DEV Community

Cover image for I built a tool to learn languages through songs — now I'm testing if anyone actually wants it
Dmitry Orlov
Dmitry Orlov

Posted on

I built a tool to learn languages through songs — now I'm testing if anyone actually wants it

#sideprojects #indiehackers #buildinpublic #languagelearning

I built a tool to learn languages through songs — now I'm testing if anyone actually wants it

I've been learning Spanish through Bad Bunny and Korean through BTS for a while. Every time I wanted to understand a song, I did the same dance:

  1. Open YouTube, play the song
  2. Open Genius in another tab, find the lyrics
  3. Copy a line into Google Translate
  4. A word doesn't make sense → copy THAT word → paste again → scroll past conjugation tables → try to figure out which meaning fits this line
  5. Repeat for the next word. And the next.
  6. By the time I understood 4 lines, the song was over and I'd spent 10 minutes

It worked once. It never became a habit, because the friction was the whole experience.

So I built Langooo — and now I need to know if I'm the only one who wanted this.

What Langooo does

You pick a language, enter any song title, and it shows you:

  • Synced bilingual lyrics — original text, translation, and transliteration (for Korean, Japanese, Thai, Cyrillic, etc.) aligned line-by-line
  • Tap any word → get its meaning in context — without leaving the song, without opening a new tab
  • Save words to your vocabulary — one tap, from the songs you actually read
  • Review with flashcards — built from your saved words, not a generic deck

That's it. No fill-in-the-blank games. No daily limits. No energy bars. No paywall on the words you already found.

11 languages supported. Enter any song — if it's not in the catalog yet, the pipeline builds the page.

What already exists vs. what's a hypothesis

I want to be honest about where this stands:

What's real:

  • The product is built and deployed (vanilla JS PWA + Node backend + SQLite — no framework, no bundler, cheap to run)
  • The pipeline works: enter a song → synced bilingual lyrics → tap-to-word → save → flashcards
  • SEO infrastructure, Stripe is live
  • It works on mobile (it's a PWA)

What's NOT proven:

  • Whether anyone besides me actually wants this
  • Whether people will pay for it when Google Translate + Genius is free
  • Whether the "song-reader who learns" is a real segment or just me

Current status: ~0 users, $0 MRR. This post is part of fixing that.

How I'm testing demand

Two things, in parallel:

1. A 3-minute survey for language learners who listen to music

If you learn a language AND listen to music in it AND have ever tried to understand the lyrics — I'd love 3 minutes:

Fill out the survey

No signup, no email required. I just want to know: what's your current process? What's the hardest part? What do you wish existed?

2. Try it yourself

Go to langooo.app, pick a language, enter a song. No signup to read your first one. Tap a word. See if the friction disappears.

If it sucks — tell me why. I'd rather hear it now than after building the wrong thing for 6 months.

The technical side (for the dev.to crowd)

Some details for fellow builders:

  • Stack: vanilla JS PWA (no React, no Vue, no bundler) + native Node.js HTTP server (no Express) + SQLite (WAL mode) + Stripe
  • Why no framework: I wanted the cheapest possible hosting and the fastest possible dev loop. The whole app is ~130KB of vanilla JS. It loads fast, works offline (PWA), and costs almost nothing to run.
  • Lyrics pipeline: I'm running a self-hosted lyrics.ovh instance on the backend for lyrics fetching — it's an open-source lyrics API. That feeds into AI translation (Google Translate first, then AI re-translates in the background and overwrites the cache) → persistent translation cache. LRCLIB is the fallback for synced/timed lyrics.
  • SEO: server-rendered song pages, sitemap, structured data, IndexNow. This is the distribution bet — if the pages rank for "[song] lyrics translation," that's free traffic. Currently very early.
  • CEFR scorer: built a pure-function English difficulty scorer (A1–B2) for potential "song difficulty" pages
  • Transliteration: offline maps for Cyrillic, Thai, Devanagari, Hangul — no API dependency

What I'm trying to figure out

  1. Is "I want to understand songs in a language I'm learning" a real Job, or a nice-to-have? The survey will tell me.
  2. Will people pay for it when the free DIY exists? The survey asks this directly — and if enough people say yes, I'll open a pricing page.
  3. Is SEO a viable channel? Pages built, but barely ranking yet. Need to diagnose (KD audit + indexation check — one weekend).

If the answer to #1 is "no" or #2 is "no," I pivot.

But first — this test. If you're a language learner who listens to foreign music, I'd genuinely value your input:

And if you've built something in the language-learning space — or you've done demand-testing for a niche product — I'd love to hear how it went in the comments.

Building Langooo solo, bootstrap, in public. Follow along or reach out — I'm happy to share what I learn from this test.

Top comments (0)