# This SaaS founder showed me every metric. It's not good... but we fixed it Jason Cohen · 2026-07-21 Jason Cohen and Dmytro Krasun dissect Dmytro's SaaS ScreenshotOne ($32K MRR, 9% monthly churn, flat growth) and find a structural ceiling around $40K MRR—9% churn on a growing base cancels new revenue. The crux: after four years and thousands of customers, no single large, repeatable market for screenshots-as-an-API has emerged. They lay out four honest options: incremental improvement to push $40–50K MRR; building a different product (e.g., regression testing for marketers) leveraging screenshot expertise; selling the business; or pivoting to AI integrations (MCP server, CLI, SDKs) to get embedded in AI workflows. Dmytro commits to a month of intensive exploration of the AI path, running the business on skeleton mode, to determine if this can restart growth or if he should sell. ## Questions this episode answers ### How does a 9% monthly churn cap a SaaS company's MRR at $40K? Jason Cohen explains that Dmytro's 9% monthly revenue churn creates a growth ceiling. Because churn dollars rise as MRR grows while new customer revenue stays roughly constant (about $3K per month gross), at around $40K MRR, churn offsets all new revenue, making further growth impossible without reducing churn or increasing new customer acquisition. [18:02](https://asmartbear.podhood.com/1075cd43-b45b-4052-99b8-6804ced04706?t=1082000) ### Why did a 4-year-old screenshot API SaaS plateau despite having 1,000 customers? Jason Cohen diagnoses that Dmytro hasn't found a large, repeatable market for his screenshot API after four years and thousands of customers. Use cases are fragmented—temporary projects, integration testing, cold email—and none show a big, sticky customer segment. This absence of a clear, substantial market is the key reason growth has plateaued at $32K MRR. [50:18](https://asmartbear.podhood.com/1075cd43-b45b-4052-99b8-6804ced04706?t=3018000) ### What is the AI/MCP growth strategy for a screenshot API business? Jason Cohen proposes that Dmytro pivot to become the default screenshot tool for AI agents. By offering an MCP server, CLI, and SDKs for Python and TypeScript, and actively marketing to AI developers, his API can get embedded in countless AI workflows. AI agents may use screenshots unpredictably but at scale, potentially breaking through the $40K MRR ceiling and reaching $100K. [1:25:46](https://asmartbear.podhood.com/1075cd43-b45b-4052-99b8-6804ced04706?t=5146000) ## Key moments - **[0:00] Intro** - [0:00] Dmytro Krasun opens his full SaaS dashboard on camera for the first time - [0:23] Dmytro Krasun says it's the first time someone has looked at his numbers so openly - [1:28] Dmytro Krasun sets a new goal of $100K MRR after reaching $10K MRR felt empty - [3:20] Jason Cohen states that Dmytro's current trajectory won't reach $100K MRR - [3:44] Jason Cohen advises against jumping to solutions without diagnosing the growth problem - **[5:41] Dashboard** - [5:45] Dmytro's SaaS dashboard shows $32K MRR, 1,000 customers, and $330K annual revenue - [7:02] Jason Cohen vs Dmytro Krasun: Is inflation the real goal for wanting $100K MRR? - [7:42] Jason Cohen quotes Dmytro: 'My whole identity was getting to 10K. I got there, and then I felt empty' - [8:43] Dmytro Krasun separates his business from his personal identity to treat it as a game - **[13:19] Churn Ceiling** - [13:19] Jason Cohen identifies 9% monthly revenue churn as the main growth problem - [13:31] Dmytro's MRR growth chart shows linear growth, but acceleration has stalled at $1K/month - [14:40] Jason Cohen states that most SaaS companies grow linearly, not exponentially - [15:37] Jason Cohen tells Dmytro his linear growth is normal, but needs a 'bend' to hit $100K - [17:49] Dmytro's churn math shows a revenue ceiling of $40K MRR if churn stays at 9% - [18:53] Jason Cohen explains the churn ceiling: 9% churn on a larger base eventually overwhelms fixed new MRR - **[28:33] Use Cases** - [29:15] Dmytro confirms that customers embedding screenshots into their product is the stickiest use case - [29:24] Dmytro shares a use case: hosting providers use screenshots for regression testing after deploys - [29:57] Jason Cohen identifies the deploy screenshot use case as integration testing - [30:23] Dmytro describes cold email companies using screenshots to personalize outreach - [32:06] Dmytro mentions a niche use case: SaaS onboarding extracts website colors via screenshots - [33:12] Jason Cohen dismisses the onboarding color extraction as a niche, non-scalable use case - **[50:37] No Market** - [50:37] Jason Cohen hypothesizes that the core problem is the lack of a big, repeatable market for screenshot APIs - [50:44] Dmytro Krasun admits the missing market is the scariest question - [52:03] Dmytro's SaaS is four years old with no clear market - [52:32] Dmytro Krasun says every promising new use case (like AI agents) hasn't scaled yet - [53:01] Jason Cohen states that currently Dmytro hasn't found a market to sustain tripling to $100K - [53:39] Jason Cohen suggests Dmytro could maintain $25-30K monthly profit indefinitely with current operation - **[54:46] Option A** - [55:09] Jason Cohen defines a good market as having a consistent problem, enough customers, and willingness to pay - [56:50] Jason Cohen explains that the crux is the absence of an obvious market, which is why growth is stuck - [57:30] Jason Cohen presents Option B: build a new product using screenshot expertise for a different market - [57:44] Jason Cohen pitches an idea: marketers need regression testing, and screenshots are the tool - [58:38] Jason Cohen suggests becoming the champion of regression testing for marketers - [59:14] Jason Cohen argues screenshots are necessary because dynamic web content can hide regressions from code checks - **[1:00:56] Option B** - [1:01:39] Q: Does Dmytro like Option B, the marketer regression testing idea? A: Not hell yeah, cautious. - [1:04:13] Jason Cohen advises that curiosity about Option B should be explored despite default skepticism - [1:05:08] Dmytro Krasun says he's tired and not excited about incrementally running the current business (Option A) - [1:08:28] Jason Cohen introduces Option C: sell the business or otherwise exit, as it's ideal for someone else - [1:08:51] Jason Cohen argues Dmytro's one-person business is sellable because one person can run it - **[1:10:44] Option C** - [1:10:44] Q: Is Dmytro ready to never see screenshots again? - [1:11:46] Dmytro Krasun says selling is the most exciting option, but he doesn't hate screenshots - [1:13:31] Dmytro Krasun clarifies he doesn't hate screenshots, just the current product-market misfit - [1:13:59] Jason Cohen explains if Dmytro hated screenshots, Option B would be off the table - [1:14:49] Jason Cohen argues leveraging current screenshot expertise gives strategic advantage for a new venture - [1:15:52] Jason Cohen points out the difficulty of screenshot APIs is a moat against copycats - **[1:29:30] Option D** - [1:31:06] Q: Is Dmytro sure AI agents will need screenshots? A: Exciting but unsure. - [1:31:50] Dmytro Krasun develops a sales pitch: AI’s native screenshot capabilities are unreliable and insecure - [1:32:59] Jason Cohen vs Dmytro: Browser access for screenshots is too permissive for corporate environments - [1:33:12] Dmytro Krasun ranks Option D as the most exciting among the four options - [1:34:17] Q: Would Dmytro regret selling without trying AI integration? A: Might regret. - [1:36:12] Jason Cohen outlines a one-month plan to test Option D: build integrations, market manually - **[1:40:06] Next Steps** - [1:40:49] Jason Cohen advises dedicating all available time to testing Option D, keeping existing business at maintenance - [1:42:04] Dmytro Krasun commits to slowing the current business to focus on AI testing for a month - [1:45:10] Dmytro Krasun reveals he already built MCP, CLI, etc., but never marketed them - [1:46:07] Jason Cohen distinguishes technical documentation from promotional landing pages for AI integrations - [1:47:39] Jason Cohen notes Dmytro can afford to spend $10K on ads for learning, even if it's not profitable - [1:49:07] Jason Cohen emphasizes that discovery spending is separate from scaling marketing ## Speakers - **Jason Cohen** (host) - **Dmytro Krasun** (guest) ## Topics SaaS Metrics ## Mentioned ConvertKit (company), WP Engine (company), AdWords (product), Claude (product), Claude Code (product), Codex (product), GitHub (product), Reddit (product), Twitter (product) ## Transcript ### Intro **Jason Cohen** [0:00] Okay, so the way this happened was, I think you were complaining into the void that growth had slowed or stopped, and WTF, and someone said, "Oh, you should talk to somebody." And one way or another, we got connected into: let's do a podcast where you show the entire world all your numbers and travails, and we'll go through it, and I'll be an asshole, and maybe by the end of it, uh, we can figure something out. So here we are, which is very strange. Um, have you ever done anything like this? **Dmytro Krasun** [0:28] Uh, no. I think it's like, like somebody looking at my numbers completely, you know, like open, and talking about this? I think it's the first time, probably. **Jason Cohen** [0:38] Yeah, you know, people are always worried about giving the numbers. I don't know why, maybe because they're worried that they're not good enough. But what I figure is, whatever they are, most people watching don't have numbers that are even that good. So they're impressed, and in fact, wanting to be inspired by something that you're doing or get to the place where you are. And people that have, let's just say they have a bigger company, which does not mean they're more successful, but let's suppose that's the case, they know how hard it was to get there. They don't look down and say, you know, "Oh, that's terrible." Oh yeah, I was there. We're the same kind of person. You know, we're, we're, we're connected in a fundamental human way. Anyway, so, but it's still scary nevertheless, because it is, it's private normally. Um, so what do you want to accomplish today, besides the fact that it's not growing? Like, are you looking to grow again? And that's what we want to come away with today, is how to do that, or just diagnose it, or what? **Dmytro Krasun** [1:28] It's, uh, it's an interesting question in itself. I'm happy to see you. I appreciate taking time to talk and to discuss all this. You're way more experienced in this, and I can learn a lot. So when I posted this post about growth, it was kind of like, you know, my recurring topic of the past few months, maybe, maybe even more, I don't know. But one of my initial goals was to build business on my own and to have this 10K MRR, whatever. And then I reached this goal, and maybe I wasn't ambitious, you know, I don't know. Like, but, uh, but then I started to grow it, to grow it, and then I realized, you know, like, it's, it's pretty boring. Like, it, I noticed that there is no, like, the growth itself is not growing. And, um, it started to trouble me. Like, is there some fundamental issue? Maybe I can do something. Because if I project all numbers, I see there will be plateau. And I, I knowright now that I won't be past this plateau. I have kind of a new goal, you know, like, um, in my mind, I want to reach now 1,100,000 dollars of MRR. And it feels for me like that having the current numbers doesn't lead me to this goal. At least, like, maybe, I don't know, in 5, 10 years, 20 years. Still not bad. Still don't complain and super happy about everything, yes. But, um, it feels for me like a real interesting challenge to tackle. And, um, having the partner or somebody with whom I could discuss this and to see, like, on the outside, especially with experience, it would be super helpful. Yeah, so that's maybe. **Jason Cohen** [3:20] So you want to get to 100K MRR, but at the current course and speed, you won't get there. And it's not even close. And so you got to do something different, because if you keep doing the same things you're doing, you won't get there. **Dmytro Krasun** [3:31] Yes, yes, yes. And maybe there are some, maybe some, you know, kind of collateral questions to this. Maybe I also could sell it. Maybe I could build a new product. I don't know. But that's like another, you know, like philosophical question. **Jason Cohen** [3:44] Yeah, so, so, um, what that does is you're jumpingright from the problem to, like, what should I do about it? In other words, um, growth has slowed. That's something we have to really understand, like, why, what's going on. And then separately, the question is, what do you want to do about it? Like, one thing you could do is, and so I don't want to do this company ever again. Oh, that's one path to deal with that. Another path is to solve whatever it is we found in the problem of growth. That's another, and indeed keep running the company. That's another solution. But how do you even know what that option is if you haven't identified in detail what's causing the problem of growth,right? So, so I hear you on that. That'sright. But you can't really tackle that until you figure out what the problem is. Now, some people are too dogmatic about this thing where you have to first identify the problem in great detail and only then go do the solutions. Well, I, I just said that. So, so like, in general, I think that'sright. But the way human brains really work is what you just did. When I want to jump to the solution, and then I jump back, but wait a minute, then you jump back to the problem, you jump back, and that is how we think. And it's good because one can inform the other. Because, for example, you start saying, well, maybe I'll sell the company, you go down that path, and you realize, I would get so little for the company, and I'm just making this up because we haven't talked about it yet. I would get so little for the company, I don't like that. Okay, well, we still go back to identifying the problem, but now you have a renewed motivation to solve the problem because you didn't like one of the other options. So this kind of a path is normal. In general, it is better to figure out what's going on and really get to what some people call the crux of the problem. Like, okay, the general problem is growth has slowed or stopped. But why exactly? Like, what exactly is the challenge that's going on? We get to that, because that's the thing to solve specifically. And then we find the solution. That's usually a better path. So it's okay if we bounce around a little, but like in general, that's a better way to organize it. So do you want me to pull up this dashboard so we can get the general numbers out there? **Dmytro Krasun** [5:40] Yes, let's start. Let's start. **Jason Cohen** [5:41] Okay, here we go. So, uh, okay. ### Dashboard **Jason Cohen** [5:45] Approximately, which it's not approximately, or is it? Like, this is the numbers,right? **Dmytro Krasun** [5:49] Most of them are correct numbers. **Jason Cohen** [5:51] Okay, well, let's just go through it and you can, you can correct me if I'm wrong. So the MRR is 32K, which is great. I mean, really, like you just said, when you started the company, you wanted it to get to 10K of MRR. It's three times that. So I guess my first question actually is, what's wrong with that? Like you said, I want to, I want to get to 100, but what's wrong with 32? **Dmytro Krasun** [6:08] Um, it's a great question. So getting to 10K MRR actually was kind of, you know, like an identity crisis. Because I was living in this perpetual identity of achieving this 10K, and then I achieved it, and then I felt like kind of empty. So I kept seeking answers, like, what can I do next and so on. And I kept improving the product, working on churn, so on and so on. And then I realized, okay, so I, I, I started to slowly realize that my business is separate from me, and it can have goals on its own. And for business, it would be better to grow. And because of many reasons, like inflation, um, like, if, if, even if I'm not growing, in 10 years, probably the money the business earning now, it will be two times less of purchase. **Jason Cohen** [7:02] Yeah, this isn't the reason. This is not the reason. I'll tell you why. I'll tell you why. Because if you said, because if it were really true that you were happy with 32, except for inflation, then the challenge that you would say is, I'm very happy, but I need to grow 3% a year, or I won't keep up with inflation. How should I grow 3% a year? You would ask. But you already know how to do that. That's not the problem. That's a, that's a just, that's a rationalization for the actual thing that you feel. You want to get to 100. That's triple what it is now. **Dmytro Krasun** [7:33] Yes, yes. **Jason Cohen** [7:34] So to me, when I hear that number, what I hear is over a million dollars a year. That's what I heard. That's what you're thinking,right? **Dmytro Krasun** [7:40] Yes, yes, yes, yes. **Jason Cohen** [7:42] Yeah. So, but you also said, and I know, I know we'll get to the growth, but you said, my whole identity was getting to 10K. I got there, and then I felt empty, i.e., that wasn't really a fulfilling goal after all. And your conclusion after that was, well, when I get to 100, I won't feel empty. It'll be different that time, which it might be, because there's another zero. So you might beright. Um, but probably you're wrong. **Dmytro Krasun** [8:13] Yeah, no, no, no, no, no. Here I agree with you. I know that when I know that this one will not, like, I will not get fulfilled from this one. But if I separate, like, business from my identity, like from who I am, and