$100M+/year founder Jason Cohen reveals a brilliant startup idea built on top of Twitter

I’m joined by Jason Cohen, a four-time entrepreneur and current Chief Innovation Officer of WP Engine, to discuss his three best startup ideas: a tool for optimizing social media replies, a simpler, lower price segment, and helping people make practical, simple strategies.

🚀  My FREE 5 day email course to learn how to build a business of the future using the ACP funnel:

https://www.communityempire.co/

🎯 To build your own portfolio businesses powered by community you might enjoy my membership.

You'll get my full course with all my secrets on building businesses, peer-groups to keep you accountable, business ideas every single month and more!

Spots are limited.

https://www.skool.com/community-empire-6150/about

📬 Join my free newsletter to get weekly startup insights for free:

https://www.gregisenberg.com

70,000+ people are already subscribed.

FIND ME ON SOCIAL

X/Twitter: https://twitter.com/gregisenberg

Instagram: https://instagram.com/gregisenberg/

LinkedIn: https://www.linkedin.com/in/gisenberg/

FIND JASON ON SOCIAL

X/Twitter: https://x.com/asmartbear
LinkedIn: https://www.linkedin.com/in/jasoncohen/

Blog: https://longform.asmartbear.com/

Creators & Guests

Host

GREG ISENBERG

I build internet communities and products for them. CEO: @latecheckoutplz, we're behind companies like @youneedarobot @boringmarketer @dispatchdesign etc.

What is The Startup Ideas Podcast?

Get your creative juices flowing with The Startup Ideas Podcast. Published twice a week, we bring you free startup ideas to inspire your next venture. Hosted by Greg Isenberg, CEO of Late Checkout and former advisor to Reddit and TikTok. Subscribe so you don't miss out.

For more startup ideas, we created a database of 30+ startup ideas you can take at https://gregisenberg.com/30startupideas

Jason Cohen [00:00:00]:
To me, strategy means deciding how you're going to win. That means things such as my advantage or here's my strengths or assets, here's how I'm leveraging those specific things. I think in retrospect, when a company has a good strategy, it's easy to say what it was. So you look at Netflix and you're like, oh, DVD's to the home, so you don't have to do the walk of death. That was smart. And then when they realized streaming is even better because it's immediate, I mean, that was good. So to me, that is a strategy that's like, this is the critical idea that sets them apart. That is unique.

Jason Cohen [00:00:28]:
It's hard to follow. I would have a vision like helping thousands of companies succeed because they have the right strategy.

Greg Isenberg [00:00:36]:
Jason Cohen, welcome to the Startup ideas podcast. Great to have you here. And you've got a bunch of ideas. You're an idea machine yourself, and I just, I'm excited to dive in with you.

Jason Cohen [00:00:50]:
Yeah, you know, I think it's funny, I don't know that there's an area in the world that you can't make a startup in, especially if you have aspirations of building a profitable business for a few people and not a unicorn that employs thousands of people and makes billions of dollars. Those opportunities are fewer, although once again, you can probably disrupt anything with a sufficiently different idea, but especially if you have more modest goals. And I don't mean that in a bad way, I mean that in simply a magnitude way. It's hard to imagine an area that you can't make a business in. So to me, like the biggest thing there is a large and growing market because that means people are spending money already. More and more are doing that. That means there is a problem to be solved or something people want done. And it also means niches will constantly be blooming because as it expands, like, there's no way that the larger players can do a perfect job in all of the niches.

Jason Cohen [00:01:50]:
That's just never true. And so there's opportunities for, again, if you don't want to make a billion dollar company, then, well, there are niches that are too small for the big people to certainly target, maybe even notice, and then you can fit right in there. So I think as long as it's a large and growing market, you've already taken some of the market risk out and you're already on the right path.

Greg Isenberg [00:02:13]:
All right, so you have some nuggets of some ideas I want to get to. Which one do you want to start?

Jason Cohen [00:02:17]:
With first, whatever you want.

Greg Isenberg [00:02:22]:
How about this social media system that you have that you're thinking of maybe productizing? What's this?

Jason Cohen [00:02:31]:
No, I'm not thinking of productizing it. Nope.

Greg Isenberg [00:02:34]:
All right, well, tell us about the system.

Jason Cohen [00:02:40]:
I've used a lot of these systems where you buffer up stuff and sends out, and that's a good idea to sense the fat on a schedule. I just find that there's so much more to social media than that, than the broadcast period. And I just haven't found tools that work well for me in processing the other side of that. So, like on Twitter, well, actually, on any of them, you should be interacting with people. And so how can you do that quickly, almost like an inbox that you can move through very quickly. Some tools have things like that. There's other rules on Twitter. Like if people have a follower to following ratio that's really out of whack and you interact with them, it actually lowers your reputation and reach.

