Build 56 micro SaaS, Earn $14,400/m, with SIMPLE Tech Stack

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!