Latka logo

Founder Interview

How Arcads Reached $6M ARR and 4,000 Paid Customers with a Team of 5 (Interview with Co-Founder Romain Torres)

Interview Date
May 16, 2025
Interviewee
Romain TorresCo-Founder
Watch
Watch the full interview

Company Metrics at Interview Time

Annual Recurring Revenue (May 2025)

$6M

Paid Customers (May 2025)

4,000

Team Size (May 2025)

5 employees

Year Founded

2024

Profitable (2025)

Yes, bootstrapped

Historical Snapshot

These numbers were reported by Romain Torres during his interview with Nathan Latka recorded on May 16, 2025, and represent a historical snapshot, not current figures. See Arcads’s current numbers.

Key Takeaways

  • 01Arcads launched in January 2024 and onboarded its first customer that month
  • 02Romain recalls roughly $5,000 in MRR within the first week after launch
  • 03Arcads crossed $1 million ARR in June 2024, roughly five months after launch
  • 04As of May 2025, Arcads was at $6M ARR, up from $5M a month earlier, with about 4,000 paid customers and a team of 5 engineers
  • 05The business is fully bootstrapped and profitable with what Romain described as very healthy profit margins
  • 06Romain built about 100 AI automation workflows in Gumloop to run growth operations
  • 07Three main acquisition channels are paid ads on Meta, content marketing, and direct sales outreach
  • 08AI agents scrape competitor Facebook ad libraries daily and automatically rewrite ads for Arcads
  • 09Most Arcads customers sell digital products such as mobile apps and lead generation services
  • 10Romain and his co-founder previously built and sold a 7-figure mobile app called WeFast before starting Arcads

Company Metrics at Time of Interview

MetricValueSource
Annual Recurring Revenue (May 2025)$6MFounder interview, May 2025
Paid Customers (May 2025)4,000Founder interview, May 2025
Team Size (May 2025)5 employeesFounder interview, May 2025
Year Founded2024Founder interview, May 2025
First Week MRR (January 2024)~$5,000Founder interview, May 2025
AI Automation Workflows Built (May 2025)~100Founder interview, May 2025
Profitable (2025)Yes, bootstrappedFounder interview, May 2025

Growth Breakdown

Revenue

Arcads launched in January 2024 and, by Romain's account, reached roughly $5,000 in monthly recurring revenue within its first week. The company crossed $1 million ARR around June 2024, and by May 2025 was at $6M ARR, up from $5M a month earlier, all without external funding.

Customers

Arcads had 4,000 paid customers as of May 2025. Romain noted that most customers sell digital products such as mobile apps and lead generation services, though the platform also serves brands selling physical products.

Team

The company operates with just 5 full-time employees, all described by Romain as very talented engineers. This lean structure allows the team to move quickly and keep costs low relative to revenue.

Profitability and Funding

Arcads is fully bootstrapped and profitable. Romain declined to share an exact margin percentage but described profit margins as very healthy. The company does not rely on venture capital to fund paid advertising or growth.

Growth Strategy

Paid Ads on Meta Using Their Own Product

Romain's primary acquisition channel is running paid ads on Meta using Arcads itself to create the ad creatives. This approach simultaneously drives customer acquisition and demonstrates the product's capabilities to prospective users.

Content Marketing and Creator Partnerships

The team publishes content on LinkedIn and Twitter, and Romain also partners with content creators who already post about marketing, advertising, and AI. This extends reach without requiring a large internal content team.

Direct Sales Outreach to High Ad Spenders

Arcads is increasingly pursuing a sales-led motion by identifying companies that spend heavily on paid ads and reaching out directly to onboard them onto higher-tier subscriptions. AI agents handle the sourcing and initial outreach at scale.

AI Agent Automation via Gumloop

Romain built about 100 automated workflows in Gumloop that handle repetitive growth tasks including sourcing influencers, drafting outreach messages, generating tweet hooks, and scraping competitor ad libraries daily to surface replication opportunities in Slack.

Competitor Intelligence Loops

One of Romain's most powerful agents scrapes the Facebook ads library for competitor ads every day, pulls transcripts, rewrites the copy for Arcads using ChatGPT, and sends the result directly to Slack with a one-click replication button, keeping the team's creative pipeline constantly refreshed.

