After seeing the poor quality of a successful mobile app, David Attias decided he could do better. He built an app a year ago, and now he's building another. This small portfolio is already bringing in $10k/mo.
Here's David on how he did it. 👇
I graduated from a computer science school in Paris 13 years ago. I started working for Criteo as a technical account manager in London, then Barcelona, then NYC. After that, I worked as a freelancer for multiple banks and professional traders, designing their trading algorithms for seven years. That's where I made the most money.
In 2025, I watched a podcast featuring three teenagers who created an app called QUITTR, which reached $200K per month in 3 months. I tested the app myself, paid for it, and I was amazed by its poor quality. I concluded that if they could do it, I could too.
So, I made STOPPR, inspired by the QUITTR onboarding, to help people stop their processed sugar cravings. I reached $5K in revenue in two weeks, thanks to two viral videos from my influencers.
Three months later, I reached $14k in monthly revenue — I struggled to exceed that amount, and managing influencers consumed three-fourths of my day. Despite a 20% profit margin, I was overworking for too little money.
Now, I'm a developer relations advocate at Adapty for the French market, building their community in French-speaking countries. It's a very cool job. In parallel, I'm working on scaling another app in the USA.
I'm currently bringing in $15K/mo. Roughly 70% of that comes from my apps. The rest is from my job.
When I started my first app a year ago, I used Cursor + Figma + Claude 3.5 + Firebase.
AI was really bad at design then, so I asked my designer friend to create the Figma screens for me.
After that, I used the Figma MCP within Cursor to import the screens and asked Cursor to vibe code both the front end and backend for each screen, including animations and navigation between screens.
Even a year ago, it was 70-80% perfect. I still had to test each screen, button, navigation, and flow. But whenever I spotted an error, I just asked Cursor to fix it. The barrier to design and code was incredibly low then. It's even lower now.
Today, my stack is GPT 5.5 CLI (Command Line) extension within Cursor + Firebase. GPT 5.5 is goated for mobile app dev. Better than Claude 4.7
Also, it's worth mentioning that with the previous stack, I always spent $1k a month. With the current stack, I spend no more than $200 a month.
I've grown through influencer marketing. Initially, I'd pay 20% up front for eight videos per month — four on Instagram and four on TikTok. They had to hit a cumulative minimum views based on the averages of their last 20 videos. If they didn't hit that number with eight videos, they'd have to keep posting.
That approach generated views for significantly less than $1 CPM.
However, influencers often work with competing brands in parallel. They lacked incentive. I wanted influencers who actually wanted to drive conversions.
Now, I find one or two big influencers in the niche and offer them equity instead. I do this for both my apps. Unlimited upside.
We keep a close eye on our videos via viral.app. It connects to the influencer's socials and tracks in real time. Whenever a video starts to go viral, we promote it with TikTok and Meta paid ads as quickly as possible. That helps us ride the momentum.
I found all my influencers using the For You feed on TikTok and Instagram. I don't use any creator marketplace platforms. They're overpriced.
The main challenge is finding good influencers. Most influencers actually lack influence. They don't know how to go viral on repeat — that's a learnable skill.
The second challenge is convincing them to accept equity in the company rather than a fixed rate. This requires extensive negotiation and convincing. But I only need one or two per app.
Here's the roadmap to getting beyond $10K a month, starting the first month after you ship your app.
Don't work with UGCs or ambassadors. They don't know how to go viral. Micromanaging them takes too much time.
Find one or two big influencers in your niche.
DM them. Email them. Join their Discord or Telegram.
Get them on a call by any means necessary.
Leave a Comment
tarting from zero to building a $10k/month app portfolio in a year may sound challenging, but with the right strategy, consistency, and smart execution, it becomes an achievable goal. Focus on identifying simple app ideas that solve real problems, validate demand quickly, and launch fast. Instead of relying on one big success, create multiple small apps that generate recurring income through subscriptions, ads, or one-time purchases.
Success comes from learning, testing, and improving over time. Begin with basic tools, understand market trends, and prioritize user feedback to refine your products. By consistently building, optimizing, and scaling profitable apps, you can gradually grow a strong portfolio that creates steady monthly revenue and long-term opportunities.
can't open the app stoppr which I download from google play store....
What made the biggest impact?
hi
Totally agree. People often think bigger means more contacts, but sometimes 1-2 truly impactful relationships can drive much faster growth.
This is so usefull. Good marketing technique
This is really interesting to read as someone coming from a creative background rather than tech. I’m an artist and recently realised a workflow frustration I’ve had for years inside creative apps could potentially be turned into a real product. It’s fascinating seeing how creators and developers collaborate to build things from genuine lived experience.
Would genuinely love to hear whether this resonates with anyone else working in creative software.
Great breakdown — the equity for influencer angle makes total sense when recurring subscription revenue backs it up. Wish this model translated as cleanly when you're selling one time products or assets (the unlimited upside is harder pitch to an influencer).
The viral amplification via. paid ads the moment something starts gaining traction is a great takeaway regardless of the model though.