He Built MY Startup Idea and Sold It in 8 Days ($15K Side Hustle)

0:00Today we've got Josh on the podcast. He's going to tell the story of how he vibecoded an idea he got from ideabser.com and then sold it in 8 days for $15,000. It was a roller coaster journey. And the reason I wanted to have him on the pod is I want him to to teach you how you could do the same. And I just thought it was a really interesting story of how someone could take something from 0 to one and this whole

0:28age of ideas with AI. Enjoy the episode. [Music] Josh, welcome to the pod and uh I can't wait to hear what you have to say. Yeah, thanks for uh having me on, Greg. I'm pumped to talk shop. All right. So, maybe walk us through how this whole thing happened because it's a crazy story. It's absurd is what it is. Craig, like I don't know if you're numb to it, but like as an I was telling this to uh I went for coffee this morning. I was telling this to like a barista and he was like that's insane. Like guy guy guy guy guy guy guy guy guy guy guy guy guy

1:11guy guy guy guy guy guy guy guy guy steals an idea from the internet and then like ends up like flipping it for $15,000. Like what kind of world do you people live in? That's right. Yeah. I mean, yes. What world do we live in? Like there's Yes, we're in some sort of weird unicorn bubble. I feel like this entire story can be told in tweets. Um, and so that's sort of what we're going I'm going to attempt to do here. You posted on um what day was this? July 14th. So, this is 10 days ago. You tweet the about an

1:42AI agent that finds expired.com domains. So, I um I did not see this tweet initially. However, someone who follows me did. They they just tagged me with some, you know, with an emoji and uh and then I replied that I had plans this week cuz I I saw that tweet and I knew that I like it was going to eat at me. And then I tweet, let's see, where is it? Oh, that I'm building this. Yeah. So this is what the next day which by the way you know you know what you did here that I don't think a lot of people realize is you committed you you

2:21made yourself on the hook for doing it right. You publicly said I'm doing this thing. Yeah. I that's um that's my addiction to Twitter. So um I I just have to like it's a that's great content, right? Like if I if I say I'm posting about it. So like um my previous startup it's called bare metrics. It's like a business analytics platform. I sold it back in 2020. Uh but like in 2013 2013 I think I had a very similar tweet where uh Stripe had just started like they were still really early days. Nobody had really built anything on their API other than

3:00just like integrating it for payments. But as far as like apps built on top of Stripe, nobody else had really done anything. So, like I made a tweet that said something like SAS or like Stripe Analytics uh company um I'm building it tonight like coffee brewed. This was like at like 700 p.m. or something. And um and that was the kind of thing that like I you know over the course of about two weeks or something I built this tool that ended up like I sold for like millions of dollars back in 2020. Like but that that resulted in this or it started the impetus for it was this

3:40tweet that I made and so like I find myself doing that a lot habitually where like if I think of something I'll tweet about it and then like that just brings up you know first customers a lot of times come from these kind of tweets like it's self- serving um to some extent and it does it does provide a little bit of accountability um but yeah so like the next morning at 8:50. I tweeted about it. However, I had been up since I typically get up around like 4:30 or 5 in the morning and I had been working on it already at that point, like a few hours. Um, and so that

4:17was but that was 24 hours later I tweeted from your tweet. I I I committed to it basically and then I just kept tweeting about it in a thread and um and then by let's see what time was this 700 p.m. that same day it was live. So um that was that was sort of like the from from your idea to the time that like I actually launched it was like less than 36 hours. I think people are going to want to know how you

4:52actually did that. Yeah. So, um the the nitty-gritty of like pulling that off. There's so many things here, right? Like you see in the in the thread here, I'm talking about buying domains like I first got theai that was available for name snag now. Um actually, let's this will be great. So let let's see if I can. Okay, so in Claude um finding names for stuff is just maddening um specifically when it comes

5:21to finding domains that are available. So what I've been doing is I have um a who is MCP um which we can talk about that if you want to that that's really getting into the weeds that will check the availability of domains right here in cloud chat but like let me let me give you an example. So, um, what I did before I even figured out name snag was like I have to figure out the name of something because if I can't if I can't like name the folder I'm going to put this stuff in or like when I'm generating code, I don't I can't reference the product like if I don't