Best Quotes

So I think we have three big acquisition channels. The first one and the most important one is, like, actually paid ads. We use our own tool to promote our tool. So we do some paid ads on Meta to, like, basically show what the tool can do, and we let's just prove it works by using it.
Yeah. For the first month, it was crazy. I think first week, it was, like, 5 k of monthly recurring revenue after, like, literally one week. So it's but something crazy like that. We had some, like, crazy moments when we had to shut down the product because we had, like, a barrel peak and stuff like that. But, yeah, very early, we understood that we had something.
So, yeah, this is kind of something we are gonna keep for ourselves right now. But what I can tell you is, like, arcads is very profitable. We are, like, bootstrapped. We are, like, not, like, relying on VC money to to to to fund the the paid ads and the growth.
No. It's a little bit more. It's, like, closer to 4,000 now.
Actually, for a lot of them, I don't even need to click start. They are for example, I have one that maybe I can show you, but it might be a bit too complex, but that basically will send me messages in Slack every single day when my competitors run new ads that are worth replicating.
AI what you have to understand is AI is very good at duplicating stuff. It's I think it's gonna be good at creating very soon, but right now, it's like really world class at taking something that is good and doing something similar.
So I think one month ago, we were at 5,000,000 ARR. Now we are at six. This the team is still five people, engineers, very talented engineers, and, let's, let's keep it, growing.
No. I think we went into the the most important ones. Like, it's basically about making good very good ads for arcads because, one, it showcase how the good the product is to people who see the ads, but, also, it teaches us how we can make a better product by using the product ourselves.

What Happened Next

This interview was recorded on May 16, 2025, and captures Arcads at $6M ARR, up from $5M a month earlier by Romain's own account, with about 4,000 paid customers and a team of five engineers, bootstrapped and profitable. Those figures are a snapshot from that date and will not be updated on this page. In a later interview with Nathan Latka, recorded in November 2025, Romain said Arcads had reached $10M ARR with a team of eight. Visit the Arcads company profile on GetLatka for the current revenue, customer, and team data.

View Arcads’s current profile and metrics

Full Transcript

Intro and Overview of Arcads

Nathan Latka

00:00Imagine doing $6,000,000 of revenue with a small team of five people. I want you to meet Romain. He's building arcads, a very competitive space that enables you to launch advertising quickly for your company. Now he launched in January 2024 and hit $5,000 of revenue extremely quickly. He'll show you his launch day formula. Next, he hit a million dollars of revenue in just four months. He'll show you his million dollar framework. And lastly, he scaled to $6,000,000

00:28of revenue with a small team. He'll show you his tiny team tactics powered by an agent army of a 100 agents. Watch until the end to get all of his data. Hey, folks. My guest today is Romain Torres. He's building a new company on top of the great new world we're living in terms of AI called arcads.ai. Helps you create winning ads with AI actors. Romain, are you ready to take us to the top?

Romain Torres

00:50>> Yeah. Let's do it.

Current Revenue and Team Size Update

Nathan Latka

00:52Alright. Hey. Listen. I don't wanna bury the lead, and you're pretty public on this stuff. But the reason I had you on is because we're trying to focus on folks that are building tiny teams with with giant revenue. Can you give us an update here in May 2025? What's the team size today, and where are you guys at in terms of revenue?

Romain Torres

01:07>> Yeah. So I think one month ago, we were at 5,000,000 ARR. Now we are at six. This the team is still five people, engineers, very talented engineers, and, let's, let's keep it, growing.

Nathan Latka

01:24That's great. So just to be clear, between April and May, we're recording we're recording this 05/16/2025. You added about a million bucks of new ARR in terms of growth and added one new team member. Is that right?

Romain Torres

01:35>> Exactly. Yeah.

Nathan Latka

01:36Alright. Guys, you're gonna I'm excited for this. You're gonna wanna learn from this guy. Romain, let's not ignore your product, though. You can't make all this work unless you have a great product. What's arcads do?