Jason Cohen [00:03:22]:
So you don't want to do that. So the tool should know that and kind of show you the things that you should react to. Maybe you should even sort it not in order of time, but in order of maybe who it is that you're interacting with. And then there's the reverse. You know, you want to. Another, another thing they say on Twitter and other things is, well, there's probably people your size, maybe a little smaller, maybe a little bigger, maybe a lot bigger, that's in your zone that you want to interact with. You want to know them, you want to, et cetera. And so, for example, if they post something and they have a huge following, and you have a very cogent, useful, great reply that can get a lot of likes, and people might even follow you because they saw you that way.

Jason Cohen [00:04:02]:
And that's one of the things people say to do. I haven't really seen tools that are that good at it, but it's not even hard to do that with Twitter itself. So, like, if you go to tweetdeck, you can set up a search using a Twitter list in which you can, you know, keep updated with this set of people. Right. That's very easy to do because you can do it on all the clients and then say, search in that list for people who have tweeted and it's not a retweet. And, you know, you might have conditions like that to get to these items. Now, the trick to me of getting the comments to work is if they just posted it, then if your comment gets in there, it might start accumulating likes and snowball so that even though that thing might live for a day or 12 hours or whatever it does, perhaps you have a much higher chance of getting attention and getting this effect that I'm talking about if you were to post it immediately after their post, not if you wait 8 hours. So you can add to that filter how long it was since the tweet.

Jason Cohen [00:05:02]:
So I had like 20 minutes. Like, if it's longer than 20 minutes, I'm just not even going to look. And so that means I just go there. And there might be just two or three things there, tops at any given time, but they're highly likely to be good. And I got this idea from a related but different story where the CEO of Google was saying how it's impossible for him to use Gmail because it's just full of stuff. It's a disaster area. So he doesn't really use it, he says. But what he does is he goes in occasionally and he just answers the email that's at the top, meaning the most recent one.

Jason Cohen [00:05:36]:
And he says what he likes about this is most people get exactly the experience they expect, which is they email the CEO of Google. And of course he didn't respond. Like, why would he respond, right? But occasionally someone gets an answer in like 40 seconds and he's imagining them thinking like, I emailed the CEO Google, you answer your 40 seconds. This is a beast. Who is this person? How can this be? So it's like, yeah, that's hilarious. So that's sort of the application of that to Twitter is what I'm describing. So I think that's pretty funny. Another thing I do, and again, I have my own system for this that I've made that I'm not interested in turning into a product.

Jason Cohen [00:06:14]:
And some of it is kind of hacky with stuff like using puppet to control the browser and stuff like that. So that doesn't really scale. So another thing I like to do is even you spend a lot of time replying, like how much time do you have in your day? Probably this is too much time to invent new stuff for your cue, but also to have these really great replies that are actually good. Don't just say, yes, I agree. Something really good that's worthy of other people sharing or liking. That takes effort. So what I do is when I make that reply, I also have it. And it's like that one of those kind of, in my system, it makes a copy of that into a draft queue.

Jason Cohen [00:07:00]:
So then at other times I'll go through that, and I'll rewrite them to be standalone tweets with whatever the concept was. But in that mental mode, I don't have to come up with ideas. I've already done that. Even with the way it's worded and stuff, I'm just editing for purpose, so I can just go through very quickly. So I've spent the time having a good idea and even wording it well in the reply. But then I sort of reuse it, you might say, for the queue. So that means my queue is very full, very long, so I never worry about having to come up with an idea. Oh, the other thing is those are scheduled at least three months later, so that it's not really possible for someone to say, didn't you just say that? Because it's sufficiently removed from everyone's mind, including my own, it might as well be a brand new idea.

Jason Cohen [00:07:50]:
And so that's a way to essentially double the reuse, you might say the efforts that I do have good replies, and yet I'm not wasting that time in a sense, because I'm really filling my own queue. Also, that's another thing that you could easily make the product do that that doesn't take so much special effort.

Greg Isenberg [00:08:10]:
So I think what you're saying is really interesting, which is you've built a system for how you manage social media. And whenever you've built a system, there's an opportunity to productize that system. And sometimes it doesn't even need to be your own system, so you can pick off on how other people are creating their sausage, so to speak, like, especially in social, because everything, you know, a lot of, not everything, but a lot of it's public, you're able to see that, oh, Reg Eisenberg often replies within five minutes to these people. So I think it's really smart. Like that framework for you have a system, how do you productize it? I really like that. The other thing I really like about this idea is how do you make the UI chat? So, for example, you were talking about having replies come into this gmail like experience. I would totally use something like that if it told me, here's the ten people you should respond to right now. And it could also give me some ideas based on what I've said in the past, and because there's real tangible value associated with that, because replies are now being optimized in algorithms, I would totally pay for that.

Greg Isenberg [00:09:26]:
So I think I love this idea. And it would be also really cool if I was able to subscribe to not just like Jason Cohen's system, but Greg Eisenberg's system or other people's systems. And the UI would change based on that.

Jason Cohen [00:09:40]:
Yeah, it could be the different systems are good in different ways. I don't know. I would start with one. I think a couple other things. One is in my system. I don't say here's ten. I just show myself one at a time. So I'm completely focused.