having a separate entity, like, like business as a separate entity, then having, it's growing for me, it's like, it's kind of like a nice challenge, you know, like a nice puzzle to solve, like in that way. Yeah, so it's, it sounds nice, and I'm building. **Jason Cohen** [8:43] That's what you're saying. You're saying you've converted your mindset from this is who I am to I have this thing called a business. It's a game. And it's a good game because it pays me when I do well. And that's a different mindset now. And so it's a healthier mindset. We are not looking for personal fulfillment. So okay, that's fine. I still kind of don't believe you, because I don't know any founder who doesn't care at all. Actually, that's not true. I know some people that flip things, and I guess they don't care. But I only half believe you, but it's fine. So okay, so, so there really isn't a reason other than playing the game and making more money. It's not inflation. Is there another reason, or is it just, I just want a successful business because that makes me impressive and puts enough money in the bank that I can, let's just, let's not say it makes me happy or fulfilled, but let's say it removes all kinds of things that money can fix. Is that it? **Dmytro Krasun** [9:31] Yes, it's interesting. You get pretty quickly to the core, you know, like of the, this stuff, because I actually don't believe it 100%, what I say about, like, a narrative, but I'm like in the process of building it. Like it's, it's kind of, that's one of the reasons, like that was original post on X, I posted, yes, it's like contemplating about what I want to do. And it's like, it feels like a narrative that pulls me, you know, this one. And yes, I, like, actually I'm, I'm happy. Like, everything is good, you know, like, but, but. **Jason Cohen** [10:05] No, everything's not good, or we wouldn't be talking. Okay, so we'll, we'll leave it for now, but, but you see that this is, uh, not clear, actually. **Dmytro Krasun** [10:12] Yeah, yeah, yeah. **Jason Cohen** [10:13] And, um, the reason it matters is because later on you want to talk about, should I sell the company? In other words, should I, do I want to keep playing this game or not? And so that's when all these questions we're just asking do matter,right? Because if running the company is fulfilling, then selling it may not be. But it could be vice versa. So, so these things do matter later. Like they don't, they don't necessarily matter in terms of diagnosing what, what's happening with growth, but they will matter later. And it just gives me some context of how you're thinking. So, okay. I think, I think I understand enough. So you have 1,000 customers and a third of a million in annual revenue, 32K in MRR. And it's very profitable, of course, because I'm, I'm guessing it's just you. There's no employees, and so this is just normal business expenses, or maybe there's a consultant, or what is the few expenses that result in that profit? **Dmytro Krasun** [11:01] Yes, yes, currently it's like, uh, infrastructure, servers, many SaaS tools, and some marketing. Yes. **Jason Cohen** [11:09] So not people. **Dmytro Krasun** [11:11] No, no, no. **Jason Cohen** [11:12] Yeah. When you say I want to get to 100K, are you thinking, and still have no employees? **Dmytro Krasun** [11:18] Uh, it doesn't matter, I think. It's, uh. **Jason Cohen** [11:22] It doesn't matter? It does matter. **Dmytro Krasun** [11:26] Uh, no, no, I mean, like, if it would be, let's say, it can be, it can be with employees, yes, it can be. It can be. **Jason Cohen** [11:35] Well, but two minutes ago, were you thinking that you needed to hire someone soon? **Dmytro Krasun** [11:43] No, no. **Jason Cohen** [11:45] Hmm. Why not? **Dmytro Krasun** [11:47] Why not to hire? **Jason Cohen** [11:49] You have the money. **Dmytro Krasun** [11:53] Why not to hire? I, like, like, because I don't feel there is a need. You know, it's. **Jason Cohen** [12:02] Yes, yes, there is. You just said you're, there's a huge problem, which is you're not growing. And you want to, you want to triple the company, and it's not growing at all. There's obviously a need. But you didn't think another person could be helpful in that. Whether that other person is focused on growth, like marketing, or whether that person is focused on tech so that you could focus on marketing, regardless, that did not cross your mind. Which, which is interesting, because it tells me, like, that's not the way you're thinking of solving the problem. And maybe you'reright, maybe not. Um, maybe it makes sense to go solo to 100. I mean, especially nowadays, uh, that's, that's more possible than ever, I think. But I'm just trying to get a sense of what, how you're thinking about it. Like if you had three people, if it was 100K and you had three people that you worked with, and let's even suppose you like them, just for fun. Um, does that sound good? Or you're like, eh, that's not what I meant. **Dmytro Krasun** [12:54] Yes, yes, I'm a bit afraid, yes. I'm a bit afraid of this. **Jason Cohen** [12:58] Um, and sometimes, sometimes the issue with, is we don't know what they would do. In other words, like, I didn't hire someone because I don't know what to tell, I don't know who I would hire or what I would tell them to do. Which again, is going back to like, I'm not actually sure what the problem is, so I don't know what I'm solving. **Dmytro Krasun** [13:11] Yeah, yeah. Yes. **Jason Cohen** [13:19] Maybe, but that's not, that's not emphatic. So maybe not. Um, of course, I see the revenue churn at 9%, and I can already tell that that's my main problem. Um, maybe you've already come to that conclusion. Let's find out. ### Churn Ceiling **Jason Cohen** [13:31] Now, here's what's funny. I'm looking here at this MRR growth chart. And what I see is growth. Now, this is a one-year chart,right? 12 months? **Dmytro Krasun** [13:40] Yes, yes, 12 months. **Jason Cohen** [13:42] Okay, and so all I see is it's going up with occasional bouts of not going up. **Dmytro Krasun** [13:46] Yes, yes. **Jason Cohen** [13:47] Okay. So I have several questions. One is, it's been going up for six months straight, and then you had one month where it went sideways, and your conclusion is it's not growing anymore. But what I see is five out of the last six months it grew. So why are you saying it's not growing anymore? **Dmytro Krasun** [14:04] Oh, it was about the, um, it's maybe because of my language or how I speak. It was about the, the growth of growth. Like the growth itself, like every month the product is adding like $1,000 monthly, yes, on average, yes. And then the, the growth itself, like zero, the growth of growth, yes. So it's like. **Jason Cohen** [14:25] Yeah, I understand. **Dmytro Krasun** [14:26] That's what I mean. Like that's not the MRR itself yet. **Jason Cohen** [14:29] Yeah, we call that acceleration. **Dmytro Krasun** [14:31] Yeah, correct. Yeah, yeah. **Jason Cohen** [14:33] Right, getting faster and faster. **Dmytro Krasun** [14:35] Yeah, yeah, yeah. **Jason Cohen** [14:35] But that doesn't usually happen. **Dmytro Krasun** [14:38] Yeah, yeah, yeah. **Jason Cohen** [14:40] Usually growth is linear. **Dmytro Krasun** [14:43] Okay. But that's. **Jason Cohen** [14:46] There's a lot of startups that, that, I mean, not a whole lot, but you know, there's dozens of, of, of startups of this general size that have shared their numbers. **Dmytro Krasun** [14:53] Yes. **Jason Cohen** [14:53] And you've probably seen them. What generally happens is it looks like a line, and then sometimes there's a bend in the line, like something structurally has changed and the line bends, but then it's still a line again. **Dmytro Krasun** [15:04] Yes, yes, yes. **Jason Cohen** [15:05] Now, where you see something more like acceleration is huge runaway successes, you know, Facebook or whatever, where like, allright, it's accelerating. **Dmytro Krasun** [15:12] Yeah, yeah. **Jason Cohen** [15:13] Sure, but like you don't necessarily expect to do that. **Dmytro Krasun** [15:17] It's interesting what you're sayingright now. It's like, I have inflated expectations in a sense. That's what you say. Like, like that doesn't, like you say that usually that doesn't happen. So, but why? Like, okay. **Jason Cohen** [15:37] Yeah. Okay, there's why, but like before we get to why, like you're saying it's important that growth accelerate, and yet your growth is fine and normal. But you're, you're just not happy with that unless it accelerates, or did you just not know that? **Dmytro Krasun** [15:49] Uh, yes, it was like, if you take this initial goal of that, that maybe like, uh, that maybe I don't believe fully, yes, about 100K, that I was working backwards from it, yes. Like if I, if I want to reach this number. **Jason Cohen** [16:06] Oh, you just won't get there fast enough. Yeah, yeah, I understand what you're saying. **Dmytro Krasun** [16:09] Yeah, yeah, yeah. **Jason Cohen** [16:10] Okay, so here's, here's just some data. So this is an article of mine, so that's very self-serving to do that, but whatever. Um, because this is some of the data. So here's, here's Palli's for your MRR. So it's flat. **Dmytro Krasun** [16:22] Yes. **Jason Cohen** [16:22] And then it's, then it's something, but look at that something. It's like a line. In fact, it looks like yours, like close enough. Up, up, up, up, up, up, up, up, up, up, up. A little flat period, up, up, up, up,right? That's kind of like you. By the way, this is a million MRR. That's what that is. And then here's, uh, ConvertKit. And I mean, yes, they, they had this big growth swing, but look, for, for a long, long time, uh, years, it's this linear thing. And then even when it bends, it's a new line. See, it's still a line. That's, okay, so that's a very fast-growing line. Fantastic. But it's not, it did accelerate at this one point, but then it would turn into a line again, just at a much faster pace. That would get you to 100K in MRR, a faster pace of growth, but not necessarily constantly accelerating. Here's another one. This is ConvertKit later on, just to see it. Look at it. Look how, look how straight a line. This is, this is like 10 years. **Dmytro Krasun** [17:12] Yes, yes. **Jason Cohen** [17:13] Right? I mean, that's, that's very, very a straight line, except for a change, which I think is what you're saying is you want this change, this bend in the curve,right? Okay, so, um, so let me, so that's enough of that. I just wanted to show you what, what I meant. Um. **Dmytro Krasun** [17:27] Yes, yes, yes, I see, I see. **Jason Cohen** [17:29] Yeah, so I, so I look at your graph and you're, you don't have the bend. This is the issue. If we showed a longer picture of this. **Dmytro Krasun** [17:35] Yes. **Jason Cohen** [17:36] Um, it's what you said. Like you're saying, look, I'm adding net $1,000 of MRR per month. So I'm just seeing like, allright, so in a year it's 12. And, and so if I just stack 12s on top of each other, that's too many years. And also that's not what's going to happen over that timeframe. Um, and also your cancellation rate sets a cap, which I think you know already, because you have a thing here, new MRRs colliding with churn. You want to explain that? **Dmytro Krasun** [18:02] Um, yes, yes, it's, it's the moment when approximately the number of revenue, churning revenue, and growth, like, will annihilate each other, like, in a sense that they will be, like, if I keep growing with the same rate and my churn keeps up, like, at some moment it will be like zero. It will be like, I cannot add some MRR. So this approximately at 40K, this number. So even if I continue to grow, I will anyway, I cannot go to 100K. **Jason Cohen** [18:34] Yeah, yeah. So, so this is a way for, for other people who are listening to understand this. If marketing is adding 1K of MRR per month in new revenue, but cancellation is 9%, let's say, as the company gets bigger and bigger, that 9% is 9% of a bigger and bigger number. But marketing is still putting in 1K, 1K, 1K each month. But the churn is going up every month, because the company's growing and 9% of a bigger number is a bigger number. So effectively the dollars of churn go up every month because of this 9%. But marketing is not going up every month. It's just staying at 1K. So at some point, you're at a size where, um, the number of customers, number of customers or revenue, either one, that walks out the door in churn is equal to the amount that, uh, that new revenue, that, that, that marketing is bringing in. And at that point, the company is literally not growing. Um, and you're saying that's at 40K MRR, which makes sense to me. And a simple formula to derive this, by the way, which you probably know, because you made this, uh, this, this chart, is you take your MRR and you divide it by, or sorry, you take your new MRR every month and you divide it by your cancellation rate. So if you have one, uh, 1K MRR and you divide it by the cancellation rate of, of 9%, um, that's actually much worse. So I'm guessing 1% is probably net of cancels. Or, I'm sorry, that one, $1,000 is net of cancels. I'm guessing more than that comes in and that net of cancels, it's one. Yeah, here it is. June, um, it's 3K. That'sright. So 3K of MRR comes in, but at 9%, so divide it by 9% and you get about 30, 40K in MRR, which is about where you are. And that's why this chart was flat, because you brought in 3K, but 3K left. And you're saying, yeah, even, even in other months, this is still the case. And, uh, so I can't ever get past this because of that 9% cancellation, which that's the first thing my eyes went to, is this revenue's so high, there's no way you can grow past a very small limit. So you've seen the same thing. So the obvious question is, why is your cancellation rate so high? **Dmytro Krasun** [20:47] Yes. It's a good one. It's a good one. **Jason Cohen** [20:51] Well, what's the answer? **Dmytro Krasun** [20:55] Um, it might be a few reasons. **Jason Cohen** [20:59] Well, what do you mean might be? **Dmytro Krasun** [21:01] Yes. Like, yes, I don't know for sure. I don't know, yes. **Jason Cohen** [21:07] Why? **Dmytro Krasun** [21:08] But, uh, why I don't know for sure. **Jason Cohen** [21:10] That no one ever talks to you when they leave? You don't ask? **Dmytro Krasun** [21:12] No, no, they leave. They, they have answers, but, but I assume, but, but like, uh, but let's say I trust this like 99%. Like I have, like people leave answers and, uh, actually I shared them also, but, uh, but, but it, like there is two layers to that. For example, uh, I can say that the reason of churn because people having, uh, they say that, they say that they have short-lived use case. They just used product for some, some feature of theirs, I don't know, uploaded some file and then just canceled. **Jason Cohen** [21:54] Right. **Dmytro Krasun** [21:55] And, uh, or I can say that the reason initially that I was somehow attracting, uh, this type of customers, yes. So it's like, so it's, it's a bit mixed, but, uh, because I have like, from what I see, I have the most churn on the smallest plants, like on the cheapest plants. And, uh, like, so maybe it's not, so maybe it's not, uh, initially was a good fit for this kind of product. That's like what I think, uh, globally maybe, yes. But, but if I check per reasons, there are like many and, uh, many of them. And, uh, the most, the most popular one, like written manually by, by many people. And, uh, it's like, um, they just needed product for a month, for some feature, for some project, something like that, yes. **Jason Cohen** [22:54] So when you say temporary slash no longer needed, um, those are not the same thing. So in other words, there's a really big difference between a person who says, I know going into this that I need it for two weeks. **Dmytro Krasun** [23:08] Yes. **Jason Cohen** [23:08] And a person who says, I've been using this for two years, but now I don't need it anymore, presumably because some projects ended. **Dmytro Krasun** [23:16] Yes, yes, I agree with you. **Jason Cohen** [23:18] So do you have a sense of, of how this splits? Is it, is it mostly temporary? Or is it, or is there a lot of people who use it for two years and then, then don't need it anymore? **Dmytro Krasun** [23:28] Uh, the larger, um, yes, yes, I think it's, it's, it's mostly temporary a little bit for this one. **Jason Cohen** [23:35] Yeah, yeah. **Dmytro Krasun** [23:35] Because I know, like, I have many customers who are still, who are still using the product for like more than two years, but many of these, yes, yes. Intuitively, yes, I cannot back up with numbers, yes. But, um, it was. **Jason Cohen** [23:51] And when it's temporary, how quickly do they leave typically? **Dmytro Krasun** [23:55] Oh, I think it would be something between one to three months, something like that. It might be even just one month, you know, like just subscribe and use the product. **Jason Cohen** [24:03] So one thing that would be really good to do is to segment your customers in, let's say, the first 100 days and everything else as two different segments, because what's happening is all this stuff, now whether that's okay or not, or good or bad, even setting that aside for a minute, that's most of the people that are going through the product. Like just adding up the temporary and the free plan is enough and this kind of stuff, it's, it's like already 50%. So, um, so that's drowning out the other metrics, which are also important, which is when people, when it's not temporary and they use it all the time and they've used it for a year. In other words, when it seems like an ideal customer, that