SEO Gaid Stet 10: Multi-language SEO

Diswan hem Stet 10 blong 13-Step SEO Gaid. Multi-language SEO saesem yu organic traffic blong mekem sua olgeta maket long blong olgeta languag — sipos yu mekem rong, hem mekem duplicate content chaos.

Evri language we yu addem hem wan multiplier long yu existing content. Wan saet we gat 50 page long wan language hem gat 50 indexable URL. Addem 5 language mo yu gat 250. Addem 20 language mo yu gat 1,000. Evri wan blong ol URL ya i save rank long lokal Google result.

Bat multi-language SEO hem wan blong ol most teknikali kompleks area blong SEO. Rong implementa i mekem duplicate content, ranking dilution, mo crawl budget waste. Difrens bitwin wan saet we i correct mo wan saet we i rong long internationalisation i save kam 10x traffi difrens.

LANGR sjelf i wok long 108 language across 89 active locale — mipla i wokem ol problem ia long bigfala scale. Diswan gaid i sharem evri samting we mipla i lanem.

What Multi-language SEO Covers

Multi-language SEO i spasen 8 critical area:

Hreflang Implementa — Telim Google which page i servem which language
Locale Routing Strategies — Subdomain vs subfolder