How to Scale a Subscription App to $25K/Day on Meta Ads
How do you scale a subscription app with Meta ads?
Start with an offer whose value is obviously outsized for the price, then test the packaging rather than the product. Format Finder moved from a single $97 plan to $29, $59, and $97 tiers, and the single change that moved AOV the most was toggling annual billing on by default. Free trials failed outright because the cost to acquire a trial on Meta never backed out. Target customers with a need rather than a want, since need-driven buyers churn less and complain less, and turn your best users into your ad creators by paying them a percentage of the spend their ads carry.
Scaling a subscription app with Meta ads comes down to the offer, the pricing packaging, and where the creative comes from. Brad Ploch and Zach Stuck sit down with Naveed, who went from affiliate marketing to a 45-person performance agency to an ecommerce brand to Format Finder, the app he built in three weeks and now runs at $20,000 to $25,000 a day in Meta ad spend. You'll learn how he builds an offer, the pricing tests that worked and the one that failed, the campaign structure behind the spend, and the creator flywheel that turns his own users into his best-performing ads.
We discuss:
- How to tell whether an offer is worth running before you spend a dollar
- Why problem-solution fit matters more than unit economics at the start
- The move from a single $97 plan to $29, $59, and $97 tiers, and why the free
trial never backed out on Meta
- The pricing change that moved AOV the most, and it was not the price
- Why customers with a need churn less than customers with a want, and how a
realtor funnel tested that theory
- Why he runs one CBO split by geo, no ABO testing, and no interest targeting
- The creator flywheel: users who hit a million views make his ads and take 5%
of the spend those ads carry
- What a one-click AI video editor actually costs to run, and how SaaS margins
survive per-user API fees
- How he runs the ad account, customer service, and product development through
AI agents in Telegram and Slack
- Why the harness around the model matters more than the model
Get the newsletter: https://www.scalabilityschool.com/scalability-newsletter
Learn more: https://www.scalabilityschool.com
Connect with Naveed https://x.com/navuud
To connect with Andrew Foxwell send an email Andrew@foxwelldigital.com
To connect with Brad Ploch send him a DM at https://x.com/brad_ploch
To connect with Zach Stuck send him a DM at https://x.com/zachmstuck
Learn more about the Foxwell Founders Community at https://foxwellfounders.com
Learn more about the The Hive Haus Creators Community at http://HiveHausUGC.com
Scalability School is a weekly show from Andrew Foxwell, Brad Ploch, and Zach
Stuck about what is actually working right now in DTC growth. No theory, no
recycled LinkedIn takes, just the numbers and decisions behind real brands and
real ad accounts.
Chapters
00:00 How to scale a subscription app with Meta ads
00:29 Meet Naveed: Switch Research to Format Finder
02:45 From affiliate marketing to a 45-person performance agency
04:06 Getting banned by Meta and what it exposed
05:45 Why the ecommerce inventory cycle burned him out
08:10 How One Peak turned viral videos into a course business
10:10 Building Format Finder in three weeks with AI
11:50 What the app does, from format quiz to finished cut
13:15 Why the offer is the biggest lever nobody talks about
16:18 How to price a subscription without killing conversion
19:15 The pricing change that moved AOV the most
19:55 Why customers with a need beat customers with a want
22:05 A word from Northbeam
23:20 Why your ad does the targeting, not your audience settings
24:45 Ad account structure: one CBO, no ABO testing
25:31 The creator flywheel: paying users 5% of ad spend
28:35 How the video editor cut churn and unlocked scale
29:35 SaaS margins when every user burns API tokens
32:50 Why the TAM for creator tools is enormous
37:50 Running ads, support, and dev through AI agents
Full Transcript
(00:00) thing. I feel my whole history has been around identifying a good offer, that's from the affiliate days where if I can't visualize or understand the value this provides to your who your customer then you know, spending my own money as an affiliate isn't going to make So Everything. Everything comes down to problem solution And now let's take a listen to the Scalability School podcast.
(00:30) Welcome back to another episode of the Scalability School podcast. Zach's here. Andrew's not. And we have a very fun guest today who I'll introduce in a second. I feel like last time Zach and I were left up to our own devices, I said something questionable about not being here. He's alive and well. I think he's just cranking a webinar for for Foxwell founders right now.
(00:47) So he's he's doing great. He's doing great. It's just not he's just not here today. But we are pumped to have Naveed here with us to talk through his journey in from e-commerce into what he's doing right now. So I'll give you give a little bit of background. My experience with coming across like your work was with Switch Research years ago, right? So I followed you on Twitter, saw the work that you're doing, and then recently, it's like every couple of months I see we're at 100 and I see we're at 4 million AR and I see we're at 10 million RR.