data and that, those effects are being drowned out by the people where, again, it's not necessarily a problem, it's just that it's not this ideal sustainable business. And so, notright now, obviously, but in future, that would be a good idea to completely segment those customers so you can get information about when it's temporary, what is that about? Is it certain kinds of use cases? Do I get those customers in certain channels? Am I saying the wrong thing? Am I attracting the wrong thing? Is it fine? It's just, they're just not long-term customers. Uh, and I should have a, uh, either a different sort of an offering, like maybe a different price point. There's like the, the 100-day plan or something where they pay a certain amount. It's more like one-time revenue because that's what they actually want. Or maybe don't do that, but segment them anyway, just so that you're not worried when they cancel. You're like, I expected that, you know? And, and but you would tune things like, well, how much can I spend on things like marketing or even tech support on a person that is definitely going to churn in the next 100 days? Definitely. And you go, ah, a little bit, but very little,right? Because this is not a lot of revenue and I have to think of it like one-time revenue, uh, because it effectively is. And so I have to, the rest of my business metrics or business, um, decisions have to be aligned with that. But then there's these other kinds of customers where I do expect them to be with me for years. And so I can afford more spend to acquire them. I can afford to work with them more in support. Maybe there's integrations they want help with that's very technical. Um, but that's worth it because that means they're here for three years because they've built you into some system that they have, some permanent system. So it's worth it. So you want to be able to segment them so that you can be, you can react and behave and, and spend money separately. And, and that's hard. Like, uh, maybe you have to ask ahead of time, even if this is just in the product, what this is for or what this is like, so that you can know ahead of time. Because people just sign up, how would you know? Well, you could ask them what it's for, what they're doing, this and that. You also could have different pricing plans that would naturally self-select people so that the people who, for whom they know it's temporary, it would be more attractive for them to pick a certain pricing plan. And then for people where it's, um, where, where it's long-term, there would be a different pricing plan that'd be attractive to them. And then they, they somewhat self-segregate. That would be good. There's different ways, again, not just jumpingright to the conclusions, of course, but there's different ways to, to figure out what that segmentation is, because it may be that on the temporary, all you can afford to do is stuff that, that's incrementally free, not, not free, but incrementally free, like social media and SEO, where, um, of course it's not free either in time or money, but, um, but getting one more person to see it is, is free. Um, as opposed to advertisement, where that probably doesn't work for the temporary customer. There's just no price that's, that's not, you know, ridiculous, uh, where you can afford to buy customers that are going to, you know, spend a total of 20 bucks,right? Um, but you could afford to buy someone that's going to spend thousands of bucks. So do you have a senseright now of what those use cases or those things might be? Or is that just an interesting homework problem for the rest of today and tomorrow? Um. **Dmytro Krasun** [28:04] Use cases for customers who would stay or for customers who are leaving? **Jason Cohen** [28:10] Both, of course. **Dmytro Krasun** [28:11] For both. Um, yes, yes, I have, uh, I have examples of this. **Jason Cohen** [28:18] Well, what's a category? Like, think about the customers who stay forever and are happy. Um, I'm sure there's not just one kind, but what's an example of why that is? What have they done, just generally speaking? **Dmytro Krasun** [28:33] What have they done? ### Use Cases **Jason Cohen** [28:35] Like, why are they using it forever? Did they build it into? For example, one answer could be, sometimes the act of taking a screenshot on the web gets built into a feature in a product that they sell their customers. So unless they rip that feature out, they're going to be a customer forever. And they certainly don't want to build all that infrastructure themselves. That's a distraction. They've got their own product to work on. They don't want to do this. This is hard. But since it's a built-in feature, like they're just going to stay forever unless they rip the feature out, which typically features are not ripped out. Therefore they stay forever. And so the, the, in that case, the, the generalization would be people who build in screenshots as part of the product that they themselves sell. **Dmytro Krasun** [29:15] Yes, yes. **Jason Cohen** [29:17] So that might be an example. Is that an example? And if, and are there other examples of categories like that? **Dmytro Krasun** [29:24] This one, like one of the most, the stickiest ones are these that you described. And also there is kind of category. For example, I have a company who are screenshotting hundreds, thousands of websites after deploying updates and then checking if there are any changes and then reverting back, uh, their deploys. But it's kind of, it's not like the main feature, you know, but it's like, it's, it's very important for them and they don't want to spend on this kind of task. Like it's, uh, like secondary importance. I don't know how to say it. **Jason Cohen** [29:57] Okay, that's called integration testing. Uh, or, or, or, um, or end-to-end testing, something like that. Okay. Um, it, it, do people build it into products like I said, or did I just make that up and that, that's not whatever happens? **Dmytro Krasun** [30:10] No, no, it's not. This one is not like a part of the product, I think. It's like for, but for. **Jason Cohen** [30:14] Okay, that never happens. Besides integration testing, what's another example of a, of a category that actually does happen among the people who stay forever? **Dmytro Krasun** [30:23] Who stay forever. For example, companies, um, a few examples, I think. Company that is sending cold emails. And in cold emails, it's like one of their main features. And in cold emails, they have templates for, let's say, cold email with a screenshot in it of the prospect website and, uh, something about it. And it's like one of the main features, let's say, of, uh, such a customer. Yeah. So it's, it's like maybe one of the top three, top five features of such a customer. **Jason Cohen** [31:00] Gotcha. So, so someone's doing some kind of personalized automation campaign, like cold emailing. And part of how you get someone's attention is you put their screenshot or their website in the email. Um, whether just to get their attention or because you're pointing something out or something like that. Um, in the same way that there people say, although I disagree with this, but there's that thing that where you're supposed to use the other person's name when you, all the time, uh, because, because that means, because they like it. I just think that's weird. And then I think that you're doing that on purpose and that means I don't trust you. And so to me, it has the opposite effect. But okay, this kind of like that. If you, if you show me my own stuff, then, uh, it seems more personalized in any way. Maybe I'll look at it and like it doesn't seem quite so cold. **Dmytro Krasun** [31:47] Yes, yes. **Jason Cohen** [31:47] Okay. So integration testing of, of websites or web apps, who knows? Um, cold, or, well, things where you want to customize some kind of outbound-like emails, um, with their, with their website. Okay. Is there any other things? Any other categories among the people that stay forever? **Dmytro Krasun** [32:06] Uh, like forever, yes. It's a dream. But, uh, for example, in some, in some SaaS products in onboarding sessions where they need to fetch like company website and extract some colors from it and to personalize their SaaS for, um, for this, uh, for their customer and to speed up onboarding. Uh, yes. **Jason Cohen** [32:36] Um, that sounds really cool, but I would be shocked if 100, if you had 100 customers doing that. **Dmytro Krasun** [32:42] No, no, not, not 100. Yes. It's like, I'm, I'm, I think about use cases where like I know like customers who are still using the product and it's like. **Jason Cohen** [32:52] Yeah, but how many people do that? The onboarding sessions with the colors, how many people do that? **Dmytro Krasun** [32:58] Yes. It's, it's not a lot. Yes, it's not a lot. **Jason Cohen** [33:00] Is it more than one? **Dmytro Krasun** [33:02] Yes, yes. **Jason Cohen** [33:03] Is it two? **Dmytro Krasun** [33:07] Uh, it feels for me it's like, yes, it's a, it's a good question. It's a good question. **Jason Cohen** [33:12] Okay. So that the answer is it's not 10. Okay. So, so that one, I, that one's fine. I, I, I mean, so, so I mean, of course there's going to be one-off cases, which is awesome. They just won't be patterns that you can easily, they'll be easy to exploit. These other ones feel a lot more exploitable to me. This one sounds really fun, actually. Really interesting. Great idea. I just, it just doesn't feel like something that, uh, that, that's widespread. That would be easy to scale or something. Okay. So, but we should spend our time mostly on the successes and getting more of them, because, uh, we could talk about why people do things temporarily, but my guess is there's a huge number of reasons why people do things temporarily. Like, why do I need a screenshot for a few weeks and not again? I mean, there's probably infinite reasons, even including just hobbyists doing weird stuff with AI and stuff that you're just trying crap. Like, there's probably all kinds of reasons. And it, and in a certain sense, it doesn't matter because you don't want any of them as customers. **Dmytro Krasun** [34:10] Hmm. Yes. **Jason Cohen** [34:12] Now, it's okay that they sign up, but they're going to sign up and leave. And so for your, for your goal of getting to 100K,right? Or really even just a goal of not capping out at your current size,right? Like the, even setting the 100K aside, just saying, I just don't want to stop growing like I am. Like just start, you know, even without a goal of where it should be. Um, these people that come in for two, for two, for, uh, two months and leave are not going to be the solution. **Dmytro Krasun** [34:38] No, no, it can be once, like, yes, it can be some nice one-time revenue, you know, having this. **Jason Cohen** [34:44] But they're not going to be how you grow, because you can't afford, you can't afford to spend time to get them because they don't give you any money. **Dmytro Krasun** [34:52] Yes. **Jason Cohen** [34:53] Okay. So, um, so, so, so that's why there's probably, you know, hundreds of use cases and it kind of doesn't matter because that's not what you want to do. It's okay when they show up and they leave, especially if you've segmented your metrics so that you're not getting confused, um, as to, um, what metrics and activities are important for you to focus on to get the actual goal, which is people that stay a long time, because that's where the growth is going to come from. That's how you're going to get your cancellation rate down,right? Versus all these people. So I'll just take the integration testing for the web, um, for websites. How many software companies do integration testing of some kind with either a website because it's marketing or a website because it's some web app and they're doing some kind of integration testing? Maybe it's in a continuous integration server, maybe it's manual, maybe it's a QA team, but somebody's doing that. How many software companies, let's just say, I mean, I don't want to be too America-centric, um, but how many companies, let's say in America, do you think there are as a power of 10 only who do integration testsright now with websites? **Dmytro Krasun** [36:07] I think this is a lot, but the, but this one specific, like the moment you said about like numbers, how many, like we talked about forever customers, yes, but how many of them I have? Like this one specific use case, it's not even like any random companies that's doing integration testing, but they're like more like a hosting provider, you know? So it's like, it's, it's much less than, uh, any company that's doing integration testing. It's a company that hosts hundreds of websites, thousands of websites. Yes. And then this number, this number, it's interesting. **Jason Cohen** [36:49] Okay. Well, let's focus on that then. This is bad news, of course, but how many, how many, how many companies in the world do you think host at least a hundred websites and thus might want something like this, uh, potentially? They just potentially might want it. We don't know if they do, but just they're in the, they're at least in the category where they might want to use screenshots to make sure they haven't screwed anything up during a deploy or whatever, whatever the use case is for the integration testing. Just as a power of 10, how many companies do you think there are that have more than a hundred websites that they're hosting? **Dmytro Krasun** [37:30] I feel there is no more than 1,000 hosting providers in the world. It feels like. **Jason Cohen** [37:34] That's not true. **Dmytro Krasun** [37:36] You think it's more? You know for sure. Probably, yes. But it feels for me, like I can probably tell 50 maybe. It's like, you know, counting AWS and, uh, many other companies like. **Jason Cohen** [37:51] Those are the largest ones in the world, but that's not what I said. **Dmytro Krasun** [37:54] Yes, yes, yes, but, but for this, like, uh, and then it feels like. **Jason Cohen** [38:00] I went to a bootstrapping conference called, uh, Microconf that I'm sure you know. And there were several, several people at that conference alone who, um, had hosting companies with more than a hundred websites in it. And those are little bootstrappers going to a conference. So yeah, there's AWS, but I mean. **Dmytro Krasun** [38:20] Yes, yes, I meant like if I, if I take the top ones and then I go to, to the smallest companies. **Jason Cohen** [38:26] Let's just call it 10,000 because there's, there's more than a thousand, but let's not assume there's just a hundred thousand. That sounds like a lot. Let's call it 10,000,right? **Dmytro Krasun** [38:34] That's it. Okay. Okay. **Jason Cohen** [38:35] Yeah. Um, getting 1% of them to do integration testing like this sounds hard, but maybe plausible. Does that sound fair? **Dmytro Krasun** [38:47] Hmm. Either it's 10,000 of them. **Jason Cohen** [38:51] No, forget that. We already said there's 10,000. We're just saying that. Now, however many there are, does it sound plausible that maybe we could get 1% of them, however many there are, we could get 1% of them maybe to be convinced that they should use some kind of screenshotting technology as part of the deploy, presumably to say screenshot, uh, let's say a hundred sample websites of your, that you're hosting, run your deploy, then screenshot them again and make sure nothing's changed, roughly speaking, because we all know that's not quite how websites work, but let's just, let's, let's keep it simple just for, um, discussion sake. That's a good idea. We actually do that at WP Engine, by the way. So it's a good idea. Um, what, uh, do you think maybe 1% of people who have this problem could be convinced theoretically that that's a good idea? **Dmytro Krasun** [39:42] Yes. Yes. I, I, I think I could use already. I have customer stories related to this. **Jason Cohen** [39:48] Okay. So 1%. Do you think you can get 10% of them to say yes to that and actually implement it? This is every, every hosting company in the world with more than a hundred websites. **Dmytro Krasun** [40:05] It's a very good question. It's a very good question. Yes. Uh, 10%. I, I would need to try at least to reach out to a few of them to see like, you know, like to see how they respond, like how they work and, uh, like they're like, they're like approval process, how do they integrate? Because. **Jason Cohen** [40:28] Well, let's put it this way. They can't just buy your software. They have to build a whole integration system. **Dmytro Krasun** [40:33] Yes. **Jason Cohen** [40:34] Or, or retrofit their existing one in a significant way. This isn't like a quick afternoon,right? **Dmytro Krasun** [40:41] Yes. Yes. **Jason Cohen** [40:43] So can we convince, can we reach out, talk to, and then convince 10% of the market, however big the market is, um, to do all that? **Dmytro Krasun** [40:55] Yes. If I could solve already a lot of like integration issues around this, that when you integrate my product, yes. And, uh, assist them maybe and provide some like more features to, to make, uh, integration smoother. Yes. Yes. **Jason Cohen** [41:14] Really? 