5:57have a name for it, it will drive me crazy. So, like the very first thing I do is name the thing. Now, started doing this thing a couple weeks ago where I'll be like, um, I need to name an app that helps people find expired domain names. Come up with words related to find or uh snipe was I think one word um or acquire related to domains. So it comes up with 20 ideas here just right out of the gate. Fine. And then it should hit that who is. Yeah. So now it starts going through using that MCP to see what's available. So I just did this

6:51for I don't know 30 minutes or an hour or something where it does all this. It'll go through all 20 of them and then at the end it might be like I couldn't find anything or there's here are the three AIS. There were no dotcoms. And I'll just say 20 more, 50 more, whatever. And I'll do this until I've got 200 random things. And because it's all within the conversation of Claude, it every time it will basically summarize like, okay, after all this, here are all the dotcoms that are available. Here are here's the dotcoms and theis that are available. Like it'll just summarize it. It keeps summarizing

7:30it for me. and and it'll even give its opinion, you know, on like what's uh what's available, what's probably, you know, what's your best bet, like what's the best pick out of all these from a a naming perspective. Um, which pe people often tell me like their big, you know, they're always like the com's not available, the is not available, like all the good domains are thousands of dollars. And it's just not it's it's not true. like there is there are available domains that are uh snappy and catchy that you could get for ten tens of dollars,

8:09right? And my first reaction to when I saw your tweet, your quote tweet around I'm building this, it's called name snag, I was like this is going to crush um just because of the name. like the name is so good that sometimes like when you have a good name to something um it kind of sets a floor for how much it's worth just because of the name and

8:36the concept. Sure. I and and a lot of this stuff I I also have a background. I went to school for design. So like I have this um I look at everything through this lens of like oh will this work from like a branding perspective or some sort of design perspective like oh I hate like if something looks weird from a like the way that the word looks or something like there's this like mental block for me where if like man I just hate how that will probably look if I try to you know do a logo for that or something like um you know if if if one

9:11word ends in the letter P and the next word starts in a letter P. I know that's like that'll be just impossible from a branding perspective or something like that. So like a lot of times I'm I'm rapidly also thinking through branding stuff as well um as these things are coming up you know um but I mean like here yeah we found we found and that are available. So that was what I did with name snag, you know, I mean, and then but that's the thing with the clawed stuff is like because it's got this all this context, I could literally just be like three more and it'll repeat the

9:45whole thing and it's so easy. It's just ridiculous how easy it is to try this stuff out and to brainstorm this stuff um and get ideas. So like this is my favorite thing as of late for ideas uh because like I want the the some tldd to be available whether that's com.ai.app co I don't know whatever it doesn't matter you can try any of that stuff. Um so that was what um was the initial um let's see if any of those were

10:20available. Yeah. So, so I don't think people are using I call it clude. I don't think people are using code like this by the way. Um I haven't seen I never I really never used it until the MCP stuff. Yeah. Like that's that's really kind of what I'm sort of nerd out over chat GPT and specifically like 03 that model and like

10:48the deep research stuff. Yeah. So, but when I could start connecting this to other stuff, it that's sort of like made it click for me. And obviously I use the their their code stuff is what I is my weapon of choice for like actually building stuff. But yeah, the domain name the naming stuff.

11:07Incredible. Kills it for me. Kills it. Okay. So, you named it. So, I named it. Um, so let me switch back to You got a really good name. Step name. So, yeah. So, I got the.com and theai AI was available. I did the.com was like 5,000 I think and I did uh but they had a payment plan. So, I was like, "Okay, whatever. I have nothing to lose here on just paying a few hundred bucks a month on this thing and we'll see." Which, by the way, a lot of people don't realize that if you find a domain that is really incredible and is $5,000, $10,000,

11:4615,000, a lot of people will rent the domain to you. Yeah. Right. And a lot of the there it's in this particular case, it was uh it was listed on um it's adom.com atom.com. [Music] And like they it's already built in like I didn't have to go and like contact the seller or anything like that. out of it.

