How I Built my AI Tool SubMagic to $667k/Month With a Tiny Team and No Funding
44:11 Share

Today we are at 8 million ARR, the company is 3 years old. We have 5 to 10 thousand signups a day. Do you have an idea of how many new paid customers you got last month? I think it was around 2,500 customers last month. We have more than 10,000 FES today. We have a very generous program. We give 30% revenue on the lifetime value of the customers.

Our churn is around 15% today. So it's very high month over month. Do you have a sense of, you said it's not big. I mean, we're talking like under 20K a month in paid ads? Maybe, probably more. I would say between 20 and 50K a month. Hey folks, my guest today is David Zetoun. He is building SubMagic.com.

Co. It is a tool that enables you to create viral shorts very quickly from short form or long form video. They're launching new products like their AI caption tool today. We're excited to have them on the show. David, you ready to take us to the top? Thank you very much for receiving me, Nathan. Pleasure to be here. You bet. You know, us Americans, all we care about is money. You know, you're pure still coming up in France. So I don't want to bury the lead here. Before we jump into your product and sort of what you do, can you share what Topline Revenue is today?

Yeah, today we are at 8 million ARR. The company is two years old. We're still very young. Most of our business is coming from the US, so from your country, Nathan, and we are proud of that as we are French originally with my co-founder. But yeah, small team, 13 people in the company today. And yeah.

That's incredible. And we're recording this on June 5th, so everyone has context. So David, this is obviously a very hot space and AI and video, you know, video is obviously tricky to deal with because the files are so huge and there's so much context you have to have. And then you have to sort of make the shorts, the captions, et cetera. So I'm really interested to learn and sort of get more into the technology. Before we do that, though, let's take people through sort of the product experience here, right? So I went ahead and recorded this just ahead of time.

Just so you could talk through it as we go through it here. So obviously we're on the website, right? It's viral shorts. We're going to go in here and I'm going to click try for free, create free shorts. One thing you do differently than others. You don't ask a bunch of qualifying questions in your onboarding. You're going to see here, you take me directly into like the email and I'm in the interface. Why'd you make that decision?

Okay, so one of the reasons why we did that is that we wanted to build like a TwiClix products and wanted to do everything in TwiClix to make the experience super easy. So this was most the focus of my co-founder that is a technical guy. He spent like eight years before building Submagic building a website builder and he learned about like the importance of simplicity and how you can make everything in TwiClix.

Editing videos... Are you saying in three clicks? In three clicks, yes. Three clicks. Yeah. And it's very, I would say, trendy to say that, but the reality is editing videos is complex. And I've been editing videos since I'm 12 years old. I'm 28 today. So I've tried all the tools, Sony Vegas, Premiere Pro, all the complex tools. And I know how to edit and it still remains complex for me. So...

Trying to get this three-click experience as much as possible on a short form content product was, I would say, the most important thing for us as a product. So we try to avoid all the distraction around the product and be really focused on what matters for the users.

Yep. This was my first moment where I was like, dang it. If I wasn't about to interview this guy, I might drop off here because I uploaded a video that was a 20 minute interview. And I'm like, Oh, it's too long. I don't want to go do the work to edit it down short. But then I saw this extra feature you, you guys launched recently. It looks like over here where it said, wait, get clips from a long video. So, so I then, I then tested that. Help me understand that. Is that, was that just a demand you got from the user base, longer videos?

Yeah, so we started with the captions on Submagic. The tool was super focused on one thing that we wanted to do it better than everybody else. And it was how to get, we can add like amazing captions on your short form content to make your content go viral.

And so we started with this, we had to be focused on that. And then later on, we had a lot of people asking us like, hey guys, it could be amazing if we could put like our long form content into Submagic and you find automatically like the best like viral moments. So because we know already how to make a content that goes viral with the editing, we added these tools called magic clips that you're trying to use. And so, yeah, it's amazing. It's great.

In this new age of AI, obviously pricing and paywalls and usage-based pricing versus seat-based pricing is a very hot topic. So I always like the moment when the user first hits that paywall experience. You guys chose to put it right here. Help me understand sort of why this was the right moment to introduce the paywall. That's a good question.

I don't know when you tried that, but it's been like, I think very recently, potentially. I did it like literally 50 minutes ago. Yeah. So we are always testing new things and we love testing things. We don't know anything. We're just trying things and see what works. So before that, before like a week ago, Magic Clips was free for everyone. They could try like at least one Magic Clips.

