speak to Carol is compelling. Like so many websites are just generic and boring. Just the fact that you have an opinion, you're specific and you're saying something is already interesting. It stands out. It's curious.

And the fact is when we buy products, we understand, even if it's not conscious, we understand that it'll have some things I want and some things I don't want. And there's some things I just don't care about. That's just natural. There's no such thing as a car. That's exactly what I want or software product. That's exactly what I want in when a business is making a intentional buying decision. Then what I just said is actually explicit, right? They'll have a process and they'll, they'll mark things this way. It's explicit.

But for most people who don't do that, they're still thinking that way. And so for every carol where it's perfect, there's 10 times, maybe even 100 times as many people who generally say, yeah, these features you have, I like it. Maybe some I don't care about, but OK, it's fine. It's there. And then these weaknesses that you have, I don't care that much. Like, yeah, I prefer if that weakness wasn't there. But it's not a deal breaker. So although it's not the perfect carol thing,

They're sort of in the zone and that they generally are OK with the strengths and weaknesses. And then as you go further out, yeah, there's people that go, OK, I like these two features, but I really wish you had these two other things that you don't. So it's really 50 50 and I'm on the fence. OK, you'll get 50 percent of those because they're on the fence. Like you're not the perfect product, but no other products are perfect. And if some other part is absolutely perfect, if they are the carol of another product, fine.

Jonathan Stark (25:35.028)
Oh, that's such a great way to put it. Yeah, yeah.

Jason Cohen (25:44.174)
But we just said, we just admitted there's not that many of those. So that means the vast majority of people are in this zone where they're like, some of this I like, some of this I wish was different, and so is all the other products. And so I'm still gonna buy something. And a lot of them will buy from you. And because you were so specific about it, it's actually easier for them to buy from you because they understand what they're getting. And I wanna underscore that with something really interesting that's not in the article. So here's a little bonus.

Jonathan Stark (26:08.052)
Yeah, they get their head around it.

Jason Cohen (26:13.742)
podcast thing. A couple of decades ago, I don't want to get into too much detail because just for time. But let's just say there's a company in Austin who ran product reviews for big companies. And I mean, like they ran them for Walmart and Target, stuff like that, right? So huge product review set of property data. Here's one of the things they found. When you talked about the negative aspects of a product, people bought the product more often.

Jonathan Stark (26:13.972)

Jason Cohen (26:43.598)
and returns were lower.

Jonathan Stark (26:46.58)
Uh, well, okay. Why do you, why did they think that was?

Jason Cohen (26:47.502)
So sales go up, returns go down. And in software, returns are annoying. In hardware or stuff, returns are death. Because one return, especially if it's broken or it never comes back, you might have to sell 10 more things to make up the profit margin of the loss of that one thing. It's really, really bad when you have returns. So the fact that returns went down is like an enormous profit maker. Plus sales went up.

And why do sales go up and returns go down when you know the bad stuff? Hopefully it's obvious. It's because you knew what the problems were and you bought anyway. Like if you see that and you go, okay, I understand there's good things. I understand what the bad things are. So I buy it. And then when it happens, including the bad things, you say, I expected that you don't say what this is crap and send it back. And so it's, that is real data from.

Thousands and probably hundreds of thousands of real products that say when you're explicit about even the weaknesses actually sales go up and Returns go down and so that's just a very that's a very long way I should probably put that in the article, but it's a very long way of saying You're speaking to Carol because you know how to do that and you certainly want to win Carol when Carol shows up Or anyone who's Carol adjacent, right? But even in the wide world a lot of people

Jonathan Stark (27:53.14)

Jonathan Stark (28:01.524)

Jason Cohen (28:06.798)
will take that set of trade -offs anyway. And by being specific, you are compelling. And in fact, you'll make people more comfortable to make this set of, whatever set of trade -offs are in their head.

Jonathan Stark (28:17.396)
Yeah, that's fantastic. I had a quick story. My audience knows I have a book called Hourly Billing is Nuts. I wrote it specifically for software developers when I was coming out of software consulting and everybody that read it, you know, if they weren't a software developer, you know, like my dad said, you know, this would be great for lawyers. This would be, you should write this less specifically. And I said, I think lawyers can figure it out, you know, configure it out, right? Because it would have really blurred my language and

it would have made it harder to understand. Even if you're not in software and you read that book, I think it's easier to understand than if I had made it more generic and abstract up a level or two to try and use words that for customer or guest or client or student, you know, and if I didn't, if I wasn't specific about, um,

those languages in that domain, people would just not understand it where if Carol is my ICP for that book, then, but she's not, well, someone's Carol adjacent is not a software developer, maybe they're a graphic designer, they can translate it in their own head because the specificity was there in the first place, even though it wasn't directed at them. It was a really interesting outcome. It took me a little bit by surprise. So.

Let's we're I see we're running out of time here. So I want I wanted to kind of wrap up on a sort of a meta question about your blog and I've been on your mailing list for a little while you send out I think it's two emails a month and They're good and they're pretty long. They're long ish not super long not too long, but like a nine ten minute read and I'm curious

Why do you do that? What is your motivation to continue doing it? I can guess a lot of things, but I'm curious what yours are.

Jason Cohen (30:08.462)
It's always been because I enjoyed writing. I enjoy the attention and adulation that comes when people like it. Um, and I never try to make money on it. There's no ads never has been. There's no, I don't know. There's no dollar subscription. I'm not selling any products. So it's so, so, so it's hopefully obvious that it's not for money because I don't have to make money at it.

I can simply make whatever I think is the best I can do. So in the process, I get to figure out what I think. And often I get halfway through a draft and realize, actually, this isn't so good, or this is wrong, or this isn't clear, I don't know, which is very valuable for me. But sometimes it does make complete sense. And then I've gotten some idea, some thought, and I've worked it out in such a way, of course, that others can share that they may agree or not.

I always think if I can make almost like the Carol thing, if I can make a really good case for it, then even if you completely disagree, I've still helped you a lot because in disagreeing with a well reasoned thing, that means you probably have worked out in much better detail exactly what you think and why, you know, it's kind of forced you to improve your argument, you might say. So even if you 180 degrees, if you're facing exactly the opposite direction,

Jonathan Stark (31:23.859)
Mm -hmm.

Jason Cohen (31:32.91)
still probably helped you be really smart about that, which is great. So I think that's fun. And probably the best response I ever get is sometimes, especially when I have things that are more, uh, personal, psychological, emotional, which I do, it's not all, um, it's not all, uh, frameworks and stuff like Carol. Um, then I sometimes get like a, I need it. I really needed this today. That's a great, that to me, that's a success, even one email. So I have about 50 ,000.

Jonathan Stark (31:35.892)

Jason Cohen (32:02.51)
newsletter subscribers if there's one that says I needed to hear this today for an article that is pretty great that's that's touching another person in a very positive constructive way that's pretty cool so that is exciting and yeah some of that's ego for sure like it's nice to have lots of readers it's nice to get nice notes and that's that's egotistical for sure and I don't I don't mind that so what as long as it's helpful it's okay you know

Jonathan Stark (32:23.316)
Mm -hmm.

Jonathan Stark (32:28.34)

Jason Cohen (32:31.278)
but I have the luxury that I don't have to make money on it so I can just do whatever I think is best. I have some recent thoughts that I haven't written about yet about that with the very brief version simply being maybe this is a way to have a legacy that's not a legacy that means my name is on a building or something or even something named after me. I've never named anything after me. I always name things after...

something else, some other title about what it is or somebody or I name it after somebody else or something like that. So not about my name, but if you have an idea and it's sticky and people like it and it gets transmitted and retransmitted and not even necessarily what I wrote, but I mean other people take up what I write. They add their own things to it. They make their own exposition and it continues to go like that. Does divorced from me even, um, in that way you have like almost like an ideal legacy.

That's a positive thing. It's not a legacy about me, the name, but it's ideas that then continue. That is a pretty neat idea for so you might say leaving a legacy quote unquote, that's apart from is in fact apart from ego or apart from money or accumulating power or something like this. It's none of those things, but it is leaving some kind of imprint or some kind of contribution to how people think. Uh, even if you can't measure that and you can't really say, right.

Jonathan Stark (33:34.484)

Jonathan Stark (33:49.556)

Jason Cohen (34:00.174)
That's okay. You drop a pebble in the lake, the waves go out, then you can't see them anymore. How much of an impact do they have? Who knows? But it's kind of neat to think of them continuing on and on. That's something nice about that. So that's not a fully formed thought, but that's kind of directionally how I've been thinking about setting ideas out in the world.

Jonathan Stark (34:20.852)

Jonathan Stark (34:24.5)
That's great. Very well said. Well, that's a great place to leave it. Thanks so much for joining me, Jason.

Jason Cohen (34:30.446)
Thanks for having me. It's always fun to talk about these things and there's an infinite number of things to talk about, so maybe we'll do it again.

Jonathan Stark (34:36.596)
Yeah, great. Yeah, we're on the same page with tons of things, but we have such different backgrounds. It fascinates me when two people kind of arrive at the same conclusions about so many things, even though they're from different worlds. Well, so where can people go to find out more about what you're doing? What's the best place?

Jason Cohen (34:55.854)
So asmartbear .com, bear like the animal, that's it, that's where all these articles are. And I'm on Twitter and threads and stuff, the links are all there. I'm asmartbear on Twitter, for example.

Jonathan Stark (35:09.236)
Fantastic. All right. Well, I'll put all that in the show notes. Thanks again for joining me. All right, folks, that's it for this week. I'm Jonathan Stark and I hope you join me again next time on Ditching Hourly. Bye.

Jason Cohen (35:12.526)
Thank you.

Creators and Guests

Jonathan Stark
Jonathan Stark
The Ditching Hourly Guy • Author of Hourly Billing Is Nuts • Former software developer on a mission to rid the world of hourly building
Jason Cohen
Jason Cohen
Keyword, buzzword, half-truth, adjective, hey look at me! (founder of two unicorns: https://t.co/Cc4OvZx0T9, https://t.co/JTEGCe7Zq6)
Jason Cohen - Why AI is Different, the Paradox of Niching Down, and Writing as Legacy
Broadcast by