12:08It was essentially all automated. Um, which was why I was able to buy it that morning and then like buy I could use it that evening, you know, I was not having to negotiate a bunch of stuff or whatever. Um, but if you're vibe coding something and you're trying to do something similar to this, doing something like finding a great domain, maybe leasing it and then maybe flipping it is a really interesting way

12:33to think about it. Yeah. So, it it lowers sort of your your risk, right? Especially if you can lease it and not even make payments towards owning it. Like I mean there there are cases where you can the person who owns it was just like I'll just lease it to you whatever and then whenever you're done with it that's the end. That was still as 9:00 that morning. I got both domains that I wanted. Um so then I I just started building. Um and we can talk about that in a second here, but like you'll see. So what time was this at? That was at that was at

13:0511:30 a.m. I was talking about how I had already built the back end for it and then like a minute later I posted this first pass at a logo. So like what I'm doing is while I'm sort of vibe coding this thing and the the AI is running in the background building this stuff. I'm illustrator which you know you could be in Illustrator, Figma, Canva, whatever designing this thing. So the first thing I did and I don't let's see I had this idea of like the dot what was sticking in my head was the idea of like these are all you care about the the TLD that

13:41you're getting the com the AI whatever so let's focus on like a dot. So that was the the initial impetus for um this sort of like dot name snag thing. Um, I'm a sucker for like slightly rounded type faces. So, that was that initial idea. I was kind of going with this green. Um, then I started thinking, well, how does this look inside an avatar? Like, you know, a little square or a circle. So, we need something smaller. And then that led to like, well, what if we do something like, you know, this where or

14:20just a slightly more interesting dot. Um, then I cleaned that up a little bit and ended up on that. Um, which obviously slaps. Just obviously. Well, it just felt felt good. And I'm, you know, and I could blue is super cliche in the tech world, but like whatever. I I like the blue now that um I posted that on Twitter like here's my first pass and there's a guy named Derek who um has sort of like whenever I post these side projects um he's like hey man I made a logo for you. So and then I'll be like oh man that's legit. Send me a Stripe invoice

15:03link and I'll send you some money. So this has happened maybe like five times now where he does this and then I just give him money uh unprompted. Um so you know you can just do things and so so then he basically sends me this um you know which a little bit cleaner uh gets rid of the rounded stuff which is fine. I think it feels slightly more professional to me. And so uh so I bring the the logo mark into Illustrator this like in thing and then that sort of where we landed was like this this

15:40iteration is what made the site. So you have you have really good branding you you took a first stab at it and then someone you know did the last mile. So, I think the the lesson there is, you know, these people are out there and, you know, even if you're not a trained illustrator, like you can use tools to to to take a first stab and, you know, and then and then get get someone for not an exorbitant amount of money to take to to to to take something to the I I think people tend to overthink this and there were lots of people who were

16:15like, well, what what tool are you using to design the logo? Assuming that it was AI generated. Yeah. Um, which is fair and also completely valid to do. So there's no reason you can't just like ask Chat GPT to make you some logo, especially in the early stages. Yeah. But you need to have an idea. You need to have some

16:36conception, right? Yeah. Yes. You should you if you think put yourself in the shoes of like an art director or something, right? Or like a creative director who's like trying to uh give some directions. It was the same way with like, you know, in the chat there about coming up with domain ideas, like I gave a few words that I wanted to kind of like focus in on, right? So then it can just act like a thesaurus based on the wording or the tone and go from there. So like as with all things AI, the more context that you can provide it, the better it'll Yeah. And you might you might want to

17:11say like, you know, I you know, pretend, you know, this could be the prompt, like pretend you're a creative director. I look up to people like Josh Pigford from name snagging. Maybe I look up to, you know, Greg Eisenberg, late checkout idea.com. You use those, you screenshot those logos, you give it the context, and then see what it pumps out. Then you have to actually I saw this really good tweet today from Jordan Singer. Um, and he said, "Here it is." The new bottleneck is no longer doing the work. It's reviewing

17:48the work. I think that's very true. Yes. Um, so you'll review the work and then probably pass that off to someone who takes it to the to the last mile. Yeah. I there's um so you you briefly mentioned maybe there. So like on May at maybe uh there's four of us working on that product and um one of them is is I I'll say another engineer because I'm doing a ton of coding but he's the actual engineer like he's the the one who's actually really knows what he's doing on the engineering side. And so much of um our process between he and I is like I'll vibe code

18:29something super fast just to test out the idea or the UX behind it or whatever and then I'll code I'll code it then I'll push it to get and then he'll go in and like essentially clean up my my mess. So, like I'm passing off a lot of this to someone who's uh has a better idea of sort of like how this affects all the other parts of the code or whatever, you know. So, there's a lot of this handing off that happens