10%? **Dmytro Krasun** [41:16] No, no, no. The number itself, I don't know. **Jason Cohen** [41:20] Well, that's all I'm asking. Can you get to 10% of the market, do you think? **Dmytro Krasun** [41:25] Um, I feel it's pretty painful problem for them. **Jason Cohen** [41:32] Yes, but that's not the question. **Dmytro Krasun** [41:36] I'm trying to see like how, like how is it really possible to sell to them like by reaching out? **Jason Cohen** [41:43] That's not the question, whether it's possible to sell. I'm assuming it is possible, but like can we convince, uh, 10% of the market to build an integration testing system or retrofit their existing one around screenshots? **Dmytro Krasun** [41:56] Yes. Yes. **Jason Cohen** [41:58] Really? Okay. I, I, I would be shocked if you could get 10% of the market. Um. **Dmytro Krasun** [42:03] No, no. It's not, you, you said could we convince to do them this. Yes. Like. **Jason Cohen** [42:09] Yeah. I would be shocked if you could convince them 10% of the market to implement all this, because it's a lot of work and, and, uh, but maybe I'm wrong. Maybe, maybe the, after all, WP Engine did it, so we had the pain. So maybe I'm wrong. And, and like they, and at, at a certain scale, you just have to, and this is an easier way. So, um, I hear that. So, uh, I would guess it's 1%, but let's, but it's 10%. Let's say it's 10%. It's not a hundred percent. We can all, we can definitely agree on that. So it's 10%. Okay. So, so there's 10,000, there's 10,000 customers, uh, potential customers, and maybe you could get to a thousand of them, and then each of them pay what? What are the customers paying for this per month? **Dmytro Krasun** [42:48] Let's say. **Jason Cohen** [42:50] It's just a power of 10. **Dmytro Krasun** [42:52] Power of 10. **Jason Cohen** [42:54] Do they pay a hundred dollars a month, $10 a month, a thousand dollars a month? **Dmytro Krasun** [42:58] Yes. It's, it's between a hundred and a thousand. **Jason Cohen** [43:01] Okay. And so 300? **Dmytro Krasun** [43:05] Uh, for example, yes. **Jason Cohen** [43:06] Okay. So if a thousand customers were paying a hundred dollars a month, uh, that's your 100K MRR. **Dmytro Krasun** [43:14] Yes. Yes. **Jason Cohen** [43:15] So we just figured out a path to 100K MRR is that you have a thousand small hosting companies. Obviously, if you can get a big one that's paying you more, that's good,right? We're not, we're not, we're not saying don't do that,right? That's not what we're saying. It's just that, like, you can't, uh, depend on that as easily. But if you could get a thousand small hosting companies to implement this, you would reach your goal of 100K MRR on that basis alone. **Dmytro Krasun** [43:40] Yes. **Jason Cohen** [43:41] Now, again, I think 10% is too high, but still, what we've done by doing this is we've established that this is a path that's plausible. Plausible doesn't mean likely or unlikely. It just means it's possible given what we said. It is, it is at least possible. Most of the ideas that you or I would have for how to grow a screenshot API company, most of them will not even plausibly work. In fact, I think this one probably does not plausibly work because I don't believe in the 10% number. **Dmytro Krasun** [44:14] No, no, I just, I just wanted to go with you, like, you know, to, to restore, to explore this area, but it feels like. **Jason Cohen** [44:20] Yeah, yeah. So I think it's, it's probably more like 1%, in which case it's another 10K a month, which is still, uh, not trivial, but pretty far away from the kind of change you want. And so while it could be good to have, so that would tell me if you believe those numbers, uh, then with that, that downgrades it from, uh, the complete strategy of the company that, uh, which it would be if you think it can achieve the goal by itself. **Dmytro Krasun** [44:46] Yes. **Jason Cohen** [44:47] It downgrades that idea to, this is a decent idea, but only if I apply a certain amount of time and money. So if this is a certain marketing campaign, maybe even an outbound sales campaign specifically to those kind of companies. And I can't afford, I cannot afford to make 10 new features and lots of custom work for them, but I can do a little bit to gain that 10, to gain that 10K of MRR. **Dmytro Krasun** [45:14] Yeah. **Jason Cohen** [45:14] Um, that's the amount of effort I'm willing to put into this. And I still have not solved my root problem, but maybe I've pushed it down the road a little. That would, that's how I would think about that if those are the numbers. So the point, of course, is we're obviously not going to automatically find all theright answers on the call today. Um, that would be very lucky if we did, but this kind of mindset is how you can have an idea of like, okay, who are my good customers that I'm already able to serve? That's what we first did. We said this web integration, we said the cold email, uh, content and stuff like that. And then what is the possible, even possible market for this that I could get? And then if I do those numbers, is it even in the ballpark of the result that I want? If it's not even within a power of 10 of the ballpark of what I want, then even if it's just too, it's not the answer. **Dmytro Krasun** [46:10] Okay. **Jason Cohen** [46:10] Right. The numbers are just too small. Now, obviously we made up all those numbers and they're not exactlyright, of course, but if it's, but if it's off by a whole power of 10, it's just wrong. It's just not the, not the likely way forward. So, um, let's try it again with the custom emails just to, just to run it, run it again. So of course it doesn't have to be emails, but that does seem like the, that does seem like the most common thing. 'Cause if I'm calling on the phone, I don't know what I need a screenshot for exactly. So this case of, um, I'm doing outbound emails and I want to include a screenshot for whatever reason to get their attention or it's relevant, whatever. **Dmytro Krasun** [46:47] Yes. **Jason Cohen** [46:48] How many companies, for how many companies does it make sense for them to include a screenshot? Maybe they don't do it today, of course, but that it would make sense for them to do that. What, um, how many companies do you think do that? Obviously, millions of companies send emails. So that's a very, very big number, but how many of them want to include a screenshot do you think? **Dmytro Krasun** [47:13] It's a good one because I have like two different types of customers doing that. But if we talk about forever customers, like customers who perpetually like. **Jason Cohen** [47:23] That's all we're talking about because anything else is not something you can afford to invest in. **Dmytro Krasun** [47:28] Yes. **Jason Cohen** [47:28] Whether that means marketing, whether that means ads, whether that means support, whether that means adding features, you can't afford in that sense to, um, to do stuff for, uh, customers that don't stay. So they're, they're fine. We're happy if they come and they give us some money and they leave. As long as we didn't spend money on them and as long as their metrics don't poison the metrics of our ideal real customers, then we're okay with it. Um, but we're not focused on them. Can't be. That's not how you're going to get to where you want to be. Of course, 'cause their cancellation rate is atrocious. And that means you're never going to build, uh, um, the kind of business you want off of that. So the customers who do the emails, uh, how many really want to do that forever? **Dmytro Krasun** [48:19] I also want to answer something like 1,000, but, uh, yeah, but, but I don't have, and I don't have many of these too, you know, like it's like if you go by, if you go category by category, by use case by use case. **Jason Cohen** [48:33] Yeah. **Dmytro Krasun** [48:35] I'm not sure I have like one. I need to check this. I need to, I for sure I need to. **Jason Cohen** [48:41] If there was an obvious one, you'd know just, just by because you wouldn't need data to know if it was one jumped out at you,right? Okay. So let's, we don't have to go through that. A thousand is so low that that tells me you don't think people do this really. **Dmytro Krasun** [48:58] Mm-hmm. **Jason Cohen** [48:58] Because that's, because that's like 0.001% of all companies who send emails want to do it this way, which is, you're probablyright. In other words, people just don't want to do this. And you can find every once in a while, you can find a strange company that does it, but like this is not, this is not something to build a business around is what I'm hearing from you. **Dmytro Krasun** [49:16] Yes. Yes. It's like, for example, you want to email potential prospects about like how you want to improve their design of the website. **Jason Cohen** [49:24] I get it. But, but when you say a thousand companies in the whole world would want to do it, you're telling me either, um, this is, this is so rare. It's, you know, that, that just, it's just silly to even talk about it, or you're just so not confident that this is something people that, that people would want, even if you tried to sell it to them. **Dmytro Krasun** [49:46] Yes. I feel that there's not, not so many companies that really do it, but those, but those who are doing this, like they're forever customers. **Jason Cohen** [49:55] Yeah. So the, so once we start with a number like a thousand, that's the entire market. And again, you might be wrong, but the point is if that's your conviction, it's, there you go, like close enough. **Dmytro Krasun** [50:07] Yes. **Jason Cohen** [50:08] Then it doesn't really matter what the other numbers are. How many can you convert at what price? It doesn't matter 'cause it's just not enough. Like your whole company's a thousand customers now. So like that just isn't going to do it. So, um, so, allright. So let's take a step back and say, allright, there isn't a use case that's a healthy use case in a clear market that you can go target. There's just a bunch of scattered stuff people sometimes do with screenshots. **Dmytro Krasun** [50:37] Yes. **Jason Cohen** [50:37] It isn't really a pattern that leads to a market big enough to get to your goal. Is that possible? ### No Market **Dmytro Krasun** [50:44] This is the, this is one of the scariest questions, but, but still need to ask. Yes. It might be, it might be the real question. Like, like. **Jason Cohen** [50:55] I think it's more than it might be a real question. I think it's the problem. I think the crux of the problem is, uh, there isn't a big pull for getting screenshots via API. That's just not something that's very important to a really clear area of the market who needs to do that forever. That doesn't mean nobody wants it. People want it. You, you listed use cases where people want it sometimes temporarily, sometimes permanently. It does exist. It's not like there's never a, never a problem to solve. That's not the case. It's just there isn't a clear market where here's a use case where literally a million people do it. And so if you, you can definitely get a nice chunk of them, it's valuable to them, so they'll pay more, they'll do it forever. I know the, of course, forever is a big word. I'm just, you know, I'm, I'm trying to be brief about, you know, about it. Um, if until further notice, they'll do it because it's integrated with something important to them. Um, and it's just, it's just hard to see what that, what that is. And you're not seeing a pattern with your current customer base that would identify that. So how old is the company? **Dmytro Krasun** [52:02] Uh, four years. **Jason Cohen** [52:03] Four years. So in four years and many, many thousands of customers because there's a thousand now, but they've come and gone as well. In four years of experience doing nothing but this and seeing thousands and thousands of customers come through, you have yet to find an example of a use case where there's a decent number of customers in the world for that use case who want this, um, and want it for more than a few months. Is that true? **Dmytro Krasun** [52:34] Yes. It's pretty close. Yes. I, I, I, you know, every time I try something new and I have this big hope, you know, like, oh, maybe there's new, like for example, AI agents now, every agent needs screenshots. Yes. **Jason Cohen** [52:47] We're getting distracted now with AI. **Dmytro Krasun** [52:49] Yes. Yes. But, but, but. **Jason Cohen** [52:50] You have yet to find it in four years,right? **Dmytro Krasun** [52:53] But, but if you look at the. **Jason Cohen** [52:54] No, I, I understand what you're saying, but, butright now you have yet to find it,right? **Dmytro Krasun** [52:59] Yes. Yes. **Jason Cohen** [53:01] Okay. So this to me is the, but before we get to the solutions and AI, maybe AI changes stuff, maybe it does, but before we get to whether AI is changing it,right now you have not found a way or you have not found a market that cares enough to sustain this company in the way that you want to, which is to triple it. What you have done is built a company where probably you could continue running it just as you are now. And let's see, let's go up to these things. I'm guessing you could run a company where, um, this and this. Um, in other words, I bet you could run a company that makes, let's just call it 25K per month in profit, maybe 30K per month in profit for years to come because you could just do what you're doing now. There'd be a lot of temporary people coming in and out. That's okay. If sometimes people will stick because they have a weird ass use case that almost no one has, but they do. So they stick. Hooray. And so you stay at this, let's say three, 30 to 40K a month top line, 25 to 30K a month bottom line. You stay there because people still need screenshots. It's just scattered like it is now. And you just continue. And, and of course, incrementally you'll improve it. Incrementally, you might save some money, some costs and make, make profit go up. Incrementally, you might make a feature that makes more people stickier. Um, you might find, uh, like some of these use cases we identified, you could do some campaigns specifically for them and maybe pick up another 10, 20, even a hundred customers in that use case. So again, that's not tripling the company, but you could probably pick up some other good customers intentionally getting closer to that 40, maybe even 45, 50, pushing up a little by having better, higher quality customers, at least some of them. ### Option A **Jason Cohen** [54:57] I, I totally believe that this is a path. So let's call this option A. Option A is you look at this, the crux of the problem, which again, I think is, um, there isn't a good market for this. There's not a market, meaning customers who have a, like a sort of consistent sort of problem that they want to solve in this way. There's enough of them, they want to pay enough, all this kind of stuff I listed before. That's what I mean by a good market. There isn't a good market for this tool. Again, this is just a, this is the crux of the puzzle. This might, I might be wrong. And like you're about to say, AI may change it. I know we'll, we'll get to that in a second, but first let's really agree on what is the crux of the current issue. So I'm proposing that the crux is this thing of there isn't a good market. We have not, in four years and thousands of customers, it is not a, uh, a mar, a good market is not materialized, which tells us either it doesn't exist or it's so hard to find, we can't find it. Option A is, uh, yeah, that'sright. There isn't one. Continue to run the company. Like I said, incrementally improving, being a little bit smarter, maybe targeting a couple of use cases to get, to just change the customer mix a little, um, to, to bump it up a little. And in this way, incrementally improve the, the company. That's option A, incremental improvement. But actually, let me, let me pause there. Do you agree that that's the crux? No good market is materialized, uh, or do you not agree that that's the crux of the, of the problem? **Dmytro Krasun** [56:24] I feel that, I feel that it might be true intuitively because, because if I found. **Jason Cohen** [56:31] Let's put it this way. You can't name one. This whole conversation, you haven't named one yet. So. **Dmytro Krasun** [56:35] Yes. Yes. Yes. Yes. And that's why like it, it feels for me like no, because I, I, I would have, I would have one use case dominating. I would see it and I would, it would be easier for me to double down. **Jason Cohen** [56:48] Exactly. **Dmytro Krasun** [56:49] And, uh, yes. **Jason Cohen** [56:50] Well, that's why it's a crux because, because if it were obvious and easy, you'd be doing it already. Like you just said, the fact is it's not there. And so you're sitting there going, it's not there. I don't, I'm stuck. I'm stuck 'cause it's, you're clearly just doing the same thing that you're doing now will not achieve a hundred K in revenue in, in, in MRR. **Dmytro Krasun** [57:07] Wow. **Jason Cohen** [57:07] And, um, and it's not obvious what to do. Like that's, that is the crux. I don't know what to do because there's no good markets I found yet. So option A is incrementally running the company. Um, another op, another possibility is there is a good market out there for this, but you haven't found it yet, which means to find it, you will need to go do something intentionally that you are not currently doing. You can't just sit back and look at your customer base and say, where's the good market? You have to go seek it or even create it. This is another possibility. So it could be, let's, let's, let's talk about the integration testing again. I could make the case, um, anyone with a website in marketing should be doing this. They have a marketing website. Marketers change websites all the time. They run A/B tests, they change headlines, they add a new landing page, they do something weird with the homepage, they screw with the pricing page, and even more subtle things. And they don't want anything else to break as a result of them doing the one thing. So they should have some sort of what engineers would call integration test. They would not call it that 'cause they're marketers. And you're saying you could say there's 10 million marketing departments in the world. They all ought to do integration testing. They don't. Um, I'm not saying I can take 10% of the market or anything nutty like that, but I think I could be the champion of marketers actually doing, uh, regression testing. I am going to bring regression testing to the world of marketing. We know, we understand in, in software why to do it. Marketers should do it, but they don't know that yet. So I'm going to educate them. I will be the one Prometheus bringing the fire of, of, uh, regression testing to the world of marketing. Um, in doing, and so I'll have a product for that. It'll involve screenshots. And, um, in doing that, um, I'll, I'll literally create a new little category in marketing that'll just be me until other people realize it's a good idea, but that's okay. Um, and why screenshots? Because maybe I should be looking at the content of the page. Ah, the thing is you can look at the HTML, you can look at the CSS, you can look at JavaScript, you can have AI do it. And at the end of the day, you don't actually don't know what it looks like. Stuff happens. A lot of that code, by the way, is dynamic and does weird real-time stuff. You really, you can't look at the source code. You gotta look at the website. In fact, with the tool, really, I'm making this up now because it's just an example,right? **Dmytro Krasun** [59:41] Yes. **Jason Cohen** [59:41] In fact, what you need to do is show a before and after screenshot and show that to the marketer and maybe show 'em 10 pages of that, 10 pages that they've configured as being important. And the human being is gonna look at them because sometimes web pages are different simply because they're dynamic. So just asking, is the page different is actually also not sufficient. 