Jason Cohen [00:09:57]:
It might be ten, it might be one, it might be zero, whatever. I'm not trying to waste time here and I'm not trying to make ten replies that may be wrong. Maybe it's better to make more replies and it's just prioritized. But that's how I do it. I want to optimize my time rather than how many things I write. I don't use AI, but I agree that it makes sense. It's really obvious. Use AI, train it on your previous responses, blah blah blah, your style.

Jason Cohen [00:10:25]:
We all know AI is only so good at replacing your style or reproducing your style and so on. Maybe there's some models that are better at that than others and so forth. A big expert in that. So I don't use that because I want it to be my own voice for real, and my own ideas. But I can absolutely see the idea of using AI to draft and then you edit it. And this obviously could save you time. It also depends on what your goal is. If your goal is just to accumulate followers because you're going to sell them a course or something like that, then I would use more AI and blah, blah, blah because it's not really about you, it's not even really about your ideas.

Jason Cohen [00:10:57]:
You're just trying to work the system and that's fine. Again, I'm not saying that it has to be bad, it's marketing. That's not what I'm doing. I'm not interested in that. I'm not selling anything. I don't even run ads on my website. I actually want to have my ideas and my style and interact with real people and that's it. And it is nice to have more followers.

Jason Cohen [00:11:17]:
I like that. But my goal is not to accumulate newsletter subscribers to sell them stuff. My goal is, I suppose, to accumulate people who are actually interested because I write unique stuff. The end. That's the reason. And so hence, like, I'm optimizing for something else and I'm not interested in AI making a response because it's just not, because I'm not just accumulating it. I really want it to be an idea that I had. That's the point for me.

Jason Cohen [00:11:44]:
The other thing that would be really interesting is I'm not trying to call anyone out, but these systems that exist already, of course there's a lot of them. They always have a newsletter or something that gives you tips and there's random tips, and it's not clear to me that those tips are really validated. They just say that. But is there data that shows, have they a b tested stuff? And it always occurs to me like, you have thousands, maybe tens of thousands, maybe hundreds of thousands of customers who've done stuff. Can you actually see what works better? Could you a b test stuff within one account or across many accounts to see what works better or other kind of data? Like maybe you paid for the huge Twitter firehose. You have a whole lot of data and you can actually analyze it and know more. I know there's looking at the algorithm itself, I get that. But that's different than saying stuff like what's viral? Like, oh, you got to use a hook and then this, and then that's.

Jason Cohen [00:12:38]:
The Twitter algorithm won't tell you that part. Right. I have a feeling it's all of those things are not based on data, that they're really just based on people saying stuff and wanting to fill a newsletter with things that seem plausible. But it seems to me like there should be data, and if there's not, who really cares what they say in the newsletter? So I think there's probably some kind of opportunity in there to say, no, this is actually a data based system, and the product itself is testing things with us, with all of us, so that the product can improve for you and actually help you. Like I said, respond within 20 minutes. Maybe it should be ten, maybe it can be an hour, maybe it depends on other factors, the bigger the account, blah, blah, blah, or I don't know. I don't know. That's the point.

Jason Cohen [00:13:23]:
I don't know. Maybe, maybe posting it, a lot of these things post at regular times. Well, maybe it's, maybe you should experiment with lots of times. That seems obvious to me, but the tools don't want to do that, so my system doesn't do that. I guess it could, but I haven't put in the effort. And I'm also just one person in one account, which probably isn't enough data to make many conclusions about anyway. This is the point is like, I think the question of how to, quote unquote, how to do social media, well, I don't think is very data based, and you could potentially argue it's not possible to be data based because it's too complicated and there's too much that the noise of what happens online overwhelms any signal you could find, especially with things like, this thing went viral and that's so unusual that you can't really analyze it. So if it's impossible, okay, I would agree with that, too.

Jason Cohen [00:14:14]:
Like, I can understand if it's impossible. I just don't feel like people are trying first and then concluding it's impossible. I think a product that really did, and even if it was a slight edge or something, that might be a reason to use that product over the many other products there are for social.

Greg Isenberg [00:14:31]:
The idea that if I create this post, I tweet this thing and I could see a projected amount of types of replies and types, no, I don't.

Jason Cohen [00:14:41]:
Think you could do that.

Greg Isenberg [00:14:43]:
That would be like, who cares?

Jason Cohen [00:14:44]:
I don't need to project. Like I want to know. Like, one thing I think is true is Mondays and Tuesdays I get a lot more activity on Twitter, and it falls off pretty steeply on Thursday. And Fridays very suck, very quiet, maybe even quieter than Saturday, something like that. I'm not sure, though. I probably do have the data to figure that out. But I think it's something like that. If that kind of thing is true, you'd want to put more stuff out and engage more and blah, blah blah on Monday and Tuesday, and maybe save some of your stuff and not do the same thing on Fridays.