19:00um throughout the whole process, right? Even the logo here for name snag was handed off to someone who's can can probably do a better job. Yeah, that's right. Yeah, exactly. Okay, cool. So, we got that and then what happened? So, logo um so the and I'll walk I'm not going to just like cover all the tweets here, but I do think this plays a part in sort of the story here. So, um what you see me doing here is like I I post a poll about like how do we style the domain the domain? Do we do this, you know, capitalize the s lowercase whatever? uh

19:38post a screenshot of like proof that I've actually got something working. And this entire time is me making this almost this community effort on some level. Um to me what's super interesting about social is not just constantly broadcasting, right? Like it's a it's a the interactivity part. It's me posting something and someone has an idea. me posting something a logo and someone responds with a different logo. Uh, you know, it's uh someone saying like, "I like that feature. What if you did this?" And that's incredibly beneficial um both to me as I'm sort of doing this sitting here in a box by myself. Uh but it's also

20:24building up a bit of hype on the product itself. Um so I keep yapping about it. um and then get to the point where it's like, "Hey, we've launched." Yeah. Again, that was like 7:00 that night. Um basically, I worked on it for the better part of 12 hours or so that day. And um but I think later that evening, like I stayed up another couple hours and was at I mean, we did $1,000 or something in revenue within a couple hours of of

20:58launching this thing. um and subscription stuff because I had just been talking about it all day, right? Like because I was posting polls and screenshots and uh asking for feedback on things. Um I think it's more than that in my opinion. I think it's when I zoom out you cuz I know a lot of people are going to be like, "Well, Josh has 50,000 followers. Of course, he got $1,000 MR in like a in a few hours." But what if you look at I'm building Yeah, this I'm building this. Uh, click on that. Click on that. I just want to see how many

21:34impressions you got from that. 145,000. Like if you go to, you know, go to your Twitter your Twitter account. Like you don't normally get 145,000. Oh, right. You know what I mean? Yeah. Like like this is a this is like a big one and you're getting 11,000. go like to let's see buy cookies for breakfast. You know, you're getting you're getting in the single digit thousands of views

22:02and you got 145 times more impressions. Now, let's break that down. Why did you do that? What what happened? You jumped off a trend. Basically, you found a tweet that was going viral, right? In this case, it was about a startup idea. Yeah. and you quote retweeted it and which I did not realize this like mine my quote retweet uh got you know not as much as your as your original tweet but almost right um almost so it's like yes it very much though was like jumping on to uh that your your your tweet that started going quasi viral yeah So that's what people don't understand

22:50about these social platforms, specifically X too, is like find something you can there's tweets going viral every day and there's a startup idea related to um to that like I saw go to Brett from DesignJoys Twitter. So I I've had this guy on the pod once and here he said uh click the latest his latest tweet. So, there's this app that's going viral right now. I think it's making like two $300,000 a day of revenue, which is

23:22crazy, you know. Um, and he designed it. It's called Tea Dating Advice, and it's like an anonymous app to like basically score men that they go on dates with. And so, it's an obviously viral idea. And he says, "I'm the designer behind the mega viral tea dating app. If anyone wants to build the equivalent for men, my DMs are open. I have plenty of ideas." So there's yes, this guy Brett from DJ can go build this, but you can also quote retweet this and be like, I'm building this and then all of a sudden, how many views this guy, he just tweet, you know, it's gotten

23:5755,000. Like all of a sudden, you're like siphoning views and then people are following along your story and then they become your first customers. As I was posting this stuff, there were, you know, I think there were a number of comments that talked about how like, you know, if anybody else is building this, like it's game over. Josh will like beat you to it. Um, knowing that I how fast I'll build stuff. And I think that the problem is most people overthink this stuff, right? Like I don't have some I have a skill set to building fast only if only because I've put in the reps to know how to build

24:32stuff really fast. So, here's a list of what is it? 78 projects over the past like 20 years of things that I've built. And and this doesn't even this doesn't include everything that like I built and then just shut down because I like or I didn't completely finish it because I realized, oh, this isn't going to work. Um like there was a Twitter bookmark AI processing thing that like I was building. got real close to finishing it, but then I realized, oh, I'm gonna have to spend $50,000 a month for Twitter's API to like actually make this work. So, that those things don't show up on this list here. So, like there's