What Arcads Does and Who It Serves

Romain Torres

01:45>> Great question. So, basically, we help you create winning ads using AI actors. When you have a consumer product and you want to advertise it, one of the very common steps you have to go through is, like, basically record content with, oftentimes, UGC actors, so people who you pay to talk about your product. With our tool, you don't need to go through actual UGC actors. You can just type a script, select an actor from our library, generate

02:12>> a video, and in a couple of minutes, you get a video ready to use.

Nathan Latka

02:16Mhmm. And are you are most of the customers today, are they UGC brands selling a physical product?

Romain Torres

02:23>> Actually, not. We do have a lot of UGC brands selling, like, physical products, but I would say most of our customers are actually selling digital products. So mobile apps, even, like, lead generation. And everyone who has to, like, advertise to a very large number of people, arcads is often a good solution for them.

Nathan Latka

02:46Mhmm. And and what's driven most of your growth to date in terms of go to market motion? Is there affiliates here? Is there an SEO strategy? Is it just you you post great content on LinkedIn and Twitter? What is it?

Three Growth Channels Explained

Romain Torres

03:00>> So I think we have three big acquisition channels. The first one and the most important one is, like, actually paid ads. We use our own tool to promote our tool. So we do some paid ads on Meta to, like, basically show what the tool can do, and we let's just prove it works by using it. Right? That's number one. The second thing is content overall. So content means, as you say, like, me creating content, people of

03:32>> the team creating content on LinkedIn, on Twitter. But it's also sometimes partnering with other people who create good content. So content creators who are already posting about ads, about, like, how to do marketing, or about AI in general. So that's the second one. And the third one that we are pushing more and more is, like, more, you know, direct led sales sales led, like, approaches where we go find people who spend a lot of money in

04:08>> paid ads, and we try to, like, get in touch with them and onboard them on the tool on higher subscriptions.

Nathan Latka

04:16Mhmm. What tools are you using to go and try and find folks that are spending a lot on Meta or LinkedIn, etcetera? Is there you know, is it like an, you know, you know, an Ahrefs or similar web kind of tool or something else?

AI Agents for Sourcing and Outreach

Romain Torres

04:27>> There is no, like, secret weapon. We built a lot of internal tools. So very often when you want to either work with influencers in the space or with, like, big companies in the space,

04:44>> it all comes down to identifying them. So sourcing these people on social media and then reaching out to them. And it turns out AI is very good at this kind of task because like sourcing is a lot of, like, repetitive work. So it's a lot about scrolling social media, selecting some people based on some criteria, adding them to a Google Sheet, and then reaching out. And AI is very good at that. So we built some internal

05:09>> AI agents to to do that. And then once you have, like, sourced the right profiles, it comes down to how do I reach out to them in a natural way, have a conversation with them to convince them to start using our tool. And, again Mhmm. This AI can do in a very good way now with LLMs. So we ended up, like, building a lot of internal agents to do exactly this.

Nathan Latka

05:35I wanna dive into how you're building those internal agents. You had a great post go went out on 04/09/2025 here of all the different tools you're sort of building. So I wanna get into this in a second, know, how you're using Gumloop to automate this stuff. How do you decide what to automate? How are they working? How do you tweak them over time? Before we do that, though, I don't wanna gloss over those growth tactics you're

05:51using because they're obviously working really, really well. You've grown to $6,000,000 of revenue. And I meant to ask you too, when did you launch the tool so we can get sort of a growth rate?

Launch Story and Early Traction

Romain Torres

05:59>> Yeah. It was January 2024 when we first onboarded one customer. It was very private at the time. It was, like, just a couple of friends in the industry who were working in the mobile app industry. That's how we got the idea, by the way. We were running a a 7 figures mobile app and advertising it with, like, a lot of judicially creative, and we realized, okay. Like, this is something that's gonna be completely changed with AI,

06:28>> and we decided, okay, this is a much bigger opportunity that this mobile app studio that we are working on. So we decided to sell it and to start like building this tool. In So January 2024, we started, like, onboarding a couple of customers, and step by step, it grew, like, very quickly. In the first month, we did a lot of, like, revenue, like, straight away. So we understood we had product market fit. Now it was about

06:53>> building a better product and bidding building the growth strategy to to reach out to more customers.

Nathan Latka

06:59Do you remember how much you did in that first month? You said quote a lot of revenue.