'Cause let's say, uh, an image rotates on the homepage or there's a video. So the exact, uh, the exact frame the video's on depends, uh, when you take the screenshot or JavaScript is pulling something live. And so there's all these reasons why, um, the screen, the, it's not identical and that's okay. **Dmytro Krasun** [1:00:22] Yeah. **Jason Cohen** [1:00:22] And a human being would know that if they saw the before after, they would see the differences, um, maybe even highlight them, uh, for the human. Um, and the human could tell whether that's okay or not. And so it really, it does take that human still. Maybe it's AI in the future. I don't care. Um, so I'm gonna build a product that does all that and bring it to marketing. Now, is that a good idea? Do marketers care? Could you convince them of that? Could you make a good product that, that marketers could use given that you're an API company today? These are all questions I don't know the answers to. Maybe the whole idea is stupid, but what I'm saying, but, but the point is that this is an option B. ### Option B **Jason Cohen** [1:00:56] Option B is rather than, I mean, obviously run the company you have now because it's paying the bills and everything. So I mean, don't just stop. But option B is in order to achieve my goal, I have to build a different product, but I need to take my expertise and technology that I have into that because that's whatever that is, is gonna be part of my strategic advantage is my experience, knowledge, tech, et cetera. I obviously have to leverage that. What is a market that I could, that exists or that I think I actually could make, create that, um, that would, um, that would leverage that into a, into a product that doesn't exist or, or even if it exists a little bit that, you know, that you could make a good one of. **Dmytro Krasun** [1:01:39] Yeah. **Jason Cohen** [1:01:39] So maybe that marketing idea is good. Maybe it's bad. I don't know. Um, but this is option B is find another market that's actually good. Um, now the, how does that strike you? Like when I say all that stuff emotionally, is your reaction? Hell yeah. Like APIs are, you know, that's what I know, but goddamn it, it's dead. Like that sounds so fun. Or is your reaction something more like, yeah, you know, I get it. And if I were outgoing and I liked talking to people and I liked being a, um, and I liked being a, um, an evangelist for a new idea, I could see that, I guess, but like I just wanna be in the background doing APIs under the hood, talking to engineers, just going on Twitter and yapping with other indie developers rather than selling to marketers. I don't, I don't even trust marketers. I wouldn't like selling to marketers. Um, now obviously these two emotional states of I'm super excited or that sounds like death are not the only two emotional choices. But emotionally, when you hear option B the way I said it, how does that strike you? **Dmytro Krasun** [1:02:43] Yeah. It's interesting because I have this problem that I have many founder friends and sometimes I want to tell them about my product and I think maybe they could use it. Yes, but you know, it's very niche, it's very rare. And, uh, it's. **Jason Cohen** [1:02:59] Okay. But you're not answering the question. How does option B strike you emotionally? **Dmytro Krasun** [1:03:03] Yeah, yeah, yeah, yeah. **Jason Cohen** [1:03:04] Do you, are you like, oh yeah, hell yeah, something like that. Maybe it's marketing, maybe it's not, but yeah. Yeah. Or are you saying, or are you more like, I don't like that? Like just emotionally, not rationally, not what happens when you talk to some random founder who's not a customer. **Dmytro Krasun** [1:03:19] It's interesting. I would check it. It, it sounds like it's not hell yeah. It's not, uh, sounds like worry, but, but it's interesting. I would love to check it, you know, like to, to. **Jason Cohen** [1:03:29] Okay. So you're, you're not sure. That's fine. That, that's good. That's, that's a good answer, I think. Um, like it's, it's, it's, it's more useful when you already know that you do or don't wanna do it because it eliminates or, uh, highlights the option. So it's, it's great when, um, it's exciting or, or, or rejected. But, um, obviously it's valid and common to go, hmm, I'm gonna have to learn more before I know if I'm excited or not. Like that's, that's very valid. **Dmytro Krasun** [1:03:57] I could do it, but, but I'm not sure. Like it's, it's not like. **Jason Cohen** [1:03:59] Okay. That's a no, by the way, because you could do anything. You could, you could learn how to play, you could learn how to play the clarinet, but you're not excited to do that,right? Um, so, so, uh, but that's okay. You might get excited. It's totally true. **Dmytro Krasun** [1:04:13] Yeah. **Jason Cohen** [1:04:13] Sometimes we're like, I don't know. And then you try it and you're like, actually, this is cool. Like that, that happens all the time in life,right? So if the door's not closed, then what that means is it's a default no,right? Status quo is no. But if you're like, oh, but I'm curious, good. Then you go explore that because you, you absolutely could change your mind, um, with exposure. So that's, that's good. What about, let me go back to option A real quick, which is incrementally run the current business or run the current business, incrementally improving it, um, along these edges to make it better and be happy with the, let's, let's call it 30 K a month profit with a well, very well-run solo entrepreneur business. Um, how does that strike you emotionally? Are you like, yeah, that's who I am. That's fulfilling? Or, um, are you like, no, didn't you hear me when I said I wanted to triple the company? Like that's not good enough. Like it, maybe it should be, but it isn't. Like, like how does it, how's option A strike you? **Dmytro Krasun** [1:05:08] Um, I, I'll answer honestly. Like. **Jason Cohen** [1:05:12] I hope so. Otherwise, what's the point? **Dmytro Krasun** [1:05:15] I'm tired. Like, you know, let's say not super excited. **Jason Cohen** [1:05:22] Yeah, of course. That's, I mean, that's what you said at the top. So I'm not, I'm not shocked. It's still, like the point of these options, of course, is we're trying to list everything on the menu. **Dmytro Krasun** [1:05:32] Yeah. **Jason Cohen** [1:05:32] You're gonna end up picking something on the menu. Maybe you need to try it first before you know which, which, which one you wanna order. But of course, of course you want as big of a menu selection as possible first because, you know, the more choices that, the more likely theright choice is there. And also 'cause sometimes they intersect or collide somehow or there's some, there's some better idea that wasn't on the menu, but just going through the things on the menu, thinking them through and why do I like it or not, or what's the good or bad, just that can trigger some even better ideas. So it's always a good idea to put lots of options on the menu, even if you hate them, uh, they're useful in getting to an answer that you don't hate. So, okay, option A, you hate. Um. **Dmytro Krasun** [1:06:12] Not, not, not hate, but let's. **Jason Cohen** [1:06:15] No, but let's, let's just, yes, like, like it is, it's unfulfilling and you would, you do hate. Uh, I mean, it's fine for me to tell you what you hate, but let, let me explain what I mean though. **Dmytro Krasun** [1:06:24] Yes, yes. **Jason Cohen** [1:06:24] You would be unhappy if for five years that's what you did. **Dmytro Krasun** [1:06:29] Oh, it's pretty good. Yes, it feels like that. Yes. **Jason Cohen** [1:06:32] You would feel like you wasted five years of your life. In fact, maybe five of your best years. **Dmytro Krasun** [1:06:39] It's a good framing. Yes. **Jason Cohen** [1:06:40] How old are youright now? **Dmytro Krasun** [1:06:41] Uh, 30, 34. **Jason Cohen** [1:06:44] Right. So these are your best years because you are old enough to have experiences and some wisdom and, uh, but you're not, uh, old, you're not so old that, um, you can't think properly anymore. You can't, your attention spans, uh, lower. You need more sleep than before. Your body's starting to break down,right? Like you have a lot of wisdom, but man, you just can't, you know, this and that. Now, I, I'm not a fatalist. Of course, there's lots of things you can do to make it better or worse, but like also as, as, um, as Seinfeld says, father time is undefeated,right? So it's not good, not good news. So you're, you are in a, what you might say, the thirties are probably some kind of a optimum. And so you don't wanna spend your optimum the next optimum five years, um, incrementally running this business. What, whether other people agree or not or what they think about you, that's how you feel. **Dmytro Krasun** [1:07:36] Yes. Yes. **Jason Cohen** [1:07:37] And that's why you would hate it. 'Cause you would, even, even after the fact, you'd look back on that saying, I, I, I could have done anything else with those five years, even something that would fail. At least I would've been doing something instead of puttering around for five years in, in this mode. So, okay. So option A, not exciting. Now, what, what is interesting about that though, if I'm gonna digress a little into option A, is it is the easiest and clearest path forward is run the current business and make it a little better. Like clearly that's the easiest and clearest thing to do,right? **Dmytro Krasun** [1:08:07] Yes. Yes. **Jason Cohen** [1:08:08] Um, but you don't want to. So what, what's interesting is either you'll pick something else to do with this business like an option B, or let's, let's think of more options too. Um, but you might decide, I don't like any of these options, but what's nice about option A in general, like, is it's very clear how to do it. And so if I were to sell this business or otherwise, or maybe I don't sell it completely, I get a new partner and I, but, but I, I peel off or there's different ways of divesting yourself of the business, but still having some financial benefit,right? Uh, this is a business that someone else could run. It's one person could run it. **Dmytro Krasun** [1:08:51] Yeah. **Jason Cohen** [1:08:51] 'Cause one person is running it. Now, that's a little unfair because often one person, the, when the founder can run it, that doesn't mean that some random person can also run it. Of course, that's not what it means, but someone else who's also founder material, let's just say, could run it 'cause you've built a business that makes, um, a quarter million dollars a year in profit that one person can do, one very talented, devoted, uh, you know, experienced person. Yes. But still, that's freaking amazing. And there's many, many, many people on Twitter who wish they were in that scenario. Good. Maybe they should be. Maybe they should run this business. And there's lots of ways to do it. Maybe you continue to get a salary essentially off of it and, and it's funded that way. Maybe someone does buy it because that does happen, of course. Um, or maybe it's some combination where they give you some money now and some over time as the business supports it. Um, and that way you can get enough money in total, but not all at once 'cause that person can't afford it. There's lots and lots of ways that could go, but let's not try to solve thatright now. Let's just say there's ways. **Dmytro Krasun** [1:09:53] Yeah. **Jason Cohen** [1:09:54] Because you've built a business where, okay, it's not obvious how to get it to a hundred K. Maybe it's even impossible. Maybe, but it sure as hell can go from 30 to 40. And I would say from, from, from, from 40 to 50, like, like you said, not with the current customer base, but there's enough use cases, enough ideas. You could get some other better customers, not completely transforming the whole business, but just incrementally and get it to maybe 50 K and, and 30 K in, in revenue and 30 K a, a month in profit. I believe that entirely. Um, so that's, that's, that's an ideal business for many, many people. So that's one of the great things about option A is there's other people who would love that. That's good because that, that gives you a way out if in fact leaving the business is what you end up wanting to do. ### Option C **Jason Cohen** [1:10:44] Um, so I suppose we could call that option C. Option C is sell the company or otherwise leave the company, but don't, um, leave the company and get some sort of, um, payment for that. Like I just listed different ways you could get payment for it. So we won't, we won't try to stipulate what that is. But option C is you're like, you know what? I just don't wanna do screenshots anymore. I've done it for four years. I'm tired of it. But this is a good little business. So I wanna out, but I wanna get paid somehow that somehow that's, that works and is fair, but I wanna move on. I don't know what, move on to what, but I'm just done with screenshots. I just don't wanna do screenshots anymore. Like even if I could do that stuff with the markers, I just don't wanna do screenshots anymore. So how does that strike you when I say that, when I say I just don't wanna do screenshots anymore? Do you think, yes, exactly. That's exactly how I feel. Or are you like, no, I don't hate screenshots. I just, uh, I just need a product that's working better. **Dmytro Krasun** [1:11:46] This one, it's interesting because even if we talk about option A, yes, this one is the most exciting in a sense. It's interesting. **Jason Cohen** [1:11:57] Option C? **Dmytro Krasun** [1:11:58] Yes. Kinda, kinda, I feel like I can kinda, you know, like, um, transform myself to like it for more time, you know, but it feels like if, if it, if we talk long term, like you said, about five years, let's say if I still run this and I still do option A for five years, okay, it's, it's cool. It's nice small business, as you said, but it feels there's so much more to do. **Jason Cohen** [1:12:25] Okay. And then what's the answer to the question? **Dmytro Krasun** [1:12:28] Yes. C. **Jason Cohen** [1:12:30] No, but my question was emotionally when I said, I just don't wanna do screenshots anymore. Are you like,right? That's how I feel. I just don't wanna look at it. I just don't wanna think about a screenshot ever again. **Dmytro Krasun** [1:12:44] No, no, no. Uh, I, I, it, it feels, uh, I, I can do it. Yes, I can do it. **Jason Cohen** [1:12:54] I know you can do it. You can do anything. You can learn to play the clarinet that poorly pro probably. So please don't do that. Um. **Dmytro Krasun** [1:13:04] Uh. **Jason Cohen** [1:13:07] When I say I don't, when I say, goddamn it, I never wanna see another screenshot again, are you like, hell, exactlyright? Or do you say, no, screenshots are fine. I just, I just need a, but, but this, this product, meaning the market and the customer is selling, this product isn't working and that I can't abide. But, um, no, screenshots are fine. It's just, I just, uh, I need a market that's working. **Dmytro Krasun** [1:13:31] Yeah. So it's fine. It's, this one resonates more with me. Yes. Like it's fine. It's okay. Like it's just. **Jason Cohen** [1:13:36] It's fine. Okay. **Dmytro Krasun** [1:13:37] Probably if, yes, like. **Jason Cohen** [1:13:39] Here's, here's what, here's why that's important because, um, I hear, I did hear you that option C, sell the business and just do something else is most exciting. Gotcha. But here's, here's why this, this emotional thing's important. If you were to say, I'm just done with screenshots, then that means option B is also off the table. But if you say, I'm done with this product, but the technology and my expertise, that's fine. Then option B is still on the table. 