25:09there's 78 things here on the list, but it's pro it's it's hundreds uh of of things where I've started something, right? So, I'm very good at like the zero to one for most things, like just starting something and seeing what happens. And so I've put in the reps and there's no reason someone can't take like go to like idea browser every day and just freaking build the thing. Even if you uh I told this to people who like new new um say freelance designers who are like well how do I how do I get clients? I don't have a portfolio. Well, just like make up stuff, man. See an idea for a company

25:47and just make a logo for it or design a website or design the app or, you know, do like Brett's doing where you have some idea and he so he's designed it. He didn't like build the app, but he designed it. Like you can do that. You can just do that, right? And there's no reason you can't take something the an idea of every day on idea browser and say okay here today this week I'm going to time box myself for the week to build this thing out and I'm going to tweet about it constantly annoyingly even and eventually you'll build up a following just around that even

26:22I was talking to someone who he's you know 16 17 years old and he wants to go to business school and I was like, "What does that even mean in 2025 to go to business school?" Like, what does that even mean? Like, you should go to shipping school, you know what I mean? And you should, you know, use idea browser, use or or just come up with

26:45ideas on your own. Like, use chat GBT. Like, whatever it is, just get the reps in to know how to, you know, take an idea from 0 to one. all the stuff that we're talking about, you know, the the like to us and to you Josh, like it's it's it's intuition, but you've built it over time, right? So,

27:05people need to build that intuition. Yeah. I think that there's everybody wants and it feels AI has made it feel like there's a shortcut and to some extent on some parts of it there are but the reason you know some of the questions people were asking or wanted to know from this podcast were like what you know do you are you using some sort of like boilerplate template for code or whatever which I'm not and like a lot of that stuff is just I just know how these things are going to work because I've done it so many times. And I haven't done it because I got paid to

27:42do it, right? Like, I'm not doing this working at some company or whatever. I'm doing it because I enjoy building stuff. And so I end up getting the reps in because I I'm obsessive about it. Um, and you you just do that and you talk about it and then, you know, yes, uh, my

28:0150,000 followers or whatever is helpful. That's that's a level of distribution. But like that came from me being on Twitter for since 2007, right? Like almost 20 years and just tweeting constantly uh about stuff. Um not saying that you need to like 20 years to build up a following, but like you build up a community around the stuff that you're doing by just putting in the work. And yes, you can't really but I will say it was harder for us back in the day because we had to build followers and then then we had followers to distribute and that took a lot of hard work. Today

28:40with the for you sure for you didn't didn't exist. Didn't exist. Right. So there's people who have like the uh the tweet that went you know I think it was like one of the most viral tweets of all time. It's like share share a piece of lore. Did you see that tweet got Oh yeah. Yeah. Yes.

28:58Like go to that tweet. Absurd. Absurd. I think it got like 1.3 billion views. Um there it is. So 1.2 billion. Yeah. Give me a break. Like first of all, that person is getting what? Like like a check for half a million dollars. When I saw this account, I think it had 2,000 followers.

29:18Followers. Is it almost 30,000 now? Yeah. That was that was not possible, you know, even 10 years ago. Yeah. So, if I could summarize like my takeaway from your story is find good ideas or come up with good ideas. Um, figure, you know, pos have a positioning and brand in the market that's that sticks out. Bring people along for the journey. Um, and

29:52just keep trying to make it better. I I think there's people are so afraid of of putting out stuff that um you know there's that like that Reed Hoffman quote about like if you've launch something that's like perfect or whatever you've waited too late or something like that like like waiting to ship perfect software like you've already waited too long. You like you just have to ship stuff and uh a lot of it's going to look bad. It should it should almost by by definition should like be kind of gnarly. Um because otherwise you're starting to make assumptions about what people will use um what features they want when it's

30:34when we're talking about software. Like you just have to kind of do things and then see get feedback on it. Um and and like with namestack there was all sorts of feedback around people saying like well this doesn't work or like this is a bad idea which is fine like people are allowed to critique stuff. Um, but if anything that means like that's a good thing that you're getting that kind of feedback because it means people are paying attention. Um, before we head out, what what would you have done