Jason Cohen [00:15:20]:
Maybe do some, but not a lot. Or, I don't know. Maybe it depends on how deep your queue is. If you have a huge queue, maybe go ahead and spit it out. But if you don't, like, there might be because you're behind or whatever, or just don't want to spend as much time on it. Maybe it can be more judicious and where it puts it out, I haven't seen any of the systems do that. They could, of course, they could just take the idea, which is fine, but that's another one. So I just think it would be nice to take all these kinds of ideas and put them to the test.

Jason Cohen [00:15:50]:
And it might be different for different accounts. I don't know that you have to say, like, oh, if you word it this way, you'll get more likes. If that's actually possible. That's great. That sounds so specific that I'd be shocked. But maybe. Maybe you could look at an account and say, look over the last year, whenever you talk about these topics, you just get more. You get more reach.

Jason Cohen [00:16:11]:
I mean, that right there would be very interesting.

Greg Isenberg [00:16:13]:
Yeah, I mean, I think there's a lot of people who would love to use a product like that, even if it's just directionally correct. If you spend a ton of time creating content, and especially as the algorithms are becoming more for you, page versus follower base, where followers are becoming less and less valuable, and the type of stuff that's getting seen on people's feeds has to be. It's similar to how TikTok works. You want to make sure that. I'll tell you, who cares? I care. When I create content, I want the, I want a lot of people who signed up to get my content, to see my content. So if I know that if I change this word or if I put an image or if I post it in the AM versus PM, that I'll get 20% higher lift with 80% accuracy, that product itself, I would imagine there's a lot of people like me who would love to try that out.

Jason Cohen [00:17:15]:
Yeah. And maybe those numbers are not possible, like that kind of accuracy is. I wouldn't be surprised if that's not possible. But even if it was a small edge, again, I think social media is like that where, no, I'm not 80% sure of something. And know that one word doesn't, you know, doesn't always work. But if you just make it 60 40 instead of 50 50, or, you know, you just push it a little bit more, that's enough. Like, or that's, that may be all there is to do, but the current tools don't even do that. So.

Greg Isenberg [00:17:48]:
Totally. So lots of good ideas there. You can unbundle this idea in a ton of different ways and build something cool. Thank you for that. Your second idea, a simpler, lower price segment, I'd love for you to explain that. And for folks who don't know what segment is, could you also explain that? Yeah.

Jason Cohen [00:18:08]:
So segment is a tool where I guess you could call it clearinghouse of data. You take all the corporate data about your marketing, your websites, your product, even internal corporate data, potentially your customer base and internal information, and you can put it all together into segment. You can send it all in. Some people call that a data lake. Then segment, you can connect it all. This all takes extra effort. This is not simplistic, this is hard to do, but you can do it. Segment can connect it.

Jason Cohen [00:18:43]:
Also, some notion of a customer that, that can flow through from the website to the product, to back end systems. And then, of course, you can do stuff with it. And besides reporting, you can take action. So you might do things like, I'm going to segment all my customers for high value customers. People have been here a long time, people who just got here, but they're using a lot of tech support, people who aren't customers yet, but have other attributes that I want to segment based on, whether they've used certain features or not, whether they're certain place where they've gotten a certain distance through the onboarding process or something. You know, there's all kinds of ways that you could segment stuff and then act on the stuff, right? So maybe, oh, you've been in this segment for three days, so I'm going to send you this email. Like, there's all kinds of stuff you could do to act on that, depending on the circumstance. There's even stuff like connecting it with retargeting systems so that you spend money retargeting only certain segments of customers.

Jason Cohen [00:19:41]:
So that might be like some kind of marketing spend optimization type of thing. There's all sorts of all manner of stuff. And if you want to do everything I just said, you probably just need segment or one of their competitors. That's it what I think, and that's fine. It's a really interesting product. So when I say simpler, lower price, I mean segment is expensive and it's not simple, but it does all this stuff. So it seems to me like there's a lot of SMBs who have no metrics or extremely bad at data. Just buying segment is out of the question.

Jason Cohen [00:20:15]:
Like it's too expensive, too difficult in any way. They're not organized enough to have their data flow in and make sense. But if you could connect some of the dots, that would already be pretty useful to them and maybe motivation for them to connect a few more dots. And you keep it simple. So for example, okay, maybe it connects to HubSpot so that you can segment customers and send emails. Maybe it connects to stripe for things like billing. And maybe it just starts with that. And there's already probably hundreds of thousands of tens of thousands for sure, maybe hundreds of low hundreds of thousands of people who use those three things where you could already start to make some interesting progress on end to end data in the company, segmenting things, making decisions, getting reports, taking actions and so forth.

Jason Cohen [00:21:04]:
Segment was kind of like that at the beginning and they ended up going up market. Often when that happens, avoid is left behind. Because the original concept that was more for engineers and was flexible and simple, that concept was still right. It's still right. Today. It's just that they left that for more, for higher ARPU pastures, and that leaves some space behind. The bad news about that is whenever you're dealing with real time systems, that's a lot harder, especially for a smaller company. You don't have 24/7 teams, you don't have the time to make things that work well at scale.