(01:17) I'm like, holy on. Like we. Yes, for sure, for sure, for sure. But it's like the growth rate looks looks incredible. And I'm sure there's there's a ton that we'll get into it that. But I'll be honest, like the thing that most impressed me is the $2,000 a month you're spending on DoorDash. So I want to start there actually, like let's, let's, let's get the go to DoorDash order.
(01:36) Like what? What are we buying? What are we buying? How many meals a week is that? Yeah, it might even be more than that. We just don't cook. Is is the key component there. So my wife and I, we've got kids and almost every night it's something whether it's like sushi for dinner or, you know, snacks like 9:10 p.m.
(02:00) because I'm usually up to like two, 3 a.m. grinding away. And so I got to get that, like, last minute 7-Eleven ordering. That's that's respectable. I, I had a pretty, pretty gnarly stretch of thought myself. I've toned it down a little bit, but I'm the same way. Like, I just I don't if I'm left to my own devices, I'm probably not going to cook.
(02:20) And I'd rather just, like, sit here and then have food handed to me. Just do that. So, so so I get Cool. Sweet. Well, let's we can we can jump into some of the tactical stuff. So I think like it would be great to just get a little bit of background and just kind of share the, the 2 to 5 minute overview, like background like okay.
(02:38) How did you how did you get into E-Comm. Maybe there's maybe there's a story before you come and we can get into. But switch is my first acknowledgment into the work that you've been doing. And we'll we'll work all the way through Format Finder, and then we'll dig into some of the tactics from there. Yeah. So my background originally is affiliate marketing started that when I was like 20 years old.
(03:00) First affiliate network I ever worked with was ads for Doe run by Jason Active So I started off doing affiliate marketing, got pretty good at that, partnered with my ex partner at a performance marketing agency that we had called straw House. That's Jason Christy, and we were doing the performance marketing affiliate model for a bunch of different clients in the US.
(03:22) One of the big ones was Golden Hippo. Sure, you're familiar with them? So we we were on the scene when they first started, and they had Doctor Gundry and Probiotic America and all of those brands. And so we drove a couple hundred million dollars in revenue for for them on that affiliate commission type deal. And then we brought on other clients like Butcher Box.
(03:44) We did a ton of volume for them and very good butchers and a bunch of other ones. Giddy up. Worked with them a bunch. And then after the well, we ended up growing to about 40, 45 people. And at that point, I mean, I was like 26 at the time. And I was like, this is this is like very dependent on a few good clients that we have.
(04:08) It's the affiliate model. Right. And we were Canada's biggest video advertiser at the time for meta. And they were they were loving us. They were flying us to Toronto. And then like two weeks later, they just like randomly banned our ad accounts because they were like, oh, no more, no more golden hippo type stuff.
(04:26) We loved you last week. This week you're off. You're off the platform. We ended up getting our accounts back, but it was kind of like a it kind of showed how much of a straw house the company actually was built on, the type of performance marketing that we were doing. And and just like the model itself is being like, you need to have those, those that arbitrage working between what you're getting paid out and what you're spending on that, that CPA on meta.
(04:53) What year was this one was this was like 2013 to 2016, I want to say okay. So early. Got it. Yeah. Yeah. You guys were early. All right. Cool. Keep rolling on that. And then, so we got kicked off the meta for a little bit and then got back on. I was kind of feeling the stress of like, having 40 employees and only having a few campaigns working and it being like, oh, you know, our monthly net is like two, 300 grand and we can do this with a team of like five people.
(05:25) So I ended up selling my shares to my ex business partner there. I was like, I'm kind of I'm kind of done with this. We did pretty well with it. I want to take a break, got married, decided to have a few kids, and then, well, we had our first kid. And then for my second kid, we had a bunch of trouble with that, unfortunately.
(05:46) And it kind of that's the word like propelled the idea of switch research or started the journey of getting into building this mental health journaling brand because my wife ended up, you know, we're lucky we had access to a therapist for her for for what she went through. And, and journaling was a big part of her journey of, you know, dealing with it was it was miscarriages at the time.
(06:15) And so we decided to to build out this, this brand where we work with therapists, develop these mental health journals and then sell them through ecom. And that did really well. But there were a lot of ups and downs where we would get a campaign working. My thing was meta, like, I know how to advertise on meta, I know how to build those performance marketing style direct response campaigns, and we would get you it working really well.
(06:39) And then I would order 30, 40,000, 40,000 units of inventory, and then it would fall off a cliff. And this was like a cycle that happened for three years where making a profit run out of inventory, have to order a shit ton of inventory. I'm sure you guys know. Yeah, yeah. And so that was the kind of like I just got bored of it to some degree where I'm like, we keep going through this cycle and unless I make a change and snowshoe some of that risk across different channels or some organic, it's just going to be a cycle
(07:16) that keeps repeating. Or I would need to be better at outsourcing the generation of funnels and, and, and the actual ad campaigns that end up performing. What did the team look like? Like up until. I'd be curious if you like. Don't mind. Feel free to keep going, but I'd love to hear, like, what the team looked like at switch, because I know you're kind of like a a solo operator to some degree.