First Week Revenue and Product Market Fit

Romain Torres

07:03>> Yeah. For the first month, it was crazy. I think first week, it was, like, 5 k of monthly recurring revenue after, like, literally one week. So it's but something crazy like that. We had some, like, crazy moments when we had to shut down the product because we had, like, a barrel peak and stuff like that. But, yeah, very early, we understood that we had something.

Nathan Latka

07:27What month did you break a million in ARR? Do you remember?

Romain Torres

07:31>> I think it was yeah. The first time was in June, I think.

Nathan Latka

07:36June 2024. That's awesome.

Romain Torres

07:38>> 2024.

Nathan Latka

07:39Yeah. We shouldn't gloss over to how you built so much empathy for this space. You were running your own 7 figure mobile app and spending a bunch on ads. I'm trying to scroll back in your Twitter feed here to see if I can find that original mobile app. Did you ever post about it back here in any of these posts?

Background: WeFast Mobile App

Romain Torres

07:52>> I didn't talk about it too much, actually, because, you know, when you have a consumer product, you don't get a lot of value by talking about it. You are targeting, like, basically, the whole world. It was a weight loss app. So our target audience, well, was not necessarily, like, tech Twitter. So all my time, I was basically spending it in the ad accounts, like, making better ads to promote our product better. And, yeah, again, that's how

08:20>> we got the idea. But now I when I I I did, like, talk with about it on on Twitter, but very, very rarely. And, yeah, that's that's that's the one.

Nathan Latka

08:29WeFast?

Romain Torres

08:30>> Yeah. Yeah. Exactly.

Nathan Latka

08:31How Roy, how does a guy how does a guy build a fasting app for women? How did you how did you get into that?

Romain Torres

08:38>> I didn't mean That's a great question. So we were like, basically, we knew we wanted to to do an app. We also knew we wanted to make something that would help people be a better version of themselves. We were with my when I say with, it's my cofounder, Dylan, and I. We're very into, like, health and wellness. And we then kind of opportunistically saw that the the people who spend the most in this category are women,

09:12>> and there is no app specifically for them. And also, is a very specific pain point when you are in the weight loss industry that which is, like, having a community of other people doing the same thing as you is very important and is very motivating. But as a woman, you don't necessarily want to post pictures of before, afters in a group where you will find, like, men and women. So we thought, okay. What if we built,

09:37>> obviously, an app, but also a community that is very engaged where people will feel safe to talk about their journey and, like, go into, like like, the challenges that they face, etcetera. So, yeah, that's why we decided to build this app, and it was very community based where there is a feed, like like, you know, kind of a Facebook group integrated into the app, and people were very, very active about it and, like, sharing about a

10:00>> lot about the journey. That's how we came up with this idea, and, yeah, again, very quickly we realized, like, one of the key factors of success to grow this app was basically being very good at paid marketing, and that's how we learned really, like, the the tactics and the like, all the the the genius it takes to to to scale an app with with paid ads.

Nathan Latka

10:22I love that. And how are you comfortable sharing how much you're spending on paid ads today to grow arcads? Like, how how much per month?

Profitability and Bootstrapped Model

Romain Torres

10:28>> So, yeah, this is kind of something we are gonna keep for ourselves right now. But what I can tell you is, like, arcads is very profitable. We are, like, bootstrapped. We are, like, not, like, relying on VC money to to to to fund the the paid ads and the growth. Mhmm.

10:48>> But, yeah, we are trying to keep everything, like, super profitable.

Nathan Latka

10:51That's awesome. Are are you comfortable sharing, like, to what degree? Are we talking, like, 10% profit margins per month or, like, 60% profit margins per month?

Romain Torres

10:59>> It's healthy profit margins. Very healthy.

Nathan Latka

11:02What'd you say?

Romain Torres

11:04>> Very healthy. They they are healthy profit margins.

Nathan Latka

11:07You're not you're not gonna let me pin you down to a percentage. Exactly.

Romain Torres

11:11>> I I know where you are getting.

Nathan Latka

11:14Yeah. You you you know my MO. You know what my audience likes. They like data, but, no, it's really impressive. Bootstrapped to $500,000 a month in revenue, 6,000,000 ARR, six FTEs. And, one of the things you do really, really well, you just touched on paid ads. We touched on content. I wanna touch on this piece of content before we jump into how you're building these these agents. What are we seeing here? Is this your workstation here?

