700 users, 10,000 hours: what a side project teaches that no course does

My side project has more users than most startups that raised a seed round. It cost $0 in investment. Let me tell you what happened.

The problem

In 2016, my wife Valéria was in academia — not the gym, the academic kind. Researchers and students spent hours manually transcribing audio interviews. The available tools were expensive, inaccurate, or simply didn't exist in Portuguese.

The pain was real. It wasn't a problem invented in a pitch deck.

The first version

I built the first version of Logus Academy in NextJS with Firebase. Nothing fancy:

Audio upload
Transcription using speech-to-text APIs
Editor for text correction
Export in academic formats

It took a few weeks. The design was functional, not pretty. The code was... functional, not pretty either.

Lesson #1: Your first version will be ugly. Ship it anyway. The alternative — not shipping — is infinitely worse.

The first users

The first 10 users were Valéria's friends. Researchers who needed to transcribe interviews for dissertations and theses. The feedback was immediate and brutal (in the best way):

"The transcription gets a lot of proper names wrong"
"I need to pause and rewind the audio more easily"
"The export doesn't maintain the ABNT formatting"

Each piece of feedback was a bug report disguised as a complaint. And each complaint showed me what really mattered — versus what I thought mattered.

Lesson #2: Your first 10 users are worth more than 10,000 market surveys. They tell you exactly what to build.

Organic growth

I didn't spend a single cent on marketing. Growth came from:

Academic word of mouth — researchers recommending to colleagues
WhatsApp groups for graduate programs
Basic SEO — well-structured pages for "academic audio transcription"

In 2 years, there were over 200 users. Today there are 700+ active users with over 10,000 hours of audio transcribed.

Lesson #3: If the problem is real and the solution works, the marketing takes care of itself. A good product is the best growth strategy.

The technical decisions that mattered

Firebase was the right choice

For a one-person side project, Firebase eliminated all backend complexity: auth, database, storage, hosting. I focused 100% on the product, not infrastructure.

NextJS as frontend

Server-side rendering helped with SEO (researchers found it through Google). Performance was good enough. And I already knew React.

Not scaling prematurely

With 700 users, I still run on Firebase's free/basic plan. I didn't need Kubernetes, microservices, or "enterprise architecture." The complexity I didn't add was as important as the code I wrote.

Lesson #4: Don't solve problems you don't have yet. Scale when the pain of not scaling is real.

What no course teaches

Courses teach how to write code. Side projects teach:

How to prioritize — when you're the only dev, PM, designer, and support, you learn quickly what really matters
How to deal with real users — people with expectations, frustrations, and use cases you never imagined
How to maintain motivation — there's no boss checking in, no sprint. Motivation has to come from within
How to make decisions with incomplete information — you'll never have all the data. Decide and adjust
How to measure real success — it's not GitHub stars. It's people using what you built to solve real problems

My advice

If you're a developer and have never shipped a side project, ship one. It doesn't have to be a startup. It doesn't need investors. It doesn't need to be perfect.

It needs to solve one real problem for one real person.

The rest follows.

Logus Academy is still active at logusacademy.com.br. If you're a researcher or know someone who needs to transcribe audio, it's there.