Jason Cohen [00:21:43]:
That's a multi year difficult process for all companies, including segment, including WP engine, including things like Twitter and Facebook. Back in the day, like, no matter how many smart people you have at the company, as all those companies did, it's very hard. Google famously tried to do wave and failed for two years and completely failed. These are Google engineers. They're the best ever at all this stuff. And they're like, yeah, we couldn't make that work too hard to do that kind of real time multi user thing, it was too hard. So these are really hard things to solve. So the bad news is it's hard.

Jason Cohen [00:22:15]:
The good news is that keeps out other competitors. Also, sometimes things are hard and there's no benefit to overcoming the difficulty. It's just hard. And you're like, well, you can still do it, but why are you doing that? Startups are hard enough as it is. In this case, I think the difficulty is rewarded with a product that is actually useful, as evidenced by segments existence and in fact, segments existence, let's say ten or 15 years ago when they started. That proves it was useful, still is, and helps to keep out new competitors. So that's actually a useful reason to be hard.

Greg Isenberg [00:22:50]:
So, for context, segment was acquired in 2020 for $2.3 billion by Twilio. It's a huge, huge, huge business. And I think you're right. Oftentimes these businesses kind of get lost along the way when they have to go up market or down market and stuff like that, and it creates these opportunities.

Jason Cohen [00:23:15]:
I don't think they're lost at all. They simply changed their target market.

Greg Isenberg [00:23:20]:
Yeah, I mean, agreed. They're not lost, they're doing well. They're doing billions of dollars in revenue. I think that segment for x is the big opportunity, in my opinion. So if you can create segment for lawyers, segment accountants, segments for realtors, and focus on these different niches that make it easy for them to understand the value proposition and allow them to integrate into their systems really easily, I think that there's probably an opportunity, probably an opportunity there. Whereas they'll go on segment.com, the leading customer data platform powered by customer AI. If I'm the chief innovation officer, of a law firm or head of it at a law firm that might not speak to me as well as the leading platform for, for law firms just like you. And I think that's the opportunity that exists right now with, with someone like this who's probably trying to scale to all sorts of different niches at the same time.

Greg Isenberg [00:24:34]:
So I love this idea.

Jason Cohen [00:24:35]:
I don't think they have a niche. Niche strategy.

Greg Isenberg [00:24:38]:
Exactly.

Jason Cohen [00:24:38]:
They don't have a segment for lawyers. That's not what they're doing. Yeah, they target marketing departments at companies that are big enough to have real data estate challenges, but also can make millions, if not billions of dollars if they get it right. Or maybe save that amount of money in marketing spend that's not as useful and that's what they target. It's not a niche strategy.

Greg Isenberg [00:24:59]:
Exactly.

Jason Cohen [00:24:59]:
And that's fine. Just as we said at the top of the call, if it's a large and growing market, which corporate data is totally a large and growing market, with no doubt about it, that means niches open up that the big guys don't even want to target. Good. Don't. I like that. I like how that sounds.

Greg Isenberg [00:25:19]:
Totally. If you were starting this idea, how would you go about starting it?

Jason Cohen [00:25:27]:
I would want to know what is the vertical or niche that makes sense? I don't think lawyers are a good niche ever. The number of startups that try to do lawyers like, it's a very, very low probability success niche. It's just they don't get tech, they don't like to spend money, oddly. Cause they have a lot of money. It is true that if you get in there, you might stay in there for ten years for the same reason. So I guess that's good. But just in general, this is not a market where people readily spend money on software, understand all that stuff, know how to use it to create value for themselves and therefore stay happy, successful customers. And so those kinds of things, to me, very rare to see that be as successful.

Jason Cohen [00:26:12]:
I say that. And of course everyone can think of like one company that does do well. And I would say, right, you can think of one. You know how many startups attempt to do that? Thousands. And you can think of one. And so there you go. The companies that do really well in things like accounting and lawyers and dentists are 20 years old. And that's my point.

Jason Cohen [00:26:31]:
So I would want to know what is that niche? It could be SaaS companies. Because if you're a SaaS company, you're not afraid of technology. And in fact, you may instantly understand the value proposition and why to do it, even in the marketing department, or maybe tech in general, or maybe it is an industry vertical, maybe one where since often it's good for marketing, one where better marketing translates really directly into more revenue so they can just see it. I tried using this software, two months later my revenue was up or my costs went down noticeably. So I don't need an ROI calculator, I don't need a sales pitch. Like, you can't pry this out of my hands if you wanted to, because I already see myself like how this pays for itself ten times over. So that could be something like realtors, because it is true in that, like it's a marketing game, because all realtors are the same for the, at the first approximation, and so. Or it could be other verticals where this is true, like better marketing equals, oh, look, I found it.

Jason Cohen [00:27:33]:
It could even be something odd, like a media company. So there's a lot of companies who make money off the website, off of content, whether that's ads, which is the typical way, or maybe subscribers and paid subscriptions and so forth. Anyway, that's what I mean by media company. There it is absolutely true. If you get 20% more traffic and it's ads, you make 20% more money. It's kind of simple as that. There may be some arbitrage on paid marketing in order to do that. As long as you make more money than you pay, then this is automatically good.