'Cause option B is, of course, not really selling screenshots. It's selling regression testing to marketers or whatever, some other idea. **Dmytro Krasun** [1:14:17] Yes. Yes. **Jason Cohen** [1:14:18] Screenshots just happen to be, um, how you deliver that value. **Dmytro Krasun** [1:14:23] Yes. Yes. **Jason Cohen** [1:14:24] And so if you hate screenshots themselves and you ever wanna do screenshots again, then you don't want a product that's still doing that,right? But if you're like, no, no, no, that's fine. I just don't wanna sell the product of API screenshots. I wanna sell a different product, but screenshots, that's fine. I would say, oh, good. Then, then option B is still on the table because you just said there's so much else to do,right? If there could be something else to do that's unrelated, that's fine. That's option C. Option C is get rid of it so you can go do the unrelated thing without being distracted by this company,right? But it could be that the totally something else is still something where your current expertise and technology helps you get an edge. So like it could be that the next thing you do is play the clarinet, but then all of your current experience doesn't give you an edge. So you're just like everyone else who's learning to play the clarinet except you started late. It's actually kind of bad,right? Um, whereas if you do something that leverages what you currently do, you're ahead of the competition,right? 'Cause you're leveraging some asset that you have, which is your experience and your tech and all this stuff. So if you don't hate screenshots per se, then it's, then option B is still possible, possible because that could be that next totally different product that just happens to, uh, leverage the stuff that you, you are an expert in, and, and thus you can build that product quicker and better than a competitor who also had the same idea or wants to copy you,right? 'Cause like, let's say you did the marketing thing and now it's January of 2027 and now the marketing thing is out there and then someone else goes, oh man, that's a good idea. That's easy to do with Claude. And then they build it in Claude. Um, they're wrong because as you know, screenshots are actually really hard to do and get rid of all the bugs and all stuff's actually quite hard. And so, um, they don't know that. So while they're struggling for the next year or two to get the bugs out, you're adding the next feature and the next feature and you're just way ahead. You see what I mean? But why? Because you've already, because you're leveraging this asset and experience that you have, which is this knowledge and tech around screenshots. You're leveraging that. So that keeps you ahead, at least for a bit. Like, okay, it's not, you know, you haven't like invented Uber or anything, but, but this is good. This is a good kind of, uh, strategic advantage. It's pr, it's good. So this is what I mean. If you don't hate screenshots, option B becomes possible because that is a different product, but leveraging something you have there, thereby making a great strategy for you as opposed to option C, which is I gotta get out of the, I can't look at screenshots anymore. It's the end. I just can't. I'm, I'm burned out on screenshots. Okay. Then you need a clean break. That, that's that. So I, what I'm hearing you say is option C is the most interesting thing because there's so much more new stuff, but option B is not impossible because maybe that new thing that I wanna do, maybe if it could leverage what I already have to give me an advantage. Okay. Like, yeah, I guess that would actually be even better. **Dmytro Krasun** [1:17:24] Yes. Yes. Yes. Yes. **Jason Cohen** [1:17:27] Okay. Allright. So we've, we've established. **Dmytro Krasun** [1:17:31] If a plausible option B exists. **Jason Cohen** [1:17:33] Yeah. Yeah. Well, I mean, well, the bottom line is you said I wanna do, you said option C is exciting because there's so much more to do, which I, I took to mean there's so many other product ideas, company ideas that you could be doing that you would have a lot more excitement for. Is thatright? **Dmytro Krasun** [1:17:49] Yes. Yes. It feels like also like when we talked after the whole conversation, for me, it's obvious now that option A, it would be like we wouldn't have this talk probably. You know, there's like option A, just continue to do whatever I'm doing. It's like. **Jason Cohen** [1:18:03] Well, option A is not to continue doing whatever you're doing. It is to incrementally improve the business. So like you weren't thinking about, or you weren't,right now you don't have an outbound campaign where you've identified people who send emails with screenshots and you've, and you've identified a group of, uh, a thousand of them and you're targeting them with a marketing campaign to try to get more of them. You're not doing thatright now. So I'm saying you could. So option A is yes, run the business, but incre, but, but get much more targeted and tactical about how you're incrementally improving it so that you do bust beyond this 30 to 40 K revenue, uh, ceiling. And I can stop sharing this, I guess. Um, you do, you do bust beyond that ceiling. You're not gonna get to the a hundred K probably, but you get past this little, this, this little lump because you're being smarter about identifying these valuable use cases, targeting them better with, uh, marketing, uh, segmenting out the customers that are temporary. You're happy to have them because they represent a lot of revenue and profit every month. So no problem. But you get much smarter about who you're targeting, where you're spending your time and money on, and support and features so that you do evolve the company in a way that gets past this, um, this, this, uh, this current, um, uh, uh, plateau. So option A is not running the companies now. It's everything I just said, being a lot smar smarter about incrementally running the company. However, I agree it, it probably doesn't get you to a hundred K because, um, because the markets just aren't there,right? They're not, there's not that many good markets or there's not any good markets. There's just little pockets of things, but you could still be smarter about it. That's option A. But I still hear you that you don't wanna do it. No, no problem. I just wanna be really clear that option A is not just, just running the business as now. It's more than that. It's just I hear you that you, you say, yes, but I don't wanna do that,right? **Dmytro Krasun** [1:20:03] Yes. Kinda. Yes. Yes. Yes. **Jason Cohen** [1:20:05] But the next person, if you were to do option C and sell it, you would expect the next person to do option A, not run the business as it is now, but do option A, run the business, but better. **Dmytro Krasun** [1:20:17] Yes. Interesting. **Jason Cohen** [1:20:20] So, okay. Option A is that, option B. So I guess, uh, we won't do this now, but when you think, when you say there's something new I could do in the future, um, what, uh, uh, what I would do as you ex, as you, as you think about what that could be, um, if there's a way for screenshots to be a part of it, even a small part, like the marketing thing, that will give you an advantage. And so that would make it a better product for you. So it's not a requirement that you do that. Of course, it's, of course it's not, but it would be better because it would be the, the more you lever, the more you leverage experience and, and even technology that you have, the more that strategy is correct for you and the more it's something you've never, you know, none of it is something you're, you're, you're even, you're talented at or good at. Well, it's not clear why you are even doing that. So,right. So of course that's not a hard and fast rule. I'm just saying that would be better. **Dmytro Krasun** [1:21:21] Yes. Yes. **Jason Cohen** [1:21:23] So you were, so before we were talking, you, you were about to talk about AI and how that changes everything. So I'm guessing you're gonna tell me because of AI agents, AI needs screenshots. Therefore, if you were an MCP server for AI getting screenshots, that would be potentially amazing because then AI could use it a lot, even, even, uh, more than people know, the users of them know. And, and that could be a whole nother way to have the current API product breathe new life because it's a brand new, uh, use case. Is that what you were gonna say? Or tell me if not, what, what is the AI, uh, case you were gonna make about how that's gonna change everything? **Dmytro Krasun** [1:22:02] That's, no, no, no. That's how I thought, let's say maybe, uh, six months ago that it, it could change. Uh, but, but then I had customers who were using, uh, screenshots with, within Codex, Claude Code, and with agents. And then I realized, ah, okay. So they also, there's like, it's just a tool, you know, like they're using this, uh, store tool and there's like no one use case. It's not like this AI agents, it's not a use case actually. Also after we talk about all this, like I see the same example, marketers are using Codex and then use screenshots for some analysis of some websites for a short period of time and then they charm. It's not like it's kind of changing. Like I see many new customer profiles like product managers, marketers, designers, but it's not like, you know, they also, they also charm a lot. And, uh, it's not like, uh, some repeatable, scalable, I, I, I could say this is like agents, like AI agents use case, but it's not, it's not a use case because how people actually use it, it's, it's just a tool. Like it's, it's the same as I would say like, uh. **Jason Cohen** [1:23:20] Wait a minute. Um, do you have an MCP server? **Dmytro Krasun** [1:23:23] Yes. Yes. **Jason Cohen** [1:23:24] And nobody uses it? **Dmytro Krasun** [1:23:26] Uh, a few customers like use, they use it. Yes. Like, but, but not like, uh, a lot. No. **Jason Cohen** [1:23:33] And you advertise the MPC MCP server. Like if I search on Google for MCP screenshots, you're gonna appear? **Dmytro Krasun** [1:23:46] I don't know. It's a good question. I want to check this one. MCP screenshots. Um. **Jason Cohen** [1:23:56] Well, let's see. Uh. **Dmytro Krasun** [1:23:59] No, I think it's like. **Jason Cohen** [1:24:00] This is gonna be fun. **Dmytro Krasun** [1:24:02] Yes. No, no, no. I'm not appearing. **Jason Cohen** [1:24:06] Okay. MCP market capture and process screenshots with MCP. Screenshot as a dev tool. Okay. There's a GitHub one. Cursor community has one. Take screenshots with Playwright MCP. Oh yeah, that makes sense. But you have to have Playwright. So, okay, that's fair, but you have to have Playwright. Mm. Pulse MCP enables AI to capture and analyze screenshots from the desktop on demand. **Dmytro Krasun** [1:24:31] No, no. Oh, it's, it's desktop. No. **Jason Cohen** [1:24:35] Allright. Guide to taking screenshots with a server. Page pixels. That's fun. Okay. Let's just for fun see what this is. **Dmytro Krasun** [1:24:46] Yes. Yes. **Jason Cohen** [1:24:48] Learn how to connect. Uh, got, let's see. Uh. **Dmytro Krasun** [1:24:58] Yes. It feels like a bug. **Jason Cohen** [1:24:59] Yeah. This looks like you,right? **Dmytro Krasun** [1:25:01] Yes. It could be. Yes. Yes. **Jason Cohen** [1:25:02] Okay. So this is a competitor. **Dmytro Krasun** [1:25:05] Yes. Yes. **Jason Cohen** [1:25:06] And they're, they're, and you're not there. I didn't see you yet. **Dmytro Krasun** [1:25:10] Yes. Yes. **Jason Cohen** [1:25:11] Let's see. Where are you? They're there. There's a lot of people chatting that didn't say you. You're not on the first page, which means you're not there. So, um, okay. So obviously Google's not the end all. In fact, it's much more likely people would search, I think, in, uh, in AI for this,right? Rather than Google. So let, let's just say that little test is not conclusive, but what I would say is you're probably not giving it a good try yet. **Dmytro Krasun** [1:25:38] It might be. **Jason Cohen** [1:25:39] Right. Like it, like what I would wanna hear from you is, um, while I have 30 K of MRR from human beings doing screenshots and, you know, maybe I can go to 40 K, stuff like that. That's fine. Here's where the magic is. If I can get someone to configure an MCP server in their AI system, don't even say the word agent. It doesn't matter. Codex and Claude Code. Maybe it's an agent. Maybe it's just their Claude Code project. I don't care. If I can get 'em to do that and I can get 'em to wire that up, then AI just uses it. Like the person may not even even asked for a screenshot, but the AI decided, I need a screenshot,right? It goes off and gets it,right? And so, uh, that's the future because how many people are gonna do that, um, in the future? Uh, everyone, whether they know it or not. Okay. Or whatever it is, it's an, it's an unthinkably large number of people are gonna be doing stuff like this. Like even just saying what's the size of the market is sort of silly. It's like, I don't know, like a lot. And so the question is, when someone searches for MCP screenshot, I need to come up so that they do it. Because if I can get embedded in thousands and then tens of thousands of projects, yes, many of them are temporary. Yes, many of them only do two screenshots a month. And so they'll stay on the free plan or whatever. But that's how I'm, that's the, that's the future. That's how I'm gonna get in the future. And, and actually like maybe sometimes people for months don't do much, but then suddenly they just happen to do something with AI that requires a lot of screenshots. And so they end up paying me that one month, you know? **Dmytro Krasun** [1:27:21] Yes. Yes. **Jason Cohen** [1:27:22] Right? 'Cause their credit card's on file. So that one month they run up a bill, but then the next month they're not anymore. But they're a customer in the sense that they have an account with me. I have a more flexible pricing plan where it's pay by usage with a free tier. So most of the time it's free, but then they, they use if it's variable, fine. And when that's plugged into AI, it's easier for that to be variable and it's easier for that to just stay in. **Dmytro Krasun** [1:27:45] Yes. Yes. **Jason Cohen** [1:27:46] In fact, if person A configures the system and person B uses the system, person B might generate a lot of screenshots, uh, and not even know it because person A's the one who made it and person B doesn't even know exactly what the AI's doing or why. That's fine. This is all, this is the future. You could say I'm just, I'm just, I'm just, uh, very verbosely describing what this strategy is, what this future is, this view of how the company could transform. And so step one of all that, of course, is when someone intentionally searches for MCP screenshot, you gotta show up so that you, they pick you, they, you know, where they even have a chance to pick you. And you're not, you're not really like intentionally doing that. So it's possible that AI does save it and AI does get you to a hundred K and, and beyond with, uh, being visible that way. Maybe you need a CLI as well as MCP 'cause a lot of people don't like MCP, but okay. You, uh, clearly you can do all that. That's, that's, that part is in fact not hard actually,right? So you have all the ways. You have the MCP way, you have the direct API way, of course, which you already do. Maybe you have an SDK. **Dmytro Krasun** [1:28:52] SDK, actually. Yes. Yes. **Jason Cohen** [1:28:53] Yeah. You have an SDK for, for, for TypeScript and for, um, and for Python so that AI could get it that way. **Dmytro Krasun** [1:28:59] Yes. **Jason Cohen** [1:28:59] You have a CLI. Okay. So you have all these different ways that AI likes all these ways so that no matter how you do your AI, there's a way, um, an easy way for you to wire it up and then you promote it that way. You, you market it intentionally that way to try to get people to do that. And those Reddit threads that I just saw on Google, you need to come up there. And if you're not, then you need to say that you need to be watching Reddit and answering eventually. Hope, hopefully other people do that for you. Of course, that's better. But this is what you would do if this, so let's call this option D. ### Option D **Dmytro Krasun** [1:29:30] Yes. **Jason Cohen** [1:29:31] Is, um, it's the current product under the hood, but it's AI. AI is the future. AI's the way. So the current business will still be there. And that's how you're gonna maintain the business and maintain the profit and maybe go from 30 to 40 K just by natural activity. Um, under, you know, that, that's fine. But option D is yes, I'm not gonna, you know, stop doing that, but, but the way this company's gonna get to a hundred K is AI in, including AI, which means I need every integration possible. I need to be visible in a marketing sense in every way. Um, in, in those AI ways, I need to make, I need to talk about AI. I need other people who are influencers in AI to talk about me, all that kind of stuff. That's what needs to happen because I'm gonna grow through AI integrations and a pricing plan specifically for that, which probably is, uh, again, variable so that people can use it as much or as little as, as, as makes sense or not. Like, I don't know, you'd have to talk to customers to figure out what theright pricing plan is, but maybe something like that would make sense. So it's easy for them to sign up, but also if AI happens to use it a lot, you get paid. So, um, this would be option D. And again, like what I just articulated is unlikely to be exactly theright plan of action because we just made it up,right? But something along those lines