Romain Torres

11:35>> Yeah. Actually, that's kind of a weekend project. So, basically, my thinking was simple when I did this, and, I mean, I didn't end up spending too much time on it. Like, it was just something I do for fun. My core focus is just like focusing on growing arcads. But I was thinking, okay, with arcads now, you can create a lot of these UGC ads, right, with AI, talking avatars, etcetera. What if instead of posting them on

12:02>> paid ads, which is what I really know how to do well, what if I started posting these on organic accounts? So, basically, TikTok accounts with no no money behind it. No paid budget.

Nathan Latka

12:14Are these branded accounts or

12:16are these individual influence? Like like like digital influencers No. Not not real people but characters.

Romain Torres

12:22>> Oh, yeah. They basically, they are, like, characters. Characters you create, and you create them with using arcads, and you create a face for the account using the arcads character. So I was thinking, okay. What if I can start posting 10 times a day? Actually, I very quickly realized that if you post 10 times a day on the same account, TikTok will not push your content too much because they don't really want people to be posting that

12:49>> much. So you have to find creative ways to be posting more content on multiple accounts. So I ended up, like, thinking, what if I create multiple accounts and start posting on them? And then you'd have to do these, like, creative tactics where you have to have TikTok think you are a different person, and you have to use VPNs, multiple devices. And I was just like playing with this and experimenting with this, I decided, okay, let's record

13:16>> a video to showcase what I'm like tinkering with, and let's see if I can like get some people interested in that and, like, to give me some tips on that. And I actually have a friend who is, like, very deep into this stuff, so I didn't, like, end up spending too much time on it yet. But that's something I think very interesting to push. And I see a lot of arcads customers, usually teens like 16 years

13:41>> old, and they go really deep into this strategy using arcads, and sometimes I'm really impressed by the results they are getting. Like you said, your audience likes numbers, but I see sometimes like teens create TikTok shop accounts, and they just, like, create a ton of content on TikTok using arcads avatars to promote the supplements brands and say sometimes, like, do really, really good revenue, like, sometimes, like, 6 figures a month without any paid ads budget. So,

14:10>> yeah, I think it's a very interesting, like, thing to pursue as well.

Nathan Latka

14:14Yeah. The the the the reality between what's real, what's fake, what's a real person in real life versus a digital character creator with arcads or these other tools, it's it's a really interesting world we live in. Hey, let's wrap up here the last couple minutes with something I think is your genius, which is how you have such a, you know, a tiny you're doing a million in revenue per employee. Right? This is super rare, very hard,

14:31bootstrapped, profitable. And, I think a lot of that credit goes to what you posted here on on on April 9, right, which is your paid subscribers grow. By the way, can you share that? What do got today in terms of paid customers? Are we still at about 3,000?

Customer Count Update

Romain Torres

14:44>> No. It's a little bit more. It's, like, closer to 4,000 now.

Nathan Latka

14:484,000 now. But you credit a lot of the success to, like, you built an AI growth engineer. Like, you built an AI ghost writer. You built an AI intern. I mean, Romain, can you teach us here? Are you open to logging in to your, like, your Gumloop account and just maybe showing us how you build one of these things?

Romain Torres

15:03>> Yeah. Let me try to find one of them, like, all the good example. Tosses.

Nathan Latka

15:09Take your time because I'm putting you on the spot. I'll tee my audience up, though, while you do that just so you guys can understand what I think the future is going. It's why I wanted to feature Romain is because you can see here, like, when he posted this all the way back in 02/18/2025. Right? So a couple you know, three, four months ago saying, you know, they bootstrapped from 7 figures with paid ads only. And

15:25you can see part of the way that they're doing that is they've built these agents, right, to run these profitable playbooks over and over and over again. And so the like you see here, this AI agent replaces a 100 k per year growth engineer, manages to create a 100 videos per day. And you can see he's using this building, you know, arcads content combined with Gumloop, which I'm putting on the spot here. We'll see if he

15:45can pull up an example on Gumloop and actually teach us live. Because, Romain, I think what a lot of people think about is, you know, they're not you strike me as someone you're pretty darn technical. Right? You know how to write code. You you and you're also a great growth engineer, which is a rare combination. But for people that are just great marketing and sales folks, how could they use Gumloop to run some of these AI

16:01agent playbooks?