And so the result we had is that we have a lot of users trying Magic Clips with a conversion rate what is pretty much okay, but we wanted to see like, is it really useful to give like a free clips to these users? And let's try to just put the payroll and see what is the results.

So we are still in the experimentation, but yeah, I will let you know later what's going on with this. We might go back to what we did before, but it's just a test. And I think it's important as a company. And this is one of the reasons why Submagic, we grew that fast is that we've been able to test a lot of things very fast and to iterate. Are you using any tool to, sorry to cut you off David there, but are you using any tool to run these paywall tests like Superwall or are these all custom coded by our engineering team? No, we don't use any tool.

At the moment. And I'm not the tech guy, but I know that we, I'm pretty sure we don't use any tool. Yep. Yep. Okay. So anyway, so I hit this paywall and then I go, okay, well, let me see what the monthly price is. Let me toggle off the billing. And then I said, okay, wait, let me exit out of this and explore the interface a little bit more before I convert to paid. So clicked around a little bit.

wanted to understand more about the product. Let me fast forward here. Wanted to see more about the full pricing strategy. I think at this moment, David would be a good one to actually dive into your guys' pricing page and sort of how you guys from a business perspective decided to set up some of this stuff. So let me, let me actually just share your live website here. All right. So this is how you guys have structured. And I think to your point, you're saying this little toggle here is a new thing. You just started adding this as an add on the magic clips.

Correct. We started Magic Clips five months ago as an add-on. It was a bet because we could have put Magic Clips just available for the pro plan or the business plan, including to the plan. But we wanted to try it as an add-on because it gives so much value for the users and it's so much a product that is not directly linked for what the users used to pay before having Magic Clips. So we tried this add-on thing.

You have a really interesting pricing page because you have like most AI tools basically have this section, right? Which is the usage based upselling based off some output, right? You do, you do number of videos and then the max length of the videos is the bet you're making on the usage metric. And then, but very few of these new AI tools are still charging on the seat base. The logic is they want as many team members using it to drive these values as fast as possible. Why did you decide to restrict the,

Why did you decide to basically restrict both? To put paywalls up for both? So for the numbers of videos you can do per month, this is the only metrics we can like. This is the number one reason why people are upgrading. They need more videos. So if we don't restrict on that, it's a way for us to improve the expansion, I would say. The second thing is that

We try to divide the different plan based on the different type of users that we have, the different type of ICP. We have kind of like two types of ICP. We have the people who are using Submagic for themselves, for their own videos, small business owners starting short form contents, don't know anything about it and want to discover it. And then we have the agencies. We have so much marketing agencies, video short form content agencies, and those users have different needs.

We really wanted to try to see if you are this kind of SAP, you should take this plan. This is for you. And so that's why as a small business owner, you don't need to have five members in your team. It's not useful. None of them are inviting people.

There's a lot of folks that always try to reverse engineer virality, right? They put that watermark in the free plan and then they count how many new users they get from those free plan click-throughs every month. Was that an effective strategy or is that an effective strategy for you guys?

Yeah, there is a free Mium at Submagic. You can use Submagic for free for three credits on the top, like the trial three videos without credit card with the watermark. Yes, it's great, but it's very difficult to quantify the impact on the watermark and to see really how much people are posting videos with a watermark online and how many views this video are making.

But it's a way for the users to try the product without paying. And I think we love that. It's important for the users to try. Oh, yeah. It's a good point. Most people, when they run this virality playbook, there's like a powered by button under the thing. But to your point, you're a video. So you have no way to track if the watermark inside one of your customers' videos gets clicked and drives you a new customer. You're just betting. The extra brand exposures is a good thing for you.

Yeah, very interesting. Okay, let's go back to the early days. So you just celebrated your two-year birthday, I think. Take us back to sort of a launch date. I always like to go in here and look at sort of what perplexity tells me here. But is this accurate? Perplexity is saying 2023 is when sort of the first user was acquired. You start off with a team size of nine.

We started with, I like this table, it's great. So we started with two, just Sifei, my co-founder and I. We started, the first customer we had, it was in May 2023. Yeah, okay. So pretty much. That's great. And,

And then you went from zero to a million fairly rapidly. When did you hit a million in revenue? So you can see Perplexity gets this stuff wrong, right? That data is obviously wrong. When did you hit a million of ARR? So we hit a million in ARR three months exactly after we get our first customers. So we had our first customer the 1st of May, 2023. We reached a million ARR the 1st of August.