Jason Cohen [00:28:03]:
So if something that was more sophisticated in metrics could help media companies get more views, that equals more money. Again, now, however, you're starting to get closer to what segment actually does. But again, since it's expensive and they go after big stuff and it's integrating to lots of different systems, that may be not, there may be plenty of space there. There are certainly millions of media sites on the web, like I'm describing, millions. So there's a lot there. So the idea that there might be a niche in there, that makes sense, that seems plausible to me.

Greg Isenberg [00:28:39]:
My advice for folks is go to where you have an unfair advantage. So if you have an Internet audience for nurses or healthcare industry, that might be a good place to start. If it's for realtors, that might be a good place to start. Now, it might end up being a niche that isn't worth going after, but it's definitely the first place that you should start. Yeah, I would just say go to where you have unfair advantages for picking a niche in general. I want to move to your third idea, which just says help people make practical, simple strategies and that got my ears peaked out, perked up. So I was curious what you meant by that, that idea.

Jason Cohen [00:29:29]:
Yeah. I think the challenge here is do people want it? Probably not. Do people want to pay for it? Maybe, probably not. So this might be a pretty bad thing. It is a problem. I think it is a problem, but that's about all I can say. Do people agree it's a problem? I don't know. Are they actively looking for a solution? Not sure.

Jason Cohen [00:29:51]:
It's probably pretty bad, but I do think it's a problem, which is you hear strategy and I think everyone has a different definition of it, which is already a problem. A lot of people conjure up these academic versions of strategies, which I agree is not very useful, maybe not useful at all, and workshops like making a swat, which I think is useless. And my evidence of that is it's rare that you do a swot and come away with some kind of insight you didn't know before. And it's also rare to ever reference one again afterwards. It's not like every month you're pulling up your swot because you're using it, you don't. Okay, well then how useful is it? So I think you hear strategy, you think of these things and you're like, eh, I don't need that. And you're right. To me, strategy, for what it's worth, means deciding how you're going to win.

Jason Cohen [00:30:46]:
That means things such as, and so I'm selling to exactly this group and this is my advantage, or here's my strengths or assets and here's how I'm leveraging those specific things in my way to win in my way, in my market. That sort of thing is strategy to me. So not a tactic. Like we're going to run an ad on this website or we're going to add this feature. That's, that's important too. It's just not strategy. That's your actual plan, that's what you're going to do. Cool.

Jason Cohen [00:31:15]:
For that, being agile is also good because in detail, it's hard to say in detail, you don't know which ad is going to work, so you have to try for that sort of thing. Being agile makes sense, but for this broad thing of, like, how am I going to win just being agile and saying, I don't know, I'm just going to throw a bunch on the wall regardless of what I'm good at or not. As you just said, that's not a good plan, that's not a good way to find a good strategy. I think in retrospect, when a company has a good strategy, it's easy to say what it was. You look at Netflix and you're like, oh, DVD's to the home, so you don't have to do the walk of death. That was smart. And then when they realized streaming is even better because it's immediate, that was good. There's their whole strategy in a sentence or two.

Jason Cohen [00:32:00]:
So to me, that is a strategy. This is the critical idea that sets them apart. That is unique. That's hard to follow. Does it leverage a strength they had that I don't know. I don't know anything about that. Maybe it doesn't, I'm not sure. But it does explain how they're going to win.

Jason Cohen [00:32:16]:
If someone else had the same strategy, then it would be a race. But since their competitors didn't have the same strategy, their strategy was better. It turns out it's really hard in the moment to come up with something that straightforward and yet be that compelling. But don't you need that because it keeps your eye on the prize of what is it. All the other tactics, all the other things you're doing is in service of this strategy. This way to win. This your way to win. How do you make one of those? How do you get there? If SWOT isn't useful, what is? How do I analyze my strengths and my assets and my competitors and the market and ideas that I have and stuff that I got through customer validation, etcetera? How do I take all of those inputs and ideas and things, and then arrive at a strategy that does what I'm saying, but isn't academic and silly and 80 pages long and all that stuff, which not good.

Jason Cohen [00:33:13]:
How do I do it? So this is where I think. I don't know if you would quite say that software is the answer, because clearly this is a human endeavor. It takes creativity and so forth. Can a system help? If not collect the data, then collect those inputs, which, again, a lot of that is qualitative, but to the extent that there's quantitative, fine. To the extent that it's qualitative, fine. I mean, even if that's sticky notes in Miro, whatever stuff. And then are there workshops or frameworks or systems to help go from that to a strategy that's simple and makes sense? I don't think a tool will come up with the answers for you, but it could walk you through some sort of thing, kind of like a course does or a framework does or a book does. Could it support that? So that the busy work of that is easy and the structure of it is such that it sort of funnels you into having the right conversations.