(07:40) And I want to know, like how Big switch actually turned into and tell us about that. It did like mid like 5 to 9 million in revenue per year for about four years. And the team was like me on all the marketing, one person doing some organic and social, a customer service agent and then a business partner on that who is handling all of the inventory, the product development side.
(08:11) So, so very similar to like with straw House where I was the marketer and my business partner is the one who's either bringing in the product, developing the product, or bringing in the business doing the business. So pretty small. Yeah. Pretty small team. Yeah. Cool. Keep the inventory. Yeah. Totally. Keep keep rolling up okay.
(08:30) So we're they're getting bored. That's rich. And then I start working with some friends in town. One peak creative. They did like they blew up their just a video agency at the time. And they blew up on Instagram during Covid. They did some like ad solution is what they called it, where every week they would create a video that goes or create an ad for a random company and then ended up going viral.
(09:00) So like every week it would go viral too. So they did really well, but they got like what they did one for Ryan Reynolds Gin Brandish or Whiskey or whatever. It was. Vodka, I can't remember. And then he ended up. Yeah, yeah, that's right. He ended up commenting and reposting it and shit, and that kind of made them blow up a bit more.
(09:19) And then they decided to launch a course, and I, I was always talking with them, providing some guidance around like how to create an offer that you can sell through meta ads. So they created this course on how to go viral on TikTok and Instagram. And, and I just ran the ads for them. Kind of like as a percentage of spend type deal doing it for my friends.
(09:43) And we, we we ended up scaling that to 20 to 30 K and add spend. And it's a course which I love the digital products. You don't have inventory. So I'm like, yes, this is amazing. Is that like that or a day like what are you talking 2030? K sorry, sorry a day. That's what we scaled that course up to 2030, 20 to 30 day and spend very profitable too which is amazing.
(10:06) And then they introduced a high ticket item and then so that we've been doing that and I was doing the media buying for that up until January of this year where I started. Well, December, I started playing around with Claude and obviously all of the vibe coding tools that came out around that time, as everyone did.
(10:24) And then that's when like grind season kind of hit. I think for a lot of people where you could see the capability of these models being able to actually create, bring, bring your vision of whatever app or idea you had to life. They're very capable. So anyway, I approached one peak and I was like, hey, let me make an app for this.
(10:45) We have we have all the methodology around helping somebody go viral on TikTok and Instagram. We have these 62 or 64 different formats that are part of the kind of like the IP that we've generated over time that are proven to go viral for different niches. So let me take a crack at turning this into an app, and then took me about three weeks to develop, like the bare bones for what format finder is right now.
(11:14) Through just working with Claude and Codex and Terminal, that kind of thing. And then and then we just decided that we, we, we would run it as an upsell to the course business. That was like the first test to see how well it would do and if there was any reception for it from the audience. And it did really well as a as an upsell.
(11:37) And that was like introduce this recurring revenue because it was a monthly subscription. And then and then we just started transferring everything to that. So now I'm an equal partner in we just turned it into one peak software, and it holds all of the IP, all of the the entire everything is the app now. So all of the courses exist in the app.
(11:59) Everything is subscription based. And then the app itself. So it aims to provide creators with everything they need to go viral. So you would put it in your niche. It identifies what format of the 62 you should probably try. And that could be like a challenge, a skit, you know, a knowledge gap type format, and you play this little taste game.
(12:23) It identifies the format for you. And then from there with that format, you can generate ideas, hooks, save your ideas and hooks. It'll generate the script for you, tell you how, the shop plan, how to film it. Once you're done filming it. We have a one click video editor in the platform where you drop in all the footage and it'll use the transcript from all of the footage you've you've put in there.
(12:49) Identify kind of like the best hook, the best body payoff, all that, and it'll cut together your video for you. And if you have been added in as well, it'll add that in. And then you can export that video and you can put it through our analyze feature, which does something similar. But it it does take frames from the clip to identify the visuals, the lighting, all that.
(13:09) And it'll give you kind of a bit of a score around what you can do better or if it's ready to post. And at this point, it's really good at identifying because we've just been able to train it on, like what videos do we now have actually from our from our subscriber base have actually gone viral. So it'll be able to identify pretty well whether something is going to flop or do really well.
(13:30) But in the in-between of like whether you're going to get 100 K views or 50 or 400 like that is a bit more nuanced and it's hard to predict that, but it's something we're working on. So anyway, there's this this whole flywheel in there, plus all the courses. Yeah. That's crazy. I have a I wrote down a note I want to go back to.