Exactly. Wow. Okay. August 2023 million. And what drove most the growth from zero to 83? If you had to pick only two growth channels, what were the two most impactful? I would say short form content. I will explain just after, but I would say short form content and timing. Timing because there is a time to market, like time market fit that I really believe in business in general. Like we were at the right time with the right product.

with the arrival of ChatGPT at this time, everybody wants to make like short form. Alex Almosy was like literally booming online. Everybody wants to like start understanding like, hey, we should do short form content and we should make caption as well because everybody, like all those influencers are making caption, we should do the same.

And then short form because at the beginning of Submagic we had a lot of viral short form content coming not from us but from users saying like "Hey, I discovered this new tool called Submagic. They can allow you to make three clicks captions super easily on the product." And they showed a product like this and so many videos went viral. And this two mix was just an explosion I would say.

Yeah, yeah, it's interesting. I wasn't expecting you to give those two answers. What I thought you were going to say was Your affiliate program and product time. So I want to dig deeper here, right? So you saw that natural organic posting like happening, but you then launched your affiliate program I think fairly early on when did you launch that and and would you consider that a key to going from zero to a million quickly? We launched the affiliation program pretty much like

I think it was 30 days after we had our first customers because at the beginning of Submagic, we had zero euros to invest into marketing. And I was used to do affiliate marketing in my first company where I was myself an affiliate for other businesses. And so affiliation is amazing when you don't have a lot of revenue. It's super easy.

It's a good way to attract people and pay them only if they make sales. So we had no choice to do affiliation. So I started this affiliate campaign and I was just telling people, if you make a viral short form content talking about SOMAGIC and it goes viral, you're going to make a lot of revenue because

Myself, I did it once and the short form went, I think it did like 100,000 views and it drives like $3,000 MRR directly. And I said to my users, like, this is the proof that it's working. You should do like kind of content like this. I really literally teach them how to make viral short form content talking about some magic.

And they just like keep posting content over content over content every day because they were making revenue. The success of a good affiliation program is your affiliates needs to make revenue and needs to make money out of it. So yes, affiliation is- Can you share that number? How many total affiliates today have made money from you?

crazy man i don't know exactly the numbers we have more than 10 000 fes today uh we have a very like generous program we give 30 percent revenue on the lifetime value of the customers so if you nathan you become a fia like a become a fia tomorrow and you you just bring your customers to some magic every month that they will pay some magic you will get your commissions

And it's pretty generous. Which is rare. Most people, most people do not do that for life. They do it for the first year, then it goes to zero. So I would agree. I mean, it's a very generous program. I mean, can you quantify that? I mean, if you have a large chunk of your 8 million revenue coming in, it's from affiliates, you will have a large chunk of your cogs be affiliate payouts. Do you know how much you paid out to affiliates last month?

Yeah, yeah, we know exactly. So affiliation today represents 20% of our revenue. Like 20% of our global revenue is coming from affiliation. So 20% of the 8 million of your ARR today. So I mean, that's a big chunk. That's 2 million. I mean, that's 1.7, 1.8 million.

Wow. Okay. So, so if that's the case, if 1.8 million of ARR is coming, has, has, is active today on sub magic and has come from affiliates, divide that by 12, that's about 150,000 bucks of monthly revenue. And you're paying out 30, you said 30%, right? To most affiliates. Yeah. Yeah. So what that's like for you. Okay. So you're spending out 45,000 bucks a month in affiliate payouts, something like that.

yeah but one thing I can tell you for affiliation people might see that oh it's crazy you give 30 commission it's huge like we had a lot of people's like saying like baby you're crazy to do that first we didn't have the choice at the beginning like the company was like we had zero customers and we need to be generous and secondly like people see like are you gonna give 30 it's huge and stuff like that but they say like you don't see how much

business affiliation is bringing to your business, like without tracking the link. There are so many videos like when you make a huge video talking about Submagic, maybe like a small percentage is going to click on the link. But a lot of them are going to see the brand Submagic again and again and again. And this is for free. And this is amazing. And you cannot like quantify how much is it. So it's great.