Jason Cohen [00:34:10]:
You still have to have all the conversations, make all the decisions. But could it help you make decisions in a good way? Could you, could help you arrive at a better decision by prompting properly, challenging properly. And again, I think AI is interesting here, not because AI will come up with your strategy. I do not believe that. But could it help prompt, could it help summarize, could it help generate ideas? Not evaluate the ideas, that's you. But could it generate things? Hey, if I have these seemingly conflicting inputs, people want all this stuff, but they also want it cheap. Or it seems like the only way to succeed here is to have all this complicated stuff. But I'm just one person.

Jason Cohen [00:34:51]:
I can't make a thing that's super complicated. I have to make a simple thing. These are my puzzles to solve. These are the kinds of things that strategy is good for. Oh, yeah, that's a good puzzle. Now we're talking. A strategy is how you're going to solve that puzzle. Oh, I'm going to make a product that targets this, then they don't want it as complicated, and then whatever the solution to a puzzle like that, that's what a strategy ultimately is good for.

Jason Cohen [00:35:13]:
Having solved those and saying, here's how I'm going to solve it. Okay, nice. Could AI identify some of these conflicts to cause you to have to resolve them? Maybe. And if so, that's really helpful. It can't just solve it, but could it generate ideas? I have no idea. I think it would take some prompt engineering and some, and some practice. Maybe certain models are better than others. I don't know.

Jason Cohen [00:35:33]:
But if it could generate, like, as long as you're brainstorming here, could it generate brainstorming ideas to help you get to a good idea that you actually have? Maybe I could. I could see that more. So, so again, there's a lot of maybes because of course I haven't done this or anything, but yeah, maybe as almost like a meeting facilitator type of thing or a generator of concepts and ideas. Could it help? Or one thing AI is pretty good at is summarization. So if I have a bunch of inputs and stuff, could I ask it, like, are there commonalities? Are there certain topics or certain things? That's the kind of thing it's actually not bad at. So again, like, maybe there's a role all for it as a facilitator or something to be helpful and to help you get to a place that would otherwise be hard or even that you wouldn't have gotten to. So is there something there? No idea. But I'm interested because I feel like especially smaller companies, especially bootstrap companies, have disdain for strategy.

Jason Cohen [00:36:38]:
And again, I don't disagree with them because often when you search strategy on the Internet and you see stuff, I roll my eyes at that, too. So I agree. But again, I think I'm painting a picture of an actually useful definition of strategy and a system that could help anybody, including a solopreneur, making something that's not a unicorn, have a good strategy. But right. Sizing what that means, that feels like it would be really valuable. I don't think it's good to simply not know how you're going to win. I don't think that's good. Can we have the right sort of tools, right sort of process for that? I'm very interested in that.

Jason Cohen [00:37:18]:
And as you can tell, I don't have, like, the complete answer. I have ideas. I have specific ideas about those frameworks that I've used and some of them I've written about as well. But I don't claim to have the answer. I think, though, if you could find an answer, you could really help a ton of companies be smarter about what going about what they're doing, maybe even de risking their companies by having a good strategy, by good meaning, good for them and good for their market. You probably can't measure that at all. But that doesn't stop you from making that, the vision statement. I would have a vision like helping thousands of companies succeed because they have the right strategy.

Greg Isenberg [00:37:58]:
Yep.

Jason Cohen [00:37:58]:
I think instead of fail. And like, how do I know that? Oh, I don't. That's why it's a vision statement. It's an aspiration. It's, you know, it's the North Star. It's not measured. I don't know who succeeds and fails otherwise. No, I don't know.

Jason Cohen [00:38:14]:
But having that in your head, but that is what we're trying to do here. Even if we can't measure it, still, what we're trying to do, that's actually quite helpful, staying the course of what you are trying to accomplish.

Greg Isenberg [00:38:27]:
Absolutely. I think the insight here is when you're in a boat, it doesn't make a difference how fast you paddle. If you're going in the wrong direction, that's a problem. And there's so much value in knowing the right direction, and that's where strategy plays such a huge role in this we incubated an agency called Latecheckout. It's called LCA, the website's late checkout agency. And we work with the dropboxes and Nikes and Shopifys of the world on strategy and product design. And we also use AI and automation to figure out what their communities want, and then we help build strategies around that. However, we only focus on these big companies.

Greg Isenberg [00:39:13]:
So I think there's a huge opportunity in taking that LCA model. And to your point, solopreneurs, small companies, startups, there's definitely an opportunity to build some agency mixed with AI and data application to get people's strategies as full proof or as foolproof as humanly possible. So I like this idea, and I do like agency businesses. So I think as an agency model, it could work.

Jason Cohen [00:39:42]:
Yeah, you can even imagine an agency having this as part of what makes their agency product. So I'm using that word in the general sense, making their product unique. In other words, there's a lot of agencies that say, we help you with strategy. There's obviously various products and frameworks and books and stuff about strategy. Okay, fine. Our agency has a special system. We will help you through it, we facilitate it. We have all this experience, et cetera, et cetera.