(13:47) It's something you said early on in that story, which was kind of helping the one team develop an offer around that. And I feel like that's a it's kind of like a slept on topic in general. Like every talking about last year or two years ago, I was like creative and creative volume. I feel like the wave this year in like the last seven podcast episodes, we don't have been about landing pages, but like an offer like is probably the biggest like could be one of the biggest levers out of those three.
(14:10) And it probably has the least discussion because it's probably the hardest to crack. Is there like a process that you went through in order to develop that? I mean, I'm sure you had something similar in the affiliate days, but like, how did you even get to that? That's the thing. I feel like my whole history has been around identifying a good offer, and that's from the affiliate days where if I can't visualize or understand the value this provides to your who your customer is, then you know, spending my own money as an affiliate
(14:41) isn't going to make sense. So what problem? Everything. Everything comes down to problem solution for me and my history. So like what problem I started solving is a very clear. And then is the offer that you're, you're providing outsized in the value for the cost. And then it just comes down to the marketing. How well can you how well can you.
(15:04) What's the word like a display the problem? Or, you know, create the emotional need in the user for or meet them where they're at to, you know, are they aware of the problem or are you trying to create that awareness in them and then provide your product as the solution or for does it kind of sell itself, which some products do where, like how to go viral on TikTok and Instagram? It's pretty obvious to the user why they want to do that.
(15:35) They can like add in their own personal kind of like journey in there on their own. Whereas if you're selling something that where you have to actually introduce the problem, like getting up in the middle of the night to P9 times and then introduce your your offer as the best solution for that. Yeah. Is there a is there a like so when you when you're describing the offer to me it sounds like it's not actually maybe not primarily about the actual like unit economics of it, which I think that's a that's a part of it for sure.
(16:06) But you're actually kind of talking about the entire package, which is like the offer itself, which is you kind of led with was going viral on TikTok like that. Is that the is that the offer or like how how do you break those down. Yeah. So for me I would say like the offer. It's like the keystone of it is the, the need, the desire, the the value that you're providing.
(16:30) And then and then the unit economics because you can you can change that over what? You change all of it over time. But that's something that typically we need to split test, identify like adjust. It's like once we so for for one peak you know we developed the the offer go go viral the landing page. And then we did a bunch of split testing around price point.
(16:54) You know is this like a 2.99. Because it also depends on the platform you're running on. How much is it going to cost to acquire a customer and how good are your ads. And so do we want to have this like a $299 one time fee for this course, or is it going to be a monthly subscription? Do we? And we started off with like monthly subscription.
(17:12) I think it was like 97. And then we realized, oh, if we give them 40% off the unit economics based on the CPA, we end up making way more money. You know, our conversion rate will be much higher. AOV, it ends up being lower. But net net we make more profit and then we can scale more. So a lot of that comes down to testing.
(17:35) Yeah. And the creativity I guess in in coming up with those ideas when you guys. So when you guys were setting prices for this, I'm curious like did you just play with that a bit or like how how much of the actual price of format finder you guys mess with? So Format Finder itself, definitely. We started off with that like 97 price point, and we even started off with like a first month 58 and then 97 every month after that.
(18:04) And I think I started spending a couple thousand dollars a day to that. It was payback period was like month two. But then we we decided to test no, no, no. After we did that, we got a bunch of feedback from users and we waited a bit, so I didn't scale it. I kept it at around like that. I think that like two to 5KA day because we wanted to see, you know, where's turn going to come at.
(18:29) And then there was a huge cliff for churn when they hit that 97 renewal point. And then a lot of people were like, oh, I didn't realize it was actually 97. So and we bolted on a bunch of new features and decided to try like the three tiers, which is more typical to a SaaS model. So it's like we came from the course where it was 58 and we're like, okay, let's let's go with the tears.
(18:50) So 29 is starter package. And then I even tried Free Trial and that sucked for us when it came to like the cost to acquire a customer on Facebook was still really high for the free trial. And then it just didn't back out. Sure. But having that that 29.99 starter price, 59 and then 97 ended up being in our split test, kind of what worked best.
(19:16) And the the actual thing that move the needle the most surprisingly, was toggling annual by default. Yeah, I was just looking at that. I would literally didn't even realize that this was annual pricing at first. So yeah, yeah, yeah, that's super interesting. Yeah. Super interesting. So that that, that had a huge impact on our AOV.
(19:36) And now like we don't we don't get many complaints I guess the customer service. So people doing by accident. But yeah I think like again these individuals that are trying to go viral on TikTok or Instagram or whatever, right, like to spend $300 with, with their hope to make, you know, hopefully, you know, six figures doing this as a career.
(19:58) It's not that crazy in the scheme of things. If it's the if it's the right tool. So that is really smart. I wouldn't have thought that that like I went to the said like, oh it's immediately toggle annual. But yeah super cool. Yeah. And that's where we have found that our best customers are the ones who have a need for the product as opposed to like just a want, if that makes sense.