where the strategy is get embedded in AI and let AI do the activity for you, um, with everything else, uh, behind that, um, supporting that integrations, pricing is, and, uh, scaling and so on. Um, that's option D, let's call it. It's time to, it's time to plug into AI. How does that strike you? **Dmytro Krasun** [1:31:16] It's exciting. There is one caveat. Like I'm not, I'm not sure 100% that it might work in a way because when I myself personally use Codex or Claude Code, yes, they have access to my browser. They have access to all these tools. And how often do I really need screenshots in this use case? For example, if, if we talk about personal usage, yes. But if you talk about, it's, um, it's interesting. Yes. It's, it's one of the options I would consider at least because I didn't try. **Jason Cohen** [1:31:50] What you just did is you articulated a very good part of your sales pitch. In other words, your sales pitch to someone who says, "Why would I need that? I can just tell AI to take a screenshot." And you would say, "Well, let's think about that for a second." If it's on their server, they're not taking a screenshot. They're like pulling through the HTML. **Dmytro Krasun** [1:32:11] Yes. **Jason Cohen** [1:32:13] And often they're like, "I got a 404 error." I get that all the time. Or a 500 error doesn't let me do that or it doesn't know how to render the site. Like all the time, it, it doesn't work. And then you say, "No, no, no, but I can run it locally and in my browser, it can take the screenshot." I would say, "Ah, that's true. All you have to do is give it access to your browser, which means literally everything in your life, which is incredibly scary and in many corporate contexts, not allowed,right?" **Dmytro Krasun** [1:32:43] Yes. Yes. Yes. **Jason Cohen** [1:32:43] And so, and so in many corporate contexts, that's not an option physically 'cause they don't allow it. And just in general, that's not such a good idea, you know, just for a screenshot. They have, it also has access to all your email and your bank just 'cause you wanted to screenshot an Amazon listing. Not sure that's a great idea,right? And so, um, so this is, so what you just did is articulate a sales pitch, which I like, a good sales pitch. I like it, but it did not answer the question, which was, how does it strike you in terms of emotionally? What is, how exciting is that? Are you like, "Hell yeah, let's go get a, let's, let's go do AI screenshots"? **Dmytro Krasun** [1:33:20] It feels like this. It feels like one of the most exciting options, let's say out of four. It's pretty, I really like it. And, and because I also, I, I didn't invest in it, you know, like I didn't, uh, I didn't promote it much. It's, it's interesting at least to try, at least, uh, at least to be in the first position and to see what happens. It feels, I cannot say it's like 100% yeah, but, but, but it's one of the top. It feels like, uh, it, it, it would be one of the top priorities to, to test. **Jason Cohen** [1:33:57] Okay. Good. If, if, um, if you just went with option C and sold the business and never tried, never even tried for a week to do the AI, do you think later you'd regret not trying? Or would you say, "Eh, I mean, I had a better idea anyway"? **Dmytro Krasun** [1:34:17] Um, I might regret because like this kind of product and, uh, this like step, adore, adore in this AI experience, AI market, like I already have it. I can just open and check like how's it going. But starting from scratch again, like entering this, you know, like AI agents and all this. I, I, I think it, it, I think it will be painful experience to, to sell it and then start again from scratch without anything, you know, like. **Jason Cohen** [1:34:50] Yeah. **Dmytro Krasun** [1:34:50] From zero. Like it feels like I might not regret it, but it feels, yeah. Okay. Regret. Yes. Okay. **Jason Cohen** [1:34:59] Yeah. I mean,right. Because starting from scratch, of course, it always sounds exciting. I mean, we're, we're, uh, we're, we're entrepreneurs. So like, of course we like starting a new thing. Like that's the part of what it is,right? Um, and on the other hand, we remember, ah, but most of the time it doesn't work at all. And it's really hard. Even when it does work, it's really hard. It's a hard scrabble. Like to get to where you are now is hard and unlikely. So doing that again is like, ah, like it's really exciting. And also, god dang again, like, oh man, I don't know. I, I already got here. Do I really have to start over? So it can be again, emotionally it sounds like you're on the fence there 'cause it is exciting to start over, but you can see, you also see that emotional thing of, man, I have all this. Like if, if that next thing could still be this thing, but, but pivoting, um, that would be way easier than, you know, like, um, it sounds like you're also there. Like, and again, like it's okay not, not to really know. I mean, it's a good thing to ponder,right? It's a good thing for you to continue to ponder 'cause it's theright question. Okay. So, um, what's nice about option D is you can absolutelyright now go try it and do better than you were doing. In other words, um,right now you could go through the top or you could go through the first, let's say three to five pages of Google listings and Kaji listings and DuckDuckGo listings for this. See what kinds of things show up and start to get a feel for like what would I need to do? What do I want to do? Where do I need to be seen? Should I go talk, should I go leave a comment on one of those Reddit forums today? I, I just, I, you know, probably won't work to leave one comment, one forum, but if it's the third search result on Google, it might be worth it. Um, is there now, here's where I do actually think that with AI coding, you could do some things very quickly. Like you have your MCP server, but maybe it needs better documentation or a certain GitHub repository. I don't know. Well, with AI, you could get all that real fast,right? Maybe you have it already, but if you didn't, you could do that really fast. Um, same with a CLI. Oh, I also need a CLI. You could do that in an afternoon with Claude because you've already done all the other work,right? Because you already have the API, you already have SDKs, you already have SDKs. So like because of that, comma, you could, you could do the CLI that, that Claude likes, um, in an afternoon and put it up on a GitHub repository or, or include it in your existing one or whatever. Absolutely. You could do that in an afternoon. So there's some things like that where you could like, you know, in a couple of afternoons, it could be really pretty solid in certain areas like, um, integrations. You could, you could also say, you know what? I'm gonna have a plugin, sorry. I'm gonna have a Python, um, package. I'm gonna have a TypeScript NPM package. I'm gonna have maybe one other thing, maybe not. Um, and again, Claude could, given that you already have all the stuff and it's just wrapping an API that already exists, Claude Code could absolutely make those and you just get accounts with theright places and, and, you know. So in other words, you could have this nice array of integrations that people who do AI generally like. And I know you have some of these already, whatever. I mean, by the end of the week, you could have those, or at least by Monday, you could have those, which is a pretty strong offering in, in terms of the integrations. And now getting the word out is okay. That's a little harder, but you could just start doing stuff. You could start, you have a big Twitter account, you could start tweeting about it, you could start tweeting at certain things. You could, um, you could, again, you could get on some of these Reddit things. You could post your own things there. You could, um, is there someone who also makes YouTube videos or tweets or whatever where they're also technical and I don't know, just the kind of maybe you know them already and you could say, "Hey, I'm just trying to get this MCP thing off the ground. Maybe we could do a podcast together or maybe, maybe just as a friend, you would, you know, make a tweet about it just because like we're friends and help, help me out, you know, like you could." So obviously none of these things are quote unquote scalable or how you would, you know,right? Of course not. But that's not what you're doing on step one. Step one is how quickly can I get a decent product together quickly, I think in your case. Um, and, and like what could I do to test the waters? Just see, um, can I drum up a, drum up a little bit of attention? What is that like? And of course, then talking to those people. So you have some people using the MCP server. Now you need to talk to them with a structured conversation. You know, I have articles on this and everything about how to do that to understand them. And again, you could do some of this, let's call it quote unquote, you know, manual marketing. And, and again, if that generates even a little bit of activity, great. Like try to talk to them again, understanding this better. So you get a better view of like what is this? What could this be like? So that's, that's, that's what I think of when I think of how would you test option D? You don't have to select it, but like in the next, so it's July 1st now. By the end of the month, could you have fully tested whether option D is good? And also, um, ### Next Steps **Jason Cohen** [1:40:07] I also have in greater detail what is option D? Like we know what it is generally plug into AI, but there's all these open questions. What for, how would I find customers? What language do they use? What kinds of integrations do I need? What use cases should I talk about that are more likely for people to actually need in the context of AI specifically? 'Cause it's not necessarily the same use cases. So what are those use cases? The people that do use AI, why? And, and, and, and,right? And so by the end of the month, could you know, uh, here's what option D would be, and here's some answers to some of those types of questions that would fully articulate what option D really means. You could decide that's my goal for this month, period. I'm gonna run the business the way it is. Uh, obviously, like if there's something catastrophic, I have to fix it. I have to keep up with my tech support tickets. I can't stop doing marketing 'cause this pays the bills. But other than running the business at just the lowest possible level that's not harmful to the business or your customers, every other second is gonna be this month is gonna be spent vetting option D with the kinds of questions I just said so that I, so that I can decide A, what is option D? Exactly. What are we talking about? And B, am I super excited to go do that or not? Or, or after all of that exploration, am I like, yeah, I see it now and, uh, gosh, no, I don't want to. Like,right? That's possible. But like maybe, maybe you need to go see in order to find out,right? **Dmytro Krasun** [1:41:42] Yes. See. Ah, see, I, I, I, I, I thought about option C. **Jason Cohen** [1:41:49] Oh yeah. Maybe you need, you need to go explore in, in, in order to find out like what is option D and do I like it? **Dmytro Krasun** [1:41:58] Yes. Yes. It's a good one. **Jason Cohen** [1:42:04] So that could be a good plan for this month. Run the business at a skeleton level, basic level, and spend all of your time vetting option D, meaning that you're, um, doing all these things I said. You're creating this full integrations, you're talking to people, you're doing manual marketing to drum up some attention and just seeing what that's like. How easy is that? What's the feedback like? You're interviewing both existing customers who use MCP and these new, this new attention. You're doing interviews, um, or otherwise finding out answers to all these questions. You're gonna get a list of questions. You're gonna make a list of questions. Again, I have that, I have that, uh, article about how to interview customers,right? So you're, you're gonna have these goals like the ones I just listed. You'll come up with your own as well. And then you're gonna come up with, um, your hypotheses of what, um, what are the answers to those? You know, half of 'em are wrong. That's okay. Then you're gonna invent really good questions and then you're gonna execute that. This is what the article lays out as, as I'm sure you know already. Um, so you're gonna do that. And that's what I mean by vetting option D,right? Like you build integrations, you get some attention, you do these kind of structured interviews. Even on Twitter where it's not so structured, you can still ask these questions. Like if you have good, if you generated good questions with that method,right? You have, again, a big Twitter following. You can still, um, ask some of these questions in general. Now, of course, asking in the general Twitter audience, you're gonna get random answers. That's not very high signal compared to a real customer. So of course you're gonna modulate your, uh, what, what you do with that information accordingly. However, it's still information. It's still good, especially at this stage. That's still good. And you're leveraging something you have, which is a Twitter following, um, to get some more information. So you could do that too. And so I just made up one month just because it's, it's, um, if you only did it for three days, you wouldn't really find out. And you don't wanna be exploring for six months because it shouldn't take that much time and you don't have that kind of time and it just doesn't make sense. So it doesn't have to be the end of the month, but just a convenient something like that. Uh, you know, something like three weeks, four weeks, five weeks, something like that makes sense to give it a good, uh, attempt, especially with interviewing customers 'cause it's hard to schedule. Oh, I, I could do something next Tuesday, then on Tuesday they don't show or they reschedule,right? Like that's what happens when you try to talk to customers,right? So you need calendar time in order to actually pull that off, it turns out. So that's what's in my mind when I say something like, like, uh, four weeks or even longer. But, um, but you could set an, an intention of like that's what I'm gonna do. I'm gonna, I'm gonna spend all my time that I can, um, vetting option D. **Dmytro Krasun** [1:44:53] It expands like and like putting the business like on a slow burner, let's say, like. You know. Yes. Yes. **Jason Cohen** [1:45:04] Otherwise you don't have the time to really do it. You need to be doing it for hours a day. **Dmytro Krasun** [1:45:09] Yes. Yes. I have. **Jason Cohen** [1:45:10] 'Cause you're really trying. **Dmytro Krasun** [1:45:12] Yes. Yes. Actually, whole I have all the stuff that you mentioned, like CLI, MCP, like integration page, like everything, but I never actually really promoted it. Like I, I never was focused on it, like, you know, 100% like I'm just doing this stuff and that's it. Like I didn't do it that way. I just, you know, I shared a little bit here and there and I found some customers and, oh, okay. Okay. So it's like I, I didn't see that way. **Jason Cohen** [1:45:40] Yeah. And, and, uh, and that's fine. And so, so you've already built it. So I, I think what you could do though is you could say, um, well, I built it, but now that I'm thinking for as a marketer, I not only want to document there is this MCP, there is a CLI, there is this Python library, I want to convince someone to use it, which is different from just documentation. It's both. It's both. But like the headline of the page is different when it's just reference material versus when it's trying to convince someone to implement this for some reason. So when you get into the details of how the API works, then it's just technical documentation. Fine. But that's not what the homepage is. I say homepage, but the main page that you land on when you search for MCP screenshots isn't just a boring technical document. It's marketing. So I think what you could do is say, allright, I basically have all the tech. I'm gonna take a pass now. I'm gonna think about it as a marketer now and take a pass over that saying, wait, how do I get someone excited about this? Not just tell them what it is. And, um, and so again, maybe that's just an afternoon,right? Like does that feel a lot, but like, but you wanna do that. That's gonna be part of this is it's marketing. You're gonna wanna go to your actual homepage of the whole, um, company and also talk about this. Uh, maybe there's special landing pages for it, or maybe, maybe the MCP stuff is the landing page, but anyway, you're gonna direct some traffic there somehow. Um, SEO obviously is hard to, it's hard to do quickly. That's one of its disadvantages is that you don't just rank really fast. Nevertheless, you could try by thinking about SEO for those very pages and also with, uh, um, things to write or maybe even guest post something 'cause that might rank more quickly,right? But you could try to think how do I, how do I generate that that way? **Dmytro Krasun** [1:47:38] Yes. Yes. **Jason Cohen** [1:47:39] Uh, but again, there might be a lot of more manual marketing 'cause again, you're just trying to do it. You could even spend some money. Like the good news is your profit, you, you have good profits. You could decide to spend $10,000 in advertising for this. Again, not that you would do this forever, not that that's a great way to go. Maybe it is, maybe it isn't. I don't know. But that's not the point. That's not what we're doingright now. We're trying to find out is this a good idea and