So David, just to sum that up, what you're saying is you can directly attribute about $150,000 or about $2 million of ARR to affiliate traffic, but you're also betting a lot of your other $8 million of revenue or other $6 million of revenue may have been influenced by some affiliate branding, even if they didn't click through that link directly. Well, for sure. It makes sense for you, no, Nathan? 100%. 100%. No, I agree with you. I'm playing dumb a little bit here because we want to learn. We want to capture all the good stuff. Take us back to that first...

Yeah. The other thing I want to touch on, and again, the reason I'm hitting on all these, I like to sort of create frameworks for my audience. And one of the frameworks I like to have them think about is there's really only about 34 ways to grow your software company, right? There's cheap stuff, there's expensive stuff, there's fast stuff, and there's stuff that takes more time. You over here used influencer marketing, you used affiliates. We have Product Hunt here. You used Product Hunt early on. You did three launches. Ironically, you're one in 2023, which was what this was

four months after you launched. Take us through this launch.

Two months after we launched. First of May, third of July, someone told us like, guys, you should do a product. Because at this time we had zero customers in the US. And my co-founder and I decided to say, like, we should change the strategy. We are going fully into the US. Forget about the French country. Let's go fully. And so he said, like, there is one thing we can try is product. Apparently something good for software. So we try that. And from this day, it was the best decision ever because

Our main market became the US from this day. Is Alex Ramosi an affiliate of you guys? No. We would love to, but he's not an affiliate of so much. He loves the product. I met Alex Ramosi a year ago. No, he's not an affiliate. We would love to, but he's not. That's great. Any tactics you would give other founders who are trying to figure out how to hit the number one or number two spot on Procter & Gamble like you did?

It's really hard. I wouldn't recommend everybody to do product hunt again, to be honest. Because I think it changed a lot since we launched two years ago.

You haven't been able to really get that same sort of upvoting in the past two years. It's gotten harder. It's super hard. The energy you have to put to become number two, it's like 24 hours reaching out to everyone to say, please upvote for my product, check in my product. So it's a lot of efforts. I think when you just started and you don't have a lot of choice, it's something good to do. But we don't care if we are like eight in the day or ten when we launch again on product. It's just...

Wow. I don't see if, yeah. Okay, well, cool. Let's fast forward. So May 2023, you launch at your first million of revenue in August of 2023, a couple months later, and then about a year ago, I'm recording this of June 2025. So call it June 2024. You talk about hitting your 1 millionth user. Walk us through this milestone. What drove most of this growth?

Yeah man, it was an incredible milestone. For me, 1 million users is much more incredible than 1 million error. Because it's a 1 million period and I always like imagine myself like a stadium, 80,000 people, multiple by 12, 13, like I'm like wow. It's huge and they tried my product at least once.

I think it's a mix of the time to market that I told you, the good marketing that we did since the beginning of Submagic, we push super hard on the brand and on the marketing through affiliation, SEO, paid ads, influence marketing. And there is this kind of like word of mouth that I've been driven a lot of growth for Submagic. We are super focused on the customer experience. And as a bootstrap company, like the best marketing you can have is like word of mouth. It's free.

And so we really get inspired by Brian Chesky from Airbnb about this 10-star customer experience, how you can give an amazing experience to your users. And so we try to do super incredible things with our users, creating WhatsApp groups with them, building the product directly hand-to-hand with them. I take approximately five to six calls with users every day since the beginning of Submagic.

like everyone can book a call with the ceo of some magic every time they want and because i want to get to be super close to the users and every time like they like this experiment they want to talk about some magic around themselves it's there is no software they can offer that on the market at the moment and that makes me proud so it's a mix i would say

Yep. You'd mentioned SEO paid ads in community. The first thing that I hit once I launched a paid plan was a pop-up where you said, Hey, join our discord community. Um, and so you have discord here. You just mentioned WhatsApp as well. You got 7,600 members here. What besides the strategy for other founders listening, saying, Hey, make sure you invite people into a community, any tactics for how to manage this effectively?

managing the community is very hard uh i would say we are not the best today on discord on whatsapp we try to do something very very uh it was especially on the beginning what i would suggest to any founders is like the first hundred customer of some magic i took on stripe the information i reach out to them i say

Send me a text message, please. I will give you like a discount for the next month, but please send me a text message I need to get your number and those people I put them into a whatsapp group and I built the product with those hundred people I was super close to those people they were like we say like why she say like find hundred people that love your product and

And I try to do the kind of the same thing. And this is free. And you will understand that those people will become your best product manager ever because they will tell you what features they want and how we should do it. Like, it's so like, super like people try to find solution to coaches, to online, to whatever. The solution of your business is in the answer of your customers. Talk to them, be close to your customers. They have the answer.