(20:21) So like an example of this would be Realtors. People don't own a business. People who want to be a business. Does that make sense? Like you want to be like a health influencer and that'll be like your business where you have your personal training clients come in or whatever, versus like the, you know, the woo woo spiritual.
(20:38) I want to talk about meditation. Worst customers who like most customer service requests from them, most complaints. And I'm sure this applies everywhere, but it's just like it's something that we realized about a month ago that, like, we need to really go after the ones who have a need for this. Because for them, spending this much per month is nothing like it for for the value that they're getting.
(21:03) And then also they will move into the high ticket item, which is like a $4,000 community with coaches. It's more hands on. So we have that in the back end. And then that's like the only upsell that we have right now, which is sign up for that masterclass. Nice. Okay. So I was going to say you stopped running ads in California is what you're trying to say.
(21:22) Yeah. No, no I'm just I'm just go it. Go for it. I was I was going to get to the realtor thing because like I was looking at the attic count and it looks like that realtor funnel is relatively new, but it makes sense, right? Because like, they sell one more house and they paid for the software for six, seven years, maybe, maybe more than that.
(21:40) It's like it's like a no brainer. Okay, so maybe a couple questions around that. It sounds like you. It was pretty obvious to identify because like, they just didn't complain as much and their turn wasn't as high. Are you doing anything? I want to come back to that upsell thing too. But are you doing anything inside of meta that's like unique structure wise for the realtors, like creative for them.
(22:00) Landers for them is. Yes. Are you like anything crazy with the pixel like optimizing for those people qualified leads not qualified All right, friends, quick break. This episode is brought to you by North Beam, the marketing attribution platform that we love over here at Scalability School. If you're a marketer, I mean, you probably agree, right? Incremental testing is broken.
(22:18) They were. These tests require tons of labor spreadsheets to reference, and the designing of the test is slow and error prone. How are you supposed to know what to test in the first place? Right. There's a million factors that can make incremental testing difficult. But the good news is North Beam has completely changed the game, and they fix this.
(22:37) Only North Beam incrementally automates your incremental test design and monitoring, while protecting you from the major mistakes that plague incremental testing. North beam builds tests using your own MTA data, launches them directly in ad platforms, and automatically monitors them throughout. If something threatens validity, you're alerted before it ruins the result.
(22:58) With North Beam, you can launch tests in platform with a single click, saving you tons of time and reducing any chance for errors. And when the test concludes, your results flow straight into your MTA, meaning you can easily see results alongside the rest of your data. Incrementally. Was broken. North beam fixed Yeah.
(23:18) So that's like a that's a small split test. So like right now we are spending like 2020 5KA day us on on ads in that in that ad account and the that realtor funnel. It was at like 500 a day. It was just a small test. And it's about breakeven. It's not doing better than the rest of the funnel or like the general population stuff.
(23:43) And it's like it seems like it's because when obviously we target realtors in the copy, look, I don't do any interest targeting or anything like that. I'm like, I'm a believer that your ad does the targeting. So talking about realtors in the ad media puts it in front of realtors. But what we've noticed is like obviously the cost per click for those users is a lot higher because of the type of person they are.
(24:14) And the conversion rate with that quiz funnel isn't high enough to support the increased cost to get them there. So it ends up being at this point where that the general funnel that we're leading people to is doing better. I do still want to explore the quiz funnel as like a method. I think I can get that working, but it's just it's just a small test right now, like off the side and then in the account.
(24:45) Sorry. Go ahead. No. Go ahead, go ahead. No, I was just gonna say in the ad account itself, I just, I just run CBOs that are split up by by the location. So like, a US only CBO and then the other two are just, like, tier one, English speaking, and then a tier two, whatever, but everything gets it's I don't like I used to do the, the abo where you would test in your Abo first and then throw it into the CBO and performance that are winning.
(25:17) But at this point, I just throw everything into the one CBO campaign and just let it compete in there against everything else. We do have a pretty cool flywheel right now for generating ad content, and so we are getting a ton more, more ad content from our user base and then using it in there. Yeah, that's nice because it's like it's baked into its baked into the product.
(25:39) It's like they're making I'm making they're making content as a result of that, of using the product. Or you're like actually just reaching out to them on top of that and saying like, hey, what were your results? What does the flywheel look like? Yeah, I'm so stoked on this because it's working so well. So when somebody in the platform gets like a million views or more, we'll reach out to them and we'll be like, hey, do you want to be part of like, the UGC creator? Product, like Pro or whatever? Yeah, program we have and
(26:10) we give them the formats that are working for us in ads. So the script, everything that's working, and they'll send us back the footage and we give them 5% of ad spend. So they're incentivized that way. And so the first one we tested the beginning of last week and she heard it was Becca. She had a view, a video that got 20 million views using our platform. So we reached out.