what is the idea exactly,right? And so you could spend money, whether it's AdWords or something else on something like MCP and screenshots to drive people there, find out, then put a web chat thing there, a live web chat so that when you drive 'em there, occasionally you actually talk to them. So,right, these are all, these are all the kinds of things one does, including spending money to explore and discover. And you might say, oh my gosh, like I spent 5K and I had some conversations that was useful, but that is a terrible way to get a customer. And I would say, yeah, I believe you. Right? But, you know, it's a very different question to ask. How do I find out is a very different question than now, how do I have a repeatable sales process? And by that, I don't necessarily mean a human being doing sales. I mean the process of marketing to find someone and then selling is closing them, even if it's all hands-free,right? The question of how do I do a, uh, automated and efficient marketing and sales process is a very, very different question from is this a good idea and what is the idea? And so think about spending money and time answering that question. It's worth 5K to find out, 10K to find out, given that you're so profitable that you have the money, it's worth that to discover. Is this exciting and the next thing for the company? Or you might say, nope, this sucks. And now I'm sure because I, I put in all this time, I even spent thousands of dollars. Now I'm sure it's a bad idea. And that's fantastic to know that. You probably learned something else in the process as well that you can bring to the, the, the final decision. **Dmytro Krasun** [1:49:41] Yes. Yes. Okay. Okay. **Jason Cohen** [1:49:48] And it's gonna be a slog. Like I think, I think some of it's exciting 'cause it's new, but I, well, at least I now I'm projecting myself onto you, but I find that kind of stuff can be a slog. Doing the pa landing pages, testing things, getting ads, they usually don't work, posting on Reddit and people say nasty things. Um, you know, trying to get other people to, to tweet and finally someone does and, you know, three people go to the website, no one talks to you. Um, and just it's such a grind. I know you know this, but it's such a grind. Like you have to ask 10 people to even help before one person helps. You have to run a thousand ads before 10 people click. Um, you know, it's just, uh, you have to beg to get customer interviews, even for people that are already customers. It's just a slog, you know? So, um, I, I know, but I think you owe it to yourself to do the slog for the, and, and it's also a reason why it's useful to set a date 'cause it's useful to know I'm doing this slog, but I'm doing it for 30 days or whatever,right? I'm doing it for four weeks. It gives you like, it's easier to do it when you know like it's gonna end. And at the end of it, I'm gonna really know either this is the future of the company and I even know what that looks like, which is gonna be freaking insanely amazing, or I'm gonna say I absolutely know that this is not one of the options. And now I have a better idea of what options are okay. And now I'm sure. And that was also worth it. So time bound, being time bound helps you get through the slog part. **Dmytro Krasun** [1:51:27] Yes. Yes. My brain is exploding from all these options. Yes. **Jason Cohen** [1:51:35] Well, we only have four options. There's maybe more, but this is good. **Dmytro Krasun** [1:51:40] Oh, these, these kind of like it, it's a combination of some of them. Like it's also it's A in a sense, like I can still go like with A. Like it's also. **Jason Cohen** [1:51:53] Yeah, but not for the next month. You're not. **Dmytro Krasun** [1:51:55] Yes. Yes. Yes. Not, not for the next month. **Jason Cohen** [1:51:57] Right. You need all of your attention on the, it's gonna take all your time to do a good job in vetting option D. **Dmytro Krasun** [1:52:04] Yes. Yes. **Jason Cohen** [1:52:05] You're gonna be tempted to do option A because it's easier. I can hide in Visual Studio and add a feature. **Dmytro Krasun** [1:52:13] Yes. Yes. **Jason Cohen** [1:52:14] Right. Uh, that's easier than trying to get a customer on the phone and ask them questions. **Dmytro Krasun** [1:52:18] Yes. **Jason Cohen** [1:52:19] So you're gonna be tempted to go do option A 'cause option A is smart. They're all smart. All these options are smart. That's why they're on the list at all. 'Cause they're all fine, you know? **Dmytro Krasun** [1:52:29] Yeah. **Jason Cohen** [1:52:30] But, but it, but if you allow yourself to slip into option A, what you're doing is you're ensuring that you will not find out whether option D's a good idea. **Dmytro Krasun** [1:52:43] Yes. **Jason Cohen** [1:52:44] You will do a bad job of it. Probably conclude option D can't work, but that was an incor, but that, that you wrote, you would have arrived at that decision incorrectly 'cause you didn't really try. So that would be like the, the, the founder who makes a product and then like makes four tweets about it to their 300 followers. Sales don't immediately appear. And so they go, oh, that's stupid. I'm moving on to my next idea. And then they, they, they think they're smart 'cause they did 10 of these in a year, but in fact it's dumb because what they did is just ensured they planted a seed and never gave it water, and then it didn't come up. They're like, yeah, but I planted lots of seeds that didn't come up. Okay. **Dmytro Krasun** [1:53:24] It's interesting. In, in, in this case, it's interesting because like one month feels it's, it's good enough period to invest time and to build something. But, but will I get like all this feedback, you know, like from customers in one month? **Jason Cohen** [1:53:40] Maybe not. Maybe not. So, so of course you will decide in fact what to do. I made that, I made that up as a, as a stake in the ground because the answer's not one week and the answer's not six months. So I put that as a stake in the ground. But what's really gonna happen is just what you said, three weeks in, you're gonna see the end of this timeline coming up and you'll just decide. Yeah. I think one more week's enough where you'll say, man, I got three customer calls, uh, on the books. Of course I'm gonna see that through. Of course. Like we made up this deadline. It's, it's, we can do whatever we want. Right? So really it's just, um, a week is not enough. Six months is, is too long. You need to drive to this result because you need to decide whether it is or isn'tright so that you know whether you're picking it and what it is, or you are selecting a different option and getting going on that other option, whatever it is. You don't wanna be in this, in this, in this weird spot for six months. Also, again, I'm insisting you'll do what you want, of course. It doesn't, you don't have to do what I say, but I'm insisting that you not do option A. Okay. But if you don't do option A, you can't just do that. You can't just do nothing for a year and expect this company to be okay. Right? I mean, I know you know this,right? You, you gotta work on the company. So for a month, you can get away with that, not a year. And so this is again why, why I'm trying to time bound it 'cause you wanna, you do wanna set option A aside, but you can't do that for a really extended period of time where you'll, you'll, uh, ruin the company that you have. So. **Dmytro Krasun** [1:55:12] So, uh, what did you say? Do not, uh, go with option A for a, like, why do you insist to not go with option A? Or I just, uh, misunderstood you. **Jason Cohen** [1:55:24] Well, so I'm just explaining why I think a, a year, a, a month or maybe two, you have to figure out option D because I'm telling you to not do option A during that time. **Dmytro Krasun** [1:55:38] Ah, yes. Yes. Yes. **Jason Cohen** [1:55:39] And, and, but if you were to do that for six months and just explore and ignore the company for six months effectively, that's not okay. So you need a time bound. It's okay to ignore it for a little while while you work on option D, but that can't be a long time. So that's why I'm trying to, so if it ends up being five weeks, six weeks, of course that's fine. But I'm just, you know, we're trying to limit it to something like that so it doesn't, it's not accidentally, you know, 10 weeks and, and actually hurts the company in, in fact. **Dmytro Krasun** [1:56:09] Yes. Yes. I, I thought you meant the option A in isolation, like, like, do not go with option A at all. Yes. Yes. I see. I see. I see. **Jason Cohen** [1:56:17] Okay. Well, this is fun. Maybe there's more options, but, um, I always find that when I explore like this, um, some hypotheses I had are found to be true or directionally true, but now I have a better idea of exactly what direction and half of them are false, which is very useful. And then there's all this stuff I never thought of that I then discover. And I don't even know what they're gonna be,right? Like definitionally, I don't know what they're gonna be. So what I think is we've already identified four plausible options. Surely there are more, but hey, that's good. And I think in exploring option D, there will be new things that neither you nor I know at this moment. And whatever those are, those will help you make a better decision later, a more informed, confident decision. 'Cause after all, if you find out this is also a bad idea for product, you might be more certain that option C, selling the business, is correct. **Dmytro Krasun** [1:57:21] Yes. **Jason Cohen** [1:57:22] Because you're like, no, like the single best way to stay with the business was option D and it's wrong. And even that's wrong. Even AI can't save this business. I gotta sell it. Okay. Now you're way more sure. Like that, that's great. 'Causeright now you're, you're like, I'm excited about that, but I'm not so sure. If you could just get more certain, that would be really great, you know? So, um, or vice versa, you discover option D is super exciting, absolutely possible. And oh my gosh, we just, and it's the answer. That's how we're gonna get to a hundred K. We did it. Maybe. So I just think in the, in the exploration, who knows what else you'll discover that will help inform that or, uh, often, often what happens is option E and F appear in the process of discovery and those are better, you know, again, great, you know? Um, so I would, you know, your mindset is explore, discover. So you're not leaping to a judgment. You're not, um, making decisions about the future yet. You're exploring what is the truth? What's going on? How do people actually think? What do people actually do? What do people wanna pay for? How do people pay for it? What the, it's a discovery, not judgment. **Dmytro Krasun** [1:58:35] Yes. Yes. **Jason Cohen** [1:58:36] And so with that mindset, just for this time period, then you can go back to it, you know, then you go back to the other mindset,right? In this period, that's the mindset. You have the company running as, as with as little attention as possible, and you're in a discovery, exploring mindset for the next roughly four weeks, six weeks to be open-minded and discover about this thing. Even if the result is hell no, the only way you're gonna confidently say that is if you have truly explored with complete open mind and excitement andright, and that mindset. Only then if you still decide no, will you be very confident in that? So that's theright attitude, um, for this period. **Dmytro Krasun** [1:59:26] Yes. Yes. I feel I, I feel I need to feel it, you know, like it's an exhausted option. Like I tried everything to make it work and it doesn't. Yes. And I didn't. **Jason Cohen** [1:59:36] Yeah. I mean, it sounds like there's nothing else that even plausibly could get you there. That's incremental, the option A,right? This one at least is possible. So you gotta go find out. Yeah. **Dmytro Krasun** [1:59:46] Yes. Yes. I, I still want to check, uh, customers, like to check every customer, their use case, and maybe I will see some pattern again. Yes. But, but, but I didn't see it like it's, yes. **Jason Cohen** [1:59:57] Yes. You can go check. **Dmytro Krasun** [1:59:59] I didn't miss anything. **Jason Cohen** [2:00:00] Yeah. You know, I, I would never say don't go check,right? But, but I think you said it earlier. If it was a clear one, you would just know that already 'cause you would have 10 examples in your head immediately or 20 and you say, "Oh, yeah, yeah, yeah. There's this." And since, since it's not, like, I, I don't know why poking around would suddenly make that appear. **Dmytro Krasun** [2:00:24] It's theologic. Yes. **Jason Cohen** [2:00:25] Right. Now sometimes that does, it is the case in a bigger organization when any one person cannot see everything that's going on,right? Everyone has some fog of war in a, in a larger organization. In that case, it's quite possible that there's this thing that's true, and yet you can't, you don't know that because you don't, you can't see that. You're not close enough to that thing to know it's true. That happens constantly in a big company. It's one of the, one of the, uh, difficulties in operating in and, and managing that organization, a larger organization is that you can't know everything. Um, so then it's true. You need to dig and find out 'cause maybe there's a pattern that's obvious when you look, but that's not your situation. You do literally everything. You see every support ticket, you know, you talk to every customer who wants to be talked to, you see how they use the product, you get the feature requests and so on. You see it all. Now maybe you could poke in an interview more. Okay. Maybe, but you see it all. There is no fog of war for you in that sense. **Dmytro Krasun** [2:01:27] No, no. **Jason Cohen** [2:01:28] So like if there was an obvious big pattern, how would you not see it? Like why would you quote unquote looking be information you don't already have? Now, could there be a subtle small pattern that you never noticed? Yes. That I believe. That I, that I believe. But a subtle small pattern, that's back to option A. Subtle small patterns not how you get to a hundred K,right? That's, that is how I would think about option A is ask, "Oh, this is only 5% of my customers that do this, but they're great customers." So maybe if I just get, if I get more customers like that, then I could move that from 5% of my base doing that to 10% of my base doing that. It's still a tiny amount,right? But those are, but if those are great, profitable long-term customers, then adding those, that's very good. Incrementally, that's how you get like, "Allright, now I got from 30 to 40, 40 to 50,"right? That's how I do that. It's like I find these little pockets. Then going back to the metrics and really looking that I believe that then I'm, I agree 'cause that kind of subtlety isn't necessarily obvious when you're not looking, even when it's just you. That I agree with. But with option, um, but, but if we're talking about a big new thing that's gonna save the whole company that's lurking, it's unlikely. So you should still go look for your own sanity because if you don't go look, you'll always wonder, "I didn't look though. I didn't look. I needed, I could have been there." So I agree with you. Go look. I'm just saying why, why. Yeah. I'm just saying like, "Look, I don't expect you to find anything." And that's okay. You know, that's okay. **Dmytro Krasun** [2:03:02] Yes. Yes. **Jason Cohen** [2:03:04] Yeah. **Dmytro Krasun** [2:03:06] Yes. It's like, it's like also in all like to, to kind of close this option like A in a sense, like, like to make sure like I tried everything like and now I see for sure and I have hard data to see that. Okay. So there is no pattern I can actually see, but, but it's like, yes, it's, um, it's on the, on the side. **Jason Cohen** [2:03:26] Yeah. **Dmytro Krasun** [2:03:27] Yeah. **Jason Cohen** [2:03:28] Okay. We've been going for a long time. We're full. We have some good options. We know what to do next. So of course I'm excited to hear how it goes. So let me know how it goes. Uh, even along the way, or if you wanna, you know, poke me with questions or things you wanna share separately, that's fine. But I'm really curious. I'm really, really curious to see how this goes. I mean, this is your life. This is your whole life we're talking about. **Dmytro Krasun** [2:03:51] Yes. Yes. I, I am super grateful to you. I appreciate all this conversation and it's my mind is exploding now, you know, it's kinda. **Jason Cohen** [2:04:01] Oh, I'm, I'm. **Dmytro Krasun** [2:04:02] There is, there is a lot of information and thank you. Thank you. Thank you. **Jason Cohen** [2:04:06] Well, you're welcome. And also, uh, I'm grateful to you as well because I mean, almost no one is willing to just go share all their numbers and do all that kind of stuff. And also this is, this is, uh, selfishly great for me because I can maybe, you know, edit this and put it on YouTube and get tens of views, maybe hundreds. And, uh, the way that's gonna be great. Then I can finally retire. **Dmytro Krasun** [2:04:30] Thank you. Thank you. Yes. It was, it was actually a lot of questions you asked and like many perspectives I didn't think about. **Jason Cohen** [2:04:37] Good. **Dmytro Krasun** [2:04:38] Well, thank you. Thank you, Jason. **Jason Cohen** [2:04:40] Well, I appreciate it. And let me know how it goes. I wanna hear. **Dmytro Krasun** [2:04:43] Will do. Will do. **Jason Cohen** [2:04:44] Okay. See you. **Dmytro Krasun** [2:04:46] See you. --- This library is powered by PodHood (https://podhood.com), the podcast website platform.