It's a cheat code. Yeah, it's a cheat code, but be careful. Not all your customers will be good customers.

You have to determine if this guy, like feedback, is a good customer, if he's really the guy that we're trying to hit with our marketing or not. And does this feature is a really must-have feature or it's just a nice to have? And you have to be very careful about listening to your users and then taking good decisions. But yeah, this is a cheat code for me.

I want to get into your team here in a second, which you put up a great post on about three weeks ago. But I have to touch on the two other growth channels you said you're using aggressively, which is SEO and paid ads. When I look at your SEO traffic, I mean really fast growth from 2023 to 2025 in terms of domain rating. You saw a huge pop in organic traffic. I'm curious if this was intentional up to 9 million and then dropped back down to about 235. Walk us through what's intentional around your SEO strategy.

Yeah. So I'm not the best in SEO, to be honest. I don't know so much about it. But what I can tell you is the pics that you saw, it's a test that we have been trying. It's like we launched a free tools strategy at some point, at some magic. So we created free tools like how to compress your video, how to download video from YouTube and stuff like that. And those tools were performing so well. Like in one month, it explodes.

And the one that really exploded-- - Can you quantify that in any way? Like number of website clicks, number of customers, any kind of quantification there? - The conversion rate was really bad, to be honest. And we learned from it. We had like nine million, like at some point, one million visitors a month. Like it was crazy. Like maybe even more, I don't remember exactly, but you saw the peak. It was huge, but the reality is that those things don't convert and don't attract really good customers.

And after a month or two, we had Google that blocks that because they don't want people to download YouTube videos for like trying to like download YouTube videos. So they blocked this page. That's why the traffic went down. But it was a good learning. We still gonna make some efforts to improve this free tool strategy. But it's not what drives us the most revenue in terms of SEO today.

Yeah, it's a really good strategy. We put that up here in terms of free tools because we've seen other companies just crush it with this. But to your point, if Google stops your ability to sort of process YouTube videos because of your niche, it makes this particular tactic hard. But it's super cool to see that spike when it was working going up to like 9 million views per month. I mean, this is just insane organic traffic.

One thing I would say is like free tools can be awesome for your company, but you need to really be careful about what kind of free tool should we do. And it needs to make sense regarding the value proposition of your product. So for us, like downloading your YouTube videos is too far from why people should use Submagic.

And otherwise, doing something like compressing your videos or finding titles is much more linked to why people should buy Submagic. You need to just put your place at the place of your clients and it needs to make sense. So I think it can be awesome. I see so many great strategies about free tools. We still need to learn a lot from that.

I love that. David, I want to be respectful of your time and we're at the limit we set. Do you have 10 more minutes to keep going? Yeah, sure. We can do that for you. Okay, great. So that was free tools. Before we jump into your team, can you touch on paid ads for a second? Are you spending a meaningful amount per month on paid ads?

I know that we're spending on Facebook ads, Google ads. We're going to start LinkedIn ads soon to try. I cannot tell you exactly how much we spend on this, but it's not like huge. We're getting our old marketing is not representing something huge. We try to avoid paid marketing as much as possible. Yeah, yeah. I mean, as a bootstrapper, and you're basically, I mean, you've raised, it sounds like a small, how much have you raised total? It's a very small amount, right? We never raised.

Oh, you didn't do like a community round of like half a million bucks? No. Ah, okay, okay. I misunderstood that. So you guys are bootstrapped to eight million bucks of revenue? Yeah. Yeah, I love that. So obviously paid ads, those are the most expensive advertising you can do. Do you have a sense, I'm on your guys' Facebook page right now, do you have a sense of, you said it's not big, I mean, are we talking like under 20K a month in paid ads?

Maybe probably more, I would say between 20 and 50k a month. Maybe like around 50k I would say. Okay, it's scaling. How have you set that up? Did you hire internally someone on your team to do that? Did you use an agency? How are you doing this? Yeah, we have like someone fully dedicated to ads at the moment on Facebook ads in our team. His name is Antoine, he's an absolute legend.