(26:32) She made the content, we cut it up using our app and and then we and it's now getting like 16 k a day worth of spend all of our ads. And so this is like great for her and for me. I'm like sure 5% of ad spend like I'm the one managing the ads right now. It doesn't matter. That's like for ads that are crushing. And the CPA is like 30% lower than what we were seeing on everything else.
(26:57) So yeah, it's you're painter like 25 grand a month if that, if that holds at like 15 K, which is like no joke for, for a content creator for making one banger ad for you. So yeah, that's super cool. That's super cool. But now it's like, since then we've had like two other creators submit their own content, and so now they're all like in the platform fighting for the spend.
(27:18) That's like the best though. Like, are you going to like, they know about each other and like, what's going on? Like you're going to throw them in a group chat and make them make them competitive or keep it separate for now. And I'm wondering if we want to keep it like limited to ten spots or something. It has something we might do.
(27:34) I don't know, it depends on on how much it scales. Yeah, well, it's like everybody's I know, like tribe is getting a lot of momentum right now because like tribe enables the TikTok shop of occasion of of meta content. Like. And it sounds like you kind of have the same thing, but yeah, it's like you, you don't want to.
(27:48) It's a balance between getting a bunch of new creators, and then all of a sudden you're out of count, like, yeah, you net increase spend. But maybe they feel like I'm earning less as a percentage because mine don't spend as much. But it's like, I don't know, just keep making bangers and then, you know, you keep earning.
(28:02) Yeah, exactly. Yeah. That's sick. So they sounds like there's like a couple there's a couple of like things that made like Sep function changes and like, I've tweeted at least three times about like RR roughly is there. So it sounds like a couple of those were offer to a degree and like the annual switch and then the the creator flywheel is like recently it sounds like it's cranking.
(28:27) Is there anything else that you did like any landing page wins. Any any other adjustments you've made that made a step function change? Well, one of the big ones is like there's more on the product side. I would say it's when we launched the video editor. So a lot of it was product feedback. For the first few months, I was doing all of the customer service, and so I was able to see what people wanted in the app and what they wanted out of a tool like this.
(28:54) And so being able to reduce churn by providing that outsized value with the product itself has been a huge lever for us being able to scale ads. That's that's why one of the biggest ones I'm I'm super curious on this. Like, I don't understand pricing of like SaaS and then pricing of like AI usage. Like I'm kind of an idiot when it comes to like all this stuff, so bear with me.
(29:19) But like, I don't understand how you could do like, one click editor for a customer and like, still have margin plus a catch on like $15 a month. Like there's got to be something really interesting that you guys are doing on like the back end to think about like usage and all this other stuff too, that then like contributes to the profitability of the business.
(29:40) Like, can you talk about that a little bit like, I don't know how much you want to share on that or not know for sure. Yeah it's fine. Yeah. Like I'm super curious because like, that's the math that I was doing in my head as we were kind of running through this and I'm like, I just don't understand how this works.
(29:52) So yeah, just give us like the quick TLDR of that. Yeah. So in terms of the actual cost to run the app. So like let's say let's take like the average user maybe spends spends like 5 to $10 a month in actual API fees. But the, the key to the model is similar to I think what like anthropic is doing where they're subsidizing the tokens, where some users or power users, they're hitting their limits every day, their five hour limit, their weekly limit, whatever it is.
(30:29) And they might be not worth the $280 a month cost it is or whatever. Sure. But because of the number of users who aren't those power users, it just it averages out. Yeah. So for us, you know, our actual margin is is pretty good in that like 90%. Yeah. Got it. Cool. Yeah. API fees and like the one called editor itself that's like the cheapest because it just uses like, you know, just has to look at a few frames.
(31:00) So a vision model there and then it also just transcribes, which is pretty cheap because it was just like the transcript. Did that make a big difference on retention? Because it's like people were able to film the content, but it's like, okay, I'm going to take seven hours to edit this fucking thing. Now. I can't even get a live to see.
(31:17) What it does like. Is that, is that like, is it as simple as that? Yeah, it is, but it's also like I'm we have to curb expectations with people now a bit because they'll dump their whole camera roll in there and, and be like, make me a banger. And then they'll complain when it's like, there's no, there's no like, this is all B-roll.
(31:42) You have no, like, coherent story or video or hook or anything that you've done with this at this like, yet. And so it just comes down to us educating the person. So we're taking steps in that regard. Overall, the turn though is, is at a place where, you know, our payback period is like within two months and then everything after that is, is profit.
(32:09) And so like right now we are profitable overall. And then that's just where like it's a business decision at this point where we're like, what are we even building here? Are we building something like do we get investors like what do we do? I don't know if you guys have an answer to that. I, I'm all ears.