31:03differently in this project? Um, I I don't know that I would have done anything different because I I really there's another tweet in here which it would take me 10 minutes to find, but like um so I posted this what what day of the week was this on? Um Whoops. July 15th was what was that like a Wednesday or something? Um that was a Tuesday. So that was on a Tuesday. That Thursday I uh I occasionally will like go up take the dog and go up to the mountain cabin in the mountains and just spend a few days kind of doing whatever. Um, so like that, uh, Thursday,

31:53uh, I went and I like spent the first couple of days just building out I I scrapped the entire first thing that I built, all of the code, and literally built it from scratch again on like that Thursday. So like two days later, I built version two of this thing. Um, and I think the keeping up like if I had not sold it, I probably would have kept that same mentality of like let's scrap it and like let's build the next iteration of this. And that to me that's what AI enables is this this sort of rapid iteration on things where you have an idea and then you realize,

32:33oh, here's some things that didn't work. Okay, well, you can just build it again. like it's not we didn't spend six months building some complex piece of software. Spent a few hours and not and I'm not tied to it. I now have learned I've got new information so let me just build it again. Um and to me like that's one of the big takeaways is is just you know I I I shipped two versions of this thing in like a 48 hour period. And um you can apply that to almost any side of business including marketing, you know, growth stuff, code, whatever. You just you just keep building stuff really

33:12fast and and don't get like tied to it, right? It's everything's not your baby. Amen, brother. So, people have asked you questions. We're going to go through three or four of those questions and and answer them. Cool. Let's see. So, somebody asked here about selling. So, this is actually a decent topic for us to talk about here. So, um they asked when I knew I should sell it or do I build to sell things. So, um first off, I don't I don't typically build anything to sell it. I build it

33:42because I'm trying to learn something. Like that's my motivation is I want I'm I'm teaching myself some piece of code um or some some concept, I guess. So, I don't build stuff to sell, but I do tend to assume I will not hold on to things for a long time. So, this is the same reason that I can build V2 48 hours later because I I'm not uh I'm not so married to V1 that I built and think that I've done something amazing. Like, if I've learned what I needed to learn from it, then like let's move on. Um, and that was the case for this. You

34:18know, your idea was like you could build a 1.2 2 million ARR business, which is entirely possible. I wasn't necessarily I didn't build it because I thought that I could make $1.2 million, right? So, I sold it for like $15,000 uh eight days later. And it's 100% could have sold it for more, but like I'm not I wasn't building it to sell it. I

34:41wasn't building it to make $1.2 million. like I was building it because I wanted to learn um stuff about like who is records and then the ability for AI to like process um all this data and um you know semantic search stuff like super nerdy things that just happened to fit this project and that struck a chord for me so I decided to build it. Um, but that's sort of my motivation is almost exclusively for learning things. Um, more than anything. There's a couple questions about like selling it or like this person here's like leave money on the table and not list it on like acquired.com. Um

35:23I think there there one idea that I have had here is like it would be kind of interesting if I wasn't actually building a full scale sort of VC back startup also um along with all these side projects is uh it would be kind of fun to every week pick some idea build it out and then you just sort of sell a startup in a box. Um, I think there's this thing. So, in the past since 2023, in the past like 2 years, um, I've built all these different apps like sold for 8,000, sold for 25,000, sold

36:01for 5,000, sold for 12, sold for 24,000. Like, I think I've probably done I don't know close to $100,000 over the past couple of years just selling off random side projects that I built over the course of like a week. So, I think that there's if if your obsession is just building stuff, um you could almost make that your job, you know, like there's income to be had of just like taking an idea, building it, and then selling it. Um I

36:32think AI allows for a lot of that stuff. And there's people who don't want to get in the weeds on the AI, but do want a functioning tool, and you could sell it. Um to me, that would actually be kind of fun. Cool. That's the podcast, bro. We did it. I appreciate it. Josh, um, how many ideas did you say you've done? 70 There's 78 on this list here on on my

36:56website. So, this is the Startup Ideas podcast. You're obviously the you're an idea person. You're one of us. I appreciate you coming on. If people want to follow Josh's journey, I will include links to all the places you can get to know him better in the show notes, uh, check him out. I told him before we hit record, I told Josh that he's my my ex crush of the week. Um, because I I recently found him and I've been I've been I've been consuming his tweets, loving it. So, um,

37:27you know, keep it up. Well, thanks for having me, Greg. This was a ton of fun. Later.