He's just trying things all day long. And we have also agencies that help us crafting some good product like content ads. It's not easy, but we try stuff. And we have also an agency that we work with for Google Ads that is not part of the team, but it's an agency. Do you want to recommend him? Should we give him some love?

Yes, it's a French agency. If you want, I can give you the full name and you can put it in the description. It's called Kiebers. K-I-E-B-E-R-S. It's a French agency. Kiebers Agency. This one? No, sorry. We moved the E. K-I-B-E. Yes, exactly. Kiebers.

- Premium DDC Growth and Creative Agency. - Exactly. - Very cool. Are they also a French group that you guys found? - Yeah, it's a French agency.

This one? Yeah, exactly. Great. That's awesome. Very cool. Okay, so we'll give them some love. You told me it was really important. You've got to give some love to your team. I mean, you guys are 8 million of ARR. You are 13 people. I mean, so you're cranking $700,000 of revenue per employee, which is really impressive efficiency. While I play this awesome reel that you made on your team, walk us through your org chart today. What's it look like? So...

Okay, so the team is divided in three. We have the product team with the tech guy, we have the marketing team, and we have the customer experience team. These are for us the three most important things in a software company.

And so we have very atypical people in the team. Like everyone is atypical. We have the average age is 28. I'm 28. 70% of the team doesn't have any degrees. We all are driven by the same stuff. That is like we want to deliver value and learn from our experience. So we love learnings. And yeah,

And yeah, it's amazing. We are all remote. We met once a year in a seminar. So it was this year in Thailand two months ago. It was amazing. So this video is part of the trip. Last year, we were in Tuscany in Italy. And...

yeah it's it's great we we we work with just the top of their field they are really like top notch on on what they are doing they are super like open to discuss so if anyone wants to learn from like experience they love like exchanging the learnings so we sometimes Antoine from the growth will contact also growth manager to discuss about what they learned the last two months and and try to find like yeah Ellie is a legend is our is our CMO and how do you find these folks David how do you how do you recruit um

So the funny part is like I would say 80% of the team, so eight people out of 13, I would say kind of like this, are coming from Submagic. They are users of Submagic first. They've been using Submagic for like content they wanted to create and we just send emails to our database. We are hiring and part of the team is coming from Submagic and this is awesome because it makes like a sensation about like they know the product. They use the product so they know what they're going to work on.

interesting that makes a lot of sense are you are you was or would you recommend aura for for organic SEO it looks like arena is saying she formerly worked with you guys so I'm not managing this here directly Ellie will will probably if if any reposted this that makes sense so yes potentially yes we recommend arena I don't I don't remember you know myself I don't work with the SEO directly but um yeah yeah everyone is very free tool

yeah he's saying that strategy you guys do like really worked i mean uh four million organic monthly traffic valued at 800 000 bucks i mean good good traffic yeah yeah it brings good traffic it's good for the brand as well yes it was it was a positive things for sure but uh i will not tell you it's the most important thing on the seo but it it was uh great for sure yeah

This is great. Well, we've hit a lot of the big stuff I want to chat about. I really want to, I mean, just getting the product for a second. You know, I run, obviously, a media business. We're recording a podcast, right? So I uploaded that long-form video. When will it actually show? I guess I started that upload probably about an hour ago. Is there a progress bar somewhere? Are you sure you upload? Because normally when you upload a video, you should see it on your projects. But if you want, I can. I'm sure it's user error. Hold on.

Magic Clips. Where was the button for the longer stuff? Hold on. Magic Clips. If you click on Magic Clips on the right. From long video. You can click on the video down, the future of Magic Care Bag. Okay. So I did this about an hour ago. You tell me if I did something wrong. It says your vial is uploading, but then I don't know where to go find that upload status. Oh, okay.

That shouldn't be that. So I need to check what's happening. There might be a thing we need to figure it out. Okay. What else should I click so people can see a real demo? Just do the demo video? Yeah, you can click here. Okay, interesting. How are you deciding what are these key moments? Are you looking for syntax changes in the voice or the tonality or is it the transcription? Yeah.

Yeah, there are like hundreds of parameters that we configured with the AI about like what makes a video captivating, what makes a video valuable. So about what you say, are you talking about something sensitive or not? Are you talking about money? What are like the conflicts in the voice? What are the humorous moments? And so we build the clips like this. And we are like super proud of what kind of clips we can generate with AI today. It's amazing.