(32:28) Like, yeah, we run at a loss to just try to like, get the revenue numbers, revenue multiple to sell this thing. Yeah. I mean it feels like I mean, if you even looked up the amount like how big your Tam of your customer list is, like, it's anyone that has north of. I mean, like, even if you just looked up anyone north of 10,000 followers on on the platforms, it's like any one of those people to a certain degree could be a customer of yours.
(32:53) So it's got to be like huge. Like, have you done any research on Tam of like, how many customers you think that this is even possible? Like, I think there's thoughts like that that I have that I'm like, it's so interesting because it's like, okay, this could be useful for like you said, realtors, this could be useful for people that make UGC content.
(33:10) This could be useful for like the gamut of creators is massive. So I don't know. I think that's that's the thing that is really exciting about it to me. Yeah. No. Me too. Because I do think the Tam is absolutely enormous worldwide. This is like everybody wants to be a creator, whether it's for like the notoriety, the ego boost or to actually make money that way.
(33:36) And so if we can make that process easier in any way, that's like, that's why I can't sleep at night right now. Like I slept like in like five months since January. Yeah. I feel like the most interesting thing is, like, how good you can get that video editor to be because it's like scripts I think are pretty, like scripts are getting better and I think is doing a good job with that.
(34:02) I think knowing the formats of what can go viral, you guys obviously like one peak is done a very good job of of identifying those. But like getting that editor to to actually produce bangers with like as little context as possible. Like that is where I think the source is. Because why would you not use a tool like this then? Yeah. Yeah.
(34:21) And even getting at like 90% of the way there and only having the tools in there that somebody would need for creating content to go viral, it's like not creating, you know, landscape footage for some other use case, like a commercial or something. But like everything in their built around creators is something that I feel like, let's say Cap Cut or Canva.
(34:45) You know, they're they're they're speaking to such wide use cases that to adjust their app for this specific use case. I think we have like some time here to develop this and really get it good. Yeah. And so that's my focus. I'm like the I went I'm speedrunning the path to becoming a developer here. Yeah I'm like the sole developer on this app right now.
(35:11) Yeah it's crazy proud of me. You've got more questions in here. Yeah. So is there I this was one of the questions I asked you on Twitter. Like I was just curious, like generally speaking. So we talked about like, real estate agents is obvious because it's like, yes, I'll sell literally anything. And you've paid for this for a long time.
(35:26) It's like anybody making business content, some people are just trying to generate ad revenue. But like UGC creators, people are making content for people like us that are trying to run up e-com ad spend. It's like, is that, you know, is that cohort like, are you thinking about expanding the product out towards those type of people? Or like, it's not even worth at this point because what you have is already a big enough Tam and you want to master that.
(35:48) It is. It's just not it's not a big focus. Like it'll happen for sure. Like, personally, I've thrown put ads into the platform, the auto cut feature, and it works like you can definitely use it for. UGC there's no reason you can't, and a lot of the people who are using the app, are you actually using it for that purpose to like they're like agency style businesses who are using it for their own clients.
(36:15) And it is like it's the next thing on my list is creating that, like higher tier in the app where somebody can have, you know, the different clients that they have, like more of the agency focus. And then from there we'll just do the same flywheel where when we get somebody get get a good case study, we'll just get them to make the ad for us.
(36:36) Yeah, right. What is I'm bouncing around here because I just remembered that I had this question. You have that the, like the high ticket offer on the back end. So it's twofold. One like is there is there an additional high ticket offer because like it sounds like it was a it's a it's a group and like coaching type of thing, but maybe not like a done for you which is maybe what one peak had previously done.
(36:56) Is there like another expansion to that and then like is that really meaningful for LTV? Like, does it take great high enough that that's like unlocking quite a bit or you're just that's like it's gravy on top of everything else? Yeah, I kind of treat it as gravy. I don't calculate anything based on it based on it at this point. But it does.
(37:14) It does convert pretty well and it adds like 10 to 15% to the LTV overall. Across all across all customers. But yeah it is. Yeah. It's the it's the it's 4K I think for for a four month program they get access to the community. They get access to coaches who like look at their videos, tell them what they can do differently, group coaching calls.
(37:41) And then we are thinking about like higher ticket. the. Yeah. Cool, man. Sweet. This is awesome. I know you could probably continue to go for a long time. Like we didn't even we didn't really even get into like, the AI piece of it at all. I think it was crazy. You know, you have AI like your AI. AI building. I'm not giving you enough credit when I say this, so don't take this the wrong way.
(38:04) But like AI is building the product. Sounds like I might be running the ad account. Obviously Metis got their own AI. Are you still are you still doing that? Like is still running it or something else? Yeah. Yeah. So like my and everything is run through AI for me it's through telegram. It's like my whole command center business.
