I built a SaaS landing page with Lovable last month. Shipped it in a weekend, felt great about it.
Then I checked Google Search Console.
Zero impressions. Zero clicks. For two weeks.
Turns out, building fast with AI tools creates a problem nobody talks about: your app only exists in one language, and Google doesn't know what to do with it.
The Problem Nobody Warned Me About
When you ship a React app with Lovable, Bolt, or v0, you get:
- A working product ✅
- A deployed URL ✅
- SEO-ready pages ❌
Here's what I mean. Google crawls your site, sees English content, and indexes it for English queries. But 75% of internet users prefer content in their native language. You're invisible to most of the world.
I wanted to add French and German to reach European users. Simple, right?
What I Tried First (And Why It Failed)
Attempt 1: react-i18next
The "proper" way to do React internationalization. Here's what that looks like:
// Create translation files for EVERY string
// /locales/en/translation.json
{
"hero.title": "Build faster with AI",
"hero.subtitle": "Ship your product in days, not months",
"cta.button": "Get Started",
// ... 200 more keys
}
// /locales/fr/translation.json
{
"hero.title": "Construisez plus vite avec l'IA",
"hero.subtitle": "Livrez votre produit en jours, pas en mois",
"cta.button": "Commencer",
// ... 200 more keys (pray you didn't miss any)
}
Then wrap every single text element:
import { useTranslation } from 'react-i18next';
function Hero() {
const { t } = useTranslation();
return (
<h1>{t('hero.title')}</h1>
<p>{t('hero.subtitle')}</p>
<button>{t('cta.button')}</button>
);
}
The reality with Lovable: Every time I prompted a new feature, I had to pray the LLM would update the JSON files correctly. Half the time it didn't. I spent more time fixing translation keys than building features.
Attempt 2: Weglot
Weglot is the industry standard. Paste a script, get translations.
Then I saw the pricing: €490/year for my small landing page. That's more than my hosting, domain, and analytics combined.
For a side project testing product-market fit? Hard pass.
Attempt 3: Google Translate widget
Don't. Just don't. It looks terrible, breaks your layout, and Google specifically says machine-translated content via widgets doesn't help SEO.
What Actually Worked
I found lovalingo — built specifically for Lovable/Bolt/v0 apps.
The setup was wild:
- Paste your site URL
- Pick your languages
- Copy one prompt into Lovable
- Done
No JSON files. No wrapping every <p> tag. No €490 invoice.
Here's what convinced me it wasn't just another translation widget:
Zero-flash rendering — The translated content loads server-side. No awkward flash of English before French appears.
Automatic SEO — It generates proper hreflang tags and creates a sitemap with all language versions. This is the part that actually gets you indexed.
<!-- What lovalingo adds automatically -->
<link rel="alternate" hreflang="en" href="https://myapp.com/"/>
<link rel="alternate" hreflang="fr" href="https://myapp.com/fr/"/>
<link rel="alternate" hreflang="de" href="https://myapp.com/de/"/>
<link rel="alternate" hreflang="x-default" href="https://myapp.com/"/>
One prompt install — This is the entire setup for Lovable:
Install lovalingo translation with key [YOUR_KEY]
for languages: French, German
The LLM handles the rest. No manual file management.
The Results (30 Days Later)
| Metric | Before | After |
|---|---|---|
| Languages | 1 | 3 |
| Indexed pages | 12 | 36 |
| Impressions (non-EN) | 0 | 847 |
| Setup time | - | 5 min |
The most surprising part: I started ranking for German keywords I never targeted. Turns out there's way less competition for "KI Landing Page erstellen" than "AI landing page builder".
When This Makes Sense (And When It Doesn't)
Use this approach if:
- You're building with Lovable, Bolt, v0, or similar AI tools
- You want multilingual SEO without managing translation files
- Your budget is "indie hacker" not "enterprise"
Stick with i18next if:
- You need pixel-perfect control over every translation
- You have dedicated translators and a localization pipeline
- You're building a complex app with dynamic content that changes based on user data
The Bigger Lesson
AI coding tools optimized for speed, not global reach. That's fine — speed matters when you're validating ideas.
But if you're past validation and want actual users, remember: the world is bigger than English-speaking Product Hunt.
Adding translations isn't just about being nice to international users. It's about 3x-ing your addressable market with minimal effort.
Have you dealt with i18n in your Lovable/Bolt projects? What worked for you? Drop a comment. Peace.
Top comments (0)