One of the challenges I'm facing with all these AI tools I'm testing is I want to be able to take my memory and store it in Submagic. In other words, I want to be able to hover over this moment here and have you guys tell me the logic you use to pick this thing. You might say, this sounded like something related to money, so we flagged it. And I want to then add the memory and say, good job. My audience loves money clips. Do more of this. Do you have that ability yet to take your user's memory into your UI?

Not yet, but this is something we had the feedback on. Exactly the same thing as you said, I want to train my own Submagic AI about what I like and about what my users like. So you will be able to, as you can see next to the clip, if you click next to, you have a plus button, like thumb ups or thumb down on the right. Yes. If you select this clip, we'll know what clips did you like, what clips you didn't like. And then we can go after when you click on edit, you can then edit your clip.

and then add your captions, change your style, add your music, do everything you need, add your b-rolls on the top when it's ready. And yeah, it's very simple to use. But yeah, this is something that's going to come up on Submagic super soon. Yeah, very cool. Anything else you want to showcase on the product? If you want, I can share my screen maybe with the project. Yeah, do it. That's always good. Yeah.

you should see your project that's right so let's take this clip off so when you're in some magic uh it's it's very simple we want to get access like the users get access to everything you need in just three clicks so you can choose your style in just one the best style on the market are here and then you can edit your caption like this you can add the emojis you want

super easily change the position of the caption if you need, add your music. We put all the best trendy music on the market and also the premium music that we have a partnership with Storyblocks.

And finally, what you need to make a great short form content, you need a hook title. So if you need a hook title, you're just going to go on edit caption, generate your hook title. And this is going to make your short form much more engaging at the beginning. So this is a kind of a summary of what you're going to see in the video. So for example, I'm going to put this one, put it a little bit bigger. And then when the video is going to start, you're going to have this kind of like... Oh, cool. Yeah.

And then you can add the people. David, how do you know like, okay, imagine this gets done. We're recording on Riverside. I want to take this and publish it. I want to create a 60-second short to teach other founders how to build an affiliate program using what you just taught us. But I need, there's three screens. There's my face, there's your face, and then there's the screen share asset. Have you built technology into SubMagic where it will know when to cut to my face asking the question, your face answering, and that third track, the screenshot? Yeah.

Yeah, so let's say this is a long form to short. So this is a long clip that we take short form out of it. So if you click on the clip, you can now change the camera and decide it like, for example, you can change and say it's going to be like two faces. And you can then change to like, if it's two faces, decide it in the video which faces you want to put.

And then how do we get that third track? This is where I haven't seen any video editor do yet is the faces a lot of like people do well, but what no one's doing well yet is that third option, that screen share shot. Yeah. We don't offer the screen share shot yet, but this is coming. This is something that's going to come on some magic so far because it's so useful. And I know exactly the use case and this is on the roadmap already. So it's going to be like,

This is super cool, man. I'm so thrilled. I mean, it's so rare you find a company not only crushing it, but also bootstrapping it with a small team. You're hyper-efficient. Anything else that you want to touch on before we wrap up here? I think we talked about everything that I wanted to share with the audience that you have. I think I would just say something regarding bootstrapping a company today. It's even more true regarding AI. We are able to be bootstrapped because AI is helping us a lot.

And like without AI, we should have been maybe 40 in the company or 50 people to do what we do. So I will say like use AI as much as possible. We try to push all the teams to use AI for everything they do, for marketing, for customer support, for products. It's so useful. It's crazy. Like Cloud 4.0, I don't know if you saw the last one, is just insane.

Well, what are you using? Are you building agents in N8n or Gumloop? Can you share what are your backend tools you use for this stuff? Me personally, I use a lot of ChatGPT. I use ChatGPT for everything. I talk a lot to ChatGPT myself, like with voice notes. I explain a lot of the context of what I do. So it helps me much better with the context.

My co-founder in technical part, I couldn't speak for him, but I know that they use Cursor, they use Cloud 4.0, they use some stuff like this. I don't want to go too much into details because I might say some stupid things, but I know that they use AI a lot. But just by using GPT, you can scale so much. It's so useful.