(38:21) Everything is through telegram. I've got like, it's either Claude or Codex or Cami and I've got like 40 sessions at any given time running on my home computer. I feel like, okay, so, like, I feel like this is like one of the biggest unlocks for anyone using AI. It's like vanilla Claude or vanilla ChatGPT is okay, but it's the harness that you build around it that really matters.
(38:49) It's like for copywriting, it kind of sucks, but the harness being like the skills, the process that you put in place can actually take it to producing something you can actually run in in your ad account. So for me, my harness is built on on everything is telegram. I've got agents running, you know, on their own, identifying bugs, triaging them, bringing them to me with like a solution for for how to solve it, my ad account.
(39:21) Like when I get new creative, I literally just like, send a link to my bot with the, like the slack link. That's how lazy I am at this point. I don't even like export it. And I'm like here, use the format finder, add uploader skill and get these into the ad account and it knows it has a whole process. It knows to like look at, do an analysis on the ad account, determine what the best copy is, transcribe the video.
(39:43) It just got to identify like who it's talking to right? The ad copy like it does everything. It uploads the ads and it just tells me when it's done. And then I can I can go in and confirm. But at this point it hasn't made a mistake in months. So it just it just runs on its own and then it just, it knows to give me an analysis of every morning or every whenever I want all of our customer services run through like a slack bot that I've got set up, which is like the company brain as well.
(40:11) And it's a clawed bot in slack where it triaged all customer service as it comes in, identifies like, oh, is this a refund? This is what I want to do. This is the draft reply. I've checked the previous tickets that are very similar. All you have to do is to prove and so the customer service agent can clear. Like I'd take it a minute at this point and she just hits approve and it goes in.
(40:34) Then it goes in and it does the refund updates the user account and sends the email. And then I have another bot that comes in and just verifies that it did what it said it was going to do, just like as like a QA. And at the end of the day, there's another one that comes in and checks that whole day, make sure nothing is missed because like, you even have these redundancy layers built in because, you know, they give you peace of mind, but then it's just also like, why not? Why wouldn't you kind of every week a dollar. Yeah, exactly.
(41:01) It costs like a dollar. And so yeah, I don't know. It's sick. I'm so stoked on AI right now. I gotta love every bit of it. It's so much. Yeah. Andrew, Andrew's not here, but Andrew has a has another podcast, and it sounds like I'm gonna have to come. Come on and join it. But it's called I think it's called AI, dtc tf or something crazy.
(41:20) A lot of acronyms and I know it's DTC specific, but we'll have to just like do a side episode or we just nerd out on like the AI side of it because we're not going to have time today. But is there anything else that, like, we didn't go through that you think is like super interesting to share that you want to share? Yeah.
(41:37) Not on the unless it's on the AI side. That's like where where all my, my like time is going towards is just creating the process here. Like I feel like what we're doing would require 10 to 15 like developers on this thing, but we're being forced to create process that hasn't that no one's ever not no one's ever done.
(42:00) But like that is cutting edge, let's say with AI. So that's like one of the funnest, funnest things for me right now. Yeah. I'm gonna text I'm going to text Andrew and tell him he needs to set up the, the follow up to talk through. Yeah. Oh, man. Well appreciate you appreciate making time to to join us. It's it's been really fun.
(42:18) And I know this was our first time getting a chance to chat. And sounds like you and Zach got to reconnect for the first time in a while, so that's awesome. Yeah. Thanks for having me. Before we hit record, you mentioned you might start tweeting a little bit more and I got hold you to it, but is that where you want people to find you? Do you want people to reach out? Do you want to leave you alone? No. Yeah, definitely reach out.
(42:36) I mean, like I said before, I need help. Like, I'm like to a large degree in over my head here, like, as a developer or as somebody who's, like, building a company that might be worth something. If anyone has any advice, you know, let me know. My my Twitter handle is nav ud. So cool. I'll make sure it's linked up in the show notes.
(42:59) But yeah man appreciate you joining us. It was super fun and I'm sure we'll we'll we'll have a follow up on all things AI coming soon. Sounds good. Thanks for having me. Thanks, man. The only way that we grow this podcast is by you sharing it with your friends. Honestly, like reviews kind of don't really mean anything too much anymore.
(43:19) They're really meaningful, but they don't do a lot for the growth of the podcast. And so sharing YouTube links, sharing Spotify links, sharing Apple, whatever we call it under the podcast app now, anything you can share, the better we're going to be. Guys, anything else you want to say on this? Yeah, please go check us out on YouTube.
(43:38) Rack up those views for us. We'd love to see it and then subscribe. Make sure to subscribe on YouTube as well. And I relentlessly refresh the YouTube comments because it dictates my mental health for the day. So please say something nice about all of us. Thank you everyone. Thanks for listening. Honestly.
Made With AI