I've earned $14,440 over the last month and shipped 56 micro projects over the last year using the same tech stack. This approach has helped me double my profit, reduce software development time and expenditures, increase my clients' chances of success by 3 times, reduce the cost of testing ideas from $50k to $2k, and start launching my own micro startups while spending just a few hours per week. For those who do not know me, I'm Dennis Babych, and I'm Ukrainian.
I'm the director of a software development agency and the owner of a few micro startups. I've been in this field for more than 10 years now. Before we start, I want to say that I will give you a perspective not from a developer, but from an entrepreneur and your potential client.
Let's begin. 10 years ago, like many developers, I loved switching tech stacks. Wow, Angular came out, then React, what's new for mobile?
React Native is cool, and now let's switch everything to Flutter because it's popular. And now there's Vue. js—maybe we should move all our projects to it.
My team and I were jumping from one language and framework to another: microservices, monolith, GraphQL, Docker, Kubernetes, Load balancer, etc. We followed market trends. Yes, we gained a lot of experience.
Let me explain the main point. Over the last three years, the commercial market has changed significantly. It used to be easy to find clients with a $200k budget for development, and you could use any frameworks and technologies and learn at the client's expense.
But now the market has changed dramatically. People who want their startup used to be able to spend from $100k to $200k, but now their budgets have dropped to $20k-$50k. So you need to change something to satisfy the market.
That's why I started this YouTube channel, to get more requests from potential clients. Over the past year, I received 379 requests, 10% of which turned into contracts. But, as I mentioned earlier, the market went down, and clients from YouTube also couldn't spend large budgets on development.
So, I had to change the development concept and sales logic. I had to use the so-called CDO approach to sales. I hope you've seen the movie The Big Short.
There's a scene where they explain what a CDO is using the example of repackaging stale fish in a restaurant into a soup, making it a new, fresh dish. I'll leave a link to that wonderful moment in the description. To sell software development services, you need to repackage them and start selling them differently.
My friend, for example, packaged his software development services as CTO-as-a-Service with a subscription, while I decided to continue helping people quickly launch startups and started selling a micro startup boilerplate for $2,400. This included all the necessary blocks and modules for a quick start. The truth is, to stay competitive in the market, you need to do three things: Optimization, Reduction, Persuasion.
First, you need to optimize all your development and speed it up. How? You need to create ready-made blocks and modules that clients need.
Identify the most common requests from clients, meaning which features they want and how to deliver these working features with limited functionality. In other words, we optimize the code we deliver to our clients. After that, we reduce functionality to ensure delivery takes the minimum amount of time.
For example, deployment to AWS or Vercel? The choice is obvious here. Remember, we're talking about an MVP.
And then there's persuasion and sales methodology. Because if you can't convince the client that they need to build an MVP instead of a full product for their own benefit, you won't be able to deliver and stay competitive in the market. Additionally, this helped my clients spend no more than $2,000 on creating and testing their startups.
Whereas before, they had to spend $50k and wait six months, now they can roll out an MVP in a couple of weeks, test it, and scale it or create a new micro startup for $1k because the code is already ready. By the way, this optimization also helped me launch my own micro SaaS startups quickly, spending just a few hours on delivery and launch. So, let's break down what we chose which tech stack crystallized, and why.
By the way, if you want to learn more about launching micro startups, like this video and subscribe to the channel. It’s just two clicks for you, but it brings me a ton of joy and motivation to share more insights with you. If you want to code something quickly, then React is the way to go.
The main advantage is the vast number of ready-made UI components, as well as modules like calendars, phone inputs, validation blocks, etc. It’s worth mentioning state managers like Redux and Zustand, which help you create simple applications quickly. So, for quickly creating micro apps, the best framework that my team and I use is Next.
js. The biggest advantage is the ability to write a full-stack app in one place, thanks to API routes and the ability to work with a database that won't get to the client side (also this can be a drawback as it makes control more difficult). Another big bonus of Next.
js is its super easy and super simple deployment to Vercel. However, deploying Next. js on anything else (AWS, Heroku, Google cloud) is almost impossible because Vercel are monopolist and they makes their framework hugely dependent on their ecosystem.
I think over time there will be some Vercel-only features. Along with Next. js, we use Shadcn UI, NextAuth, and other cool tools to speed up development.