No, it's amazing. And I guess before I let you go, let's just round out sort of, again, we're recording this in June. Let's just round out sort of everything where you're at. So we know you're at 13 people today. You've grown bootstrapped about 8 million bucks of revenue. 20% of that has come from affiliates. How many new signups, just free and paid, did you get last month or a range? I think we have like,

around like 8,000 or 10,000 new signups every day. So potentially, like I don't want to take me there, but I think around like even five to 10,000 signups a day. That's wild. So 200,000 in a month, 200, 250,000 in a month on average. Yeah. Do you have a target, David, on how many of those you consider a good conversion rate to paid in their first session? Like that? No, I don't. I don't. I don't really I don't want to say like bullshit numbers, but no.

Do you have an idea of how many new paid customers you got last month? Yeah, we had around like I think it was around 2500 customers last month. So you can have the conversion rate. It's very low.

Well, yeah, but if you're not spending a ton on top of funnel, you don't need to have high conversion rate to make a lot of money to do well and to grow. So that makes a ton of sense. And so 2500 new customers and a lot of these tools are having churn issues. You're seeing 567% gross logo churn on a monthly basis. How are you guys thinking about churn? Are you comfortable sharing your churn? Yeah, sure. Churn is high for AI tool in general. I would say we have a high churn.

Our churn is around 15% today. So it's very high month over month. That's logo churn per month? What do you mean by logo? So there's two types of churn. There's revenue that's churning, right? And then there's just number of logos churning. And the reason that there's a difference is because some customers pay way less than other customers.

So the way we calculate the churn is like when I say it's 15% is like if we have 100 customers at the beginning of the month, at the end of the month, 15% of them will leave. So magic.

Yep. So I don't know if it's logo. That's going to be logo churn. Okay. Yeah. My thinking is some of those customers are probably stickier. If you take your agency cohort, they probably will be stickier than your small business cohort. For sure. And we already know that the higher plan are churning less than the lower plan.

But the churn is high and I think it's not something crazy. Of course, it's not a metric that we like because I don't know any SaaS companies that like having a high churn. But people want to try product, they want to try AI tool, they want to discover, they want to see if short form is for them. And it's fine. It's part of the usage of Submagic. And the market is so huge. It's just the beginning of people knowing about they can do short form for the business.

They have so much value by making short form to grow their activities, to make their work known about the world. It's amazing. We just launched a feature yesterday, enable the users to translate the captions into like 50 different languages. So you can speak English and have captions in French or in English or in Spanish or whatever. So you can reach much more audience from just one videos. And this is why people love short form. And I believe it's only the beginning.

My bet, and feel free to challenge me on this, my bet is these AI tools, the first tools that in a specific niche, so you'd be video editing, that allow me to store my memory. Me teaching and sticking the memory into Submagic is what creates the lock-in. And I think Churn's going to drop the second people, whoever launches the memory stuff first, I think wins the market.

You are right. I think you are right. One thing that we plan to do is just connecting users' accounts to their social media directly from SubMagic so we can know directly which video performs well on their socials coming from SubMagic. So we can learn from that. And the goal is just to make our users more successful. So what you told me just before regarding magic clips and saying, oh, I like these clips. This is what I want to see. Oh, these clips perform well on my Instagram. Or I get two leads from these videos as well.

We will learn from this to make like a better experience for our users and just make them more successful. But I think you're right. Yeah. I mean, the way I do this now, I know what my audience wants. I will take this clip, this whole thing we just recorded with you and me, and I will say, go timestamp every time David said a numerical number related to growth or finance. Add 10 seconds to each side. Create a clip.

Right. Like I know that is the rule that will go viral every time, but like I wanted to stick that in one of these video platforms so you can do it automatically for me in the future. Nathan, we have to talk again. I want you to be a beta user of Submagic and Magic Lips because I think you have great ideas.

Oh guys, David, I have totally, sorry guys, I've abused David's time here. I said this ago 20 minutes, but I just fell in love with what he's doing. I mean, what an incredible story. Again, launched in, you know, call it 2023, May 2023, first customer. Launched the affiliate program shortly after that. Hit a million bucks of ARR in three months. Did a product launch. Really leaned on

virality and timing and affiliates. They have over 10,000 affiliates. They pay 30% of lifetime value and 20% of their current revenue here in June of 2025 comes from affiliates. They're doing 8 million right now in terms of ARR. They're signing up about 200, 300,000 new signups per month, converting, you know, call it 2,500 to 3,000 into paying customers. And they're doing this all guys, just two years old, bootstrapped with a team of 13. David at Submagic, thanks for taking us to the top. Thank you, Nathan. Thank you so much.