Introduction to Gumloop and AI Workflows

Romain Torres

16:03>> Yeah. I will try to find a simple one to Okay. To give you a good intro about what you can do, and then I will try to show a little bit of a more complex one. But by the way, Gumloop, the team is very good at teaching you how to use it. They have this AI agent inside Gumloop that you can chat to and that you can, like, use to kind of understand better how to build

16:30>> them. And even the team, like, you can reach out to them and just, like, ask them to to guide you on how to do more complex workflows. So let me share my screen. Where is the yeah. Here it Riverside.

Nathan Latka

16:45It should be able share yeah.

Romain Torres

16:46>> Yeah. Perfect. I will try with a very simple one. So basically, Gumloop, what it does is it lets you connect together multiple nodes. A node is a block right here. So there is a Google Docs reader node, there is an input node, combined test, multiple nodes like this, and they will let you create the workflow that will take some input that you give it and transform it using AI. So let's jump into this example. So right

17:17>> here, I created this Google Doc, a super simple Google Doc with some examples of very good tweets. So Can you show us? This one is yeah. Let me share this one right here.

Nathan Latka

17:33This is while he brings it up, guys, this is what I think a lot of people forget to do. You have to be a great you have to be a historian almost. Like, you have to document history, what does well, because that's how you're gonna train anything like a Gumloop. But Romain, I think, is obviously very good at that. He describes it as, you know, simple Google Doc, but you can see here walk us through Romain.

17:47What have you documented?

How AI Agents Are Built and Trained

Romain Torres

17:49>> Yeah. 100%. AI what you have to understand is AI is very good at duplicating stuff. It's I think it's gonna be good at creating very soon, but right now, it's like really world class at taking something that is good and doing something similar. So, yeah, I made this list of very good Twitter posts and actually this one was from the founder of Gumloop. And, yeah, I I saw all of these Twitter posts that I think were

18:16>> very good. So that's there are, you know, 10 of them, and I added them to the Google Doc. And then I let's go back to Gumloop. I told Gumloop, okay. This is my input, and the input is what I want to talk about. So this time, I want I wanted to talk about a new feature for arcads where you can control actors'emotion. In the post, we'll show sad and angry. And so those are the two

18:43>> inputs that I gave to the Gumloop flow, the thing I want to talk about, and the list of inspiration posts from the Google Doc. And then Gumloop will combine them together. So it will, like, combine these texts with the content of the Google Doc that it will read, and then it will send that to a prompt and ask chat GPT, okay. I want you to ingest all these articles from inspiration list, and then carefully choose one

19:13>> article in the inspiration list, blah blah blah. And you basically need to do a similar post inspired from one of these articles. And I say something like, you'll be penalized if you don't select the one that is the most relevant.

19:31>> And then it will, create a new Google Doc, and it will if I run it, we can see if it give us gives us a good result. But, basically, it will write in a new Google Doc something that is inspired from the top post in this list, the most relevant one to this post I want to do right here, and it will, like, then put that into a new Google Doc. Does it make sense?

Nathan Latka

19:55How many of the yeah. Does how many that was amazing. How many of these do you have built just on your Gumloop account? Are we talking, like, five or 500?

Romain Torres

20:02>> No. More than more like, 100. Maybe 100.

Nathan Latka

20:07Wow. So how much of your day is, like, just coming in here and, like, clicking start on all these tasks? I imagine you just have, like, a bunch of computers up where it's, start start start. You get your breakfast. You hit start start start start. You come back an hour later, you, like you you just did, like, thirty days worth of work.

Competitor Ad Scraping Agent Explained

Romain Torres

20:20>> Actually, for a lot of them, I don't even need to click start. They are for example, I have one that maybe I can show you, but it might be a bit too complex, but that basically will send me messages in Slack every single day when my competitors run new ads that are worth replicating. So it will scrape the ads libraries, and, yeah, it will download the ads from Facebook ads of my competitors, and it will, like,

20:47>> take the transcript of these ads, send it to ChatGPT, and then ChatGPT will rewrite the same ads but for arcads, and it will send me that in Slack with a message like, this ad was running on your competitor's account, and now you can replicate it and click this button to replicate it.