Additionally, a huge bonus is the large pool of developers who can continue your project if something happens to your developer, plus the low entry barrier. This optimization alone helped me reduce costs by 50%. This applies to basic micro apps, but if you’ve outgrown the basics, then Vue.
js & Nuxt are what we use when we have quickly validated a client’s idea and they need to scale the project quickly. By the way, if you want to debate or suggest your solution, join my Telegram group and my Twitter right now, where I post tons of tech insights and do live chats with my community. With the backend, things are not simple either.
My team and I had to choose between Python, Node. js, and TypeScript. We started coding the backend for micro apps in TypeScript.
Why? Because it keeps you within the same stack (JS, TS), which makes it easier for any developer to create tiny apps. TypeScript allows us to do a super-fast delivery for our clients.
And I can see the happiness in their eyes. By the way, I should add that TS is not a replacement for Python. TS is not even a language—yes, you can throw stones at me.
It’s just an attempt to add typing and fix the childish problems of JS. Did it succeed? No.
But the hype was, is, and will be there. There are no specific services. This means you can write integrations with whatever you want, however you want.
However, if you plan to work with AI and not just a simple OpenAI wrapper but on medium and large projects, which, by the way, are quite common now, then after validating the idea, you need to switch to Python. In conclusion, for small projects, use TS; for something bigger, use Python. Databases.
MongoDB just doesn't scale, even though we're promised it does. Nowadays, every database can do what MongoDB does (store unstructured data). So, we chose PostgreSQL + Supabase.
Supabase is a really cool comprehensive solution that offers auth, S3-like file storage, and lambda functions. Now they’ve added vector storage—hello, custom AI. A huge bonus is that it's open-source and can be self-hosted, and you only pay for the hardware.
Deployment. Currently, I use two options. The first is Vercel for my clients, where I need to show quick results or for my side projects.
It’s fast, and convenient, but very expensive if you have more than 3 users. However, if I expect high traffic, I use Hetzner + Coolify. Yes, you'll have to set everything up once and learn how to do it, but it’s part of the experience that will always come in handy.
And the price of 5 euros for 2 CPUs, 4GB RAM, and 40GB SSD is the best on the market right now. By the way, one of my micro SaaS projects that I launched in 2 days is Design Fast 2. Many people started asking me how I created such cool animations, so I decided to share it with them.
Figma slides, animations, and 3D objects in the background that you see on the screen—you can purchase them. The link is in the video description. I also frequently use services like Resend for sending emails.
For payments, I almost always use Stripe. Even if a client has a local provider or Stripe doesn't support their country, it's always easier for them to create a company in a Stripe-supported country than to use custom solutions. Why?
Because Stripe is not perfect, but it’s fast. As I mentioned earlier, the key task is to ship fast. Clients often want an admin zone or dashboard, and for that, I use Strapi CMS.
It’s very easy to deploy and gives the client access to the necessary functions. Of course, when we talk about side projects or micro SaaS, it’s very important to measure the results of our work. For this, I use Mixpanel and PostHog.
These services allow you to monitor how many users visit your site and what they do there. This way, you can remove or add blocks that will be useful to your users. By the way, we are currently working on many AI projects.
I'm not just talking about the usual OpenAI API but about AI cold phone calls, LangChain knowledge bases for companies, employee training, and so on. For all of this, we need a fast wrapper, and in 99% of cases, I use my Micro SaaS Fast boilerplate. Clients need basic functions such as payments, subscriptions, Stripe integration, email notifications, landing pages, admin dashboards, SEO-friendly blogs, and fast deployment.
With Micro SaaS Fast boilerplate, I can deploy the minimal infrastructure in a couple of hours, making the client happy and allowing us to focus on developing their required AI solutions. The biggest bonus is the trust that builds between you and the client, which allows you to take the necessary time to create the AI or any other idea the client needs. The link to the boilerplate is in the video description.
Using various boilerplates over the last year, my team and I created 56 different micro startups using different tech stacks. But the most important factor was doubling our profit by changing our approach to delivery. To stay competitive in the market, you need to change yourself, change your delivery approaches, and create something new.
Start creating your micro SaaS today so that in a year, you'll have several launched side projects bringing you passive income on autopilot. The market needs your ideas. And I do believe in you and you can succeed, it's just a number of tries and the micro SaaS fast boilerplate is allowing you to ship code fast without busting the budget.
Just start today! See you!