Jason Cohen [00:40:13]:
And also we have these systems that we've unique systems we've built that help both of us be efficient and smarter about that. So maybe you have this internal product as an agency, it's still you human beings helping. Because again, as I said all along, I don't think the product nor AI can just magically find the right strategy. I don't believe that. I believe it can facilitate, though, or I say I believe, I wonder if it could facilitate it. So whether that's helping the agency facilitate it much more efficiently for that agency, getting to better results, faster results, the more time the agencies, people at the agency spend actually facilitating a meeting, and less time messing around with all the stuff and reading everything and summarizing and organizing. And you know, the better, the more efficient. And maybe it even works better.

Jason Cohen [00:41:03]:
I don't know. That's pretty interesting. You could imagine the agency creating material, especially for lead gen, but also just for thought leadership on this topic. So you can imagine that there is a course that they put out on strategy where they explain it the same thing. Maybe they make money on the course, but I think more likely use it as marketing to say, look how smart we are, we can do this for you. There's all kinds of things you could, places you could go as an agency where rather than make this a product where you sell software, instead, it's part of your agencies offering, and it's just part of your thought leadership and showing how you're special and you have your own philosophy, and you just know what you're doing. In other words, and I'm also like, as much as I like all this, and I have ideas about strategy, it's certainly true that there's many ways of coming up with a good strategy. There's not just one.

Jason Cohen [00:42:00]:
And so it's not even necessarily, as the agency, we have the best way to make a strategy. That's not really. That's not what it is, nor does it need to be. It just needs to be that you're really good at it. You have a way that works really well. You, the agency, you are expert at this particular way. And so, sure, there's probably 20 decent ways. It's just this is one of them, and we're really good at it.

Jason Cohen [00:42:25]:
And we've even built our own materials, our own software to facilitate it. So we're just really good at this particular way. Works really well. We're not disparaging the way that's right next to us. It's fine. But this is what we do. I think that's a very. I don't want to say humble, but that's a very realistic way where you don't have to make crazy claims about being best in the world or nonsense like that.

Jason Cohen [00:42:50]:
And yet you can make a very credible case that, like, we're really good at this. You should just hire us. Like, you could read a book. Like, read blue Ocean, you know? All right. You could read it and try to do stuff in there. You can. And we're not against that. You know, there's some good ideas in there, of course, but we know how to actually go from a to z.

Jason Cohen [00:43:09]:
We know all the letters in between. We have this really great system. Here's materials about to whet your appetite, get you interested, and, like, that sounds fun. That sounds like it'd be useful. That sounds actually practical and not. Not just silly exercises. Hmm. See? So let's go.

Jason Cohen [00:43:22]:
You know, it's not that blue ocean's bad. It's just this is our way, and if we can convince you our way is effective and maybe even fun, then let's go. I think that'd be pretty cool.

Greg Isenberg [00:43:33]:
Yeah. I think if you're thinking about building an agency business, starting with the system is a good place to start. And then you can start creating content around that system, get people bought into that. Some people will connect with that system and some people won't. And that's okay. People that will, a lot of them are going to be like, well, I can implement the system myself, but it's going to take me a bunch of time. I'd much rather work with you to help implement this and bring this to my company. I think the.

Jason Cohen [00:44:05]:
Yeah, I mean, look, that's what the authors of all these books do. Now, sometimes the book does so well that they don't have to do it personally, but usually they still have their consulting company and just have other people do it. But like the blue ocean, people get paid to teach people blue ocean. Right. And a lot of these systems are like that where it's like, oh, so that's, that's exactly what happened once their materials got very popular. But you don't have to be as popular as that to have good materials. So that when you are doing your sales calls in your ways, you have an edge over the next agency who just says, I do strategy. Here's three customers who are happy.

Jason Cohen [00:44:38]:
You can call them. Yes, we all have that. Every agency has that. So, yes, there's nothing wrong with that, but it just doesn't stand out. Like, so what?

Greg Isenberg [00:44:45]:
Yep, exactly. And it's getting more and more competitive to start in, you know, to start an agency, because what's required, it's really just a landing page and a brand and a domain name. So, yeah, I agree. The systems really help stand out. Jason, thank you so much for jamming on ideas. There's a lot of good stuff here for people who are listening to this on Spotify and Apple. Please follow, subscribe, leave a comment, and then subscribe on YouTube. Just search my name, Greg Eisenberg.

Greg Isenberg [00:45:21]:
And then for folks who want to follow Jason, I follow him on Twitter. He's an incredible follow. And he's got a blog, too. Jason, where can people find you?

Jason Cohen [00:45:32]:
Yeah, it's a smart bear letter. A smart bear like the animal, because smart bear was the company I did in the knots, as they say. Another startup. So that's where I got the handle, asmartbear on Twitter and asmartbear.com for all those articles I mentioned.

Greg Isenberg [00:45:48]:
Beautiful. All right, well, thanks for jamming. And I'm sure someone is going to take a couple of these ideas, so it'll be interesting, interesting to see where it goes. All right, see you, Jason.