Nathan Latka

21:05Wild. Can we go to the output here?

Romain Torres

21:07>> Yeah. Let's have a look at this one. So, yeah, inspired by

21:11>> post seven and send it through it through this. This new arcads feature to control emotion is insane. You pick an and, yeah, and I think I wake up every day with five fresh this one is, like, a very good one, emotional knowledge for my ads.

Nathan Latka

21:26Can I ask you a question?

Romain Torres

21:28>> Basically.

Nathan Latka

21:29What what a lot of, like, cop purely trained copywriters, they know the first thing is the hook, then it's, like, the detail, the curiosity, then it's, like, the call to action at the end. You didn't feed in your input Google Doc a breakdown of, like, here are 30 really good hooks. Here are 30 really good, like, middle parts, and here are 30 really good call to actions. You just gave it, like, the whole chunk of post.

21:51Why'd you decide to train your sort of mini Gumloop model that way?

Romain Torres

21:55>> Yeah. It's actually a very good idea, and I might have to try this as well. My understanding of, like, where ChatGPT really excels is not when you give them, like, tips on, like, this is how you write a good hook, and this is how you write a good body. But it's more like, do this, but in this different way. And that's why, like, I like to give it just one example as an input and ask it

22:22>> to replicate it, because I found it to be better just in insults

22:29>> in certain ways. But, yeah, definitely a good experiment to try. I actually have another one, a hook writer specifically. So, actually, to your point, let me share this one instead. I have a Twitter of hook writers.

Nathan Latka

22:43So This is this is the hook. This is the so you like, when you're scrolling Twitter in your free time or whatever, you're, like, bookmarking things. Like, that's good hook. That's good hook. And then maybe once per week or month, you take all those good hooks and feed it into these systems so they become evergreen, and then they that's basically you building your brain out.

Hook Writer and Content Automation

Romain Torres

22:57>> Exactly. That's exactly what I do. So I them to Notion. So I don't save them from Twitter, but it's it's the same. I copy paste them into a Notion document, a Notion database where I have all the hooks that I find good. And this one, what it does basically, we won't go into the details, but it basically generates hook ideas. So instead of just making one hook, it will take

23:22>> this Notion database, and it will generate new hooks inspired from this database for the idea I want to write about. So I input my idea here, and actually my team members also use it. And I tell how many hooks variations I want, and it will generate those for me.

Nathan Latka

23:41It's wild. Amazing. Romain, hey. I wanna be respectful of your time. You got a company to build, and you're growing quickly. I'm sure you're hiring. You're trying to optimize paid ads, all that jazz. Is there anything I should have asked you about, like, that you know is critical to your growth so far that I just you're going, man, Nathan just didn't ask about this. It's the key thing.

Closing Thoughts on Growth Formula

Romain Torres

23:57>> No. I think we went into the the most important ones. Like, it's basically about making good very good ads for arcads because, one, it showcase how the good the product is to people who see the ads, but, also, it teaches us how we can make a better product by using the product ourselves. And, yeah, it's all about, like, making good systems to come up with new ideas for ads and then making a good use of tools

24:23>> like arcads and others to generate them quickly and make very good ads quickly. So that's that's the whole secret. And if you do that, like, very repeatedly for weeks and weeks and weeks, your ads performance tends to improve. And at the end, you tend to end up having very good results.

Nathan Latka

24:38Guys, Romain Torres. 2021, 2022, 2023 is building WeFast apps that, oh my gosh, we're spending so much time on paid ads and creatives. I gotta go build a tool for this. He launched arcads in January 2024. One customer, but went to, like, five k of new MRR in the first seven days, broke a million bucks of ARR in June 2024. And now today here in May 2025, six full time employees, 6,000,000 ARR, and well over

25:0150, well over a 100 sort of agents built in Gumloop that help him operate his growth channels at a very high velocity. Paid ads on Meta content, influencers now exploring with direct sales. Romain, thank you for taking us to the top.

Romain Torres

25:14>> Yeah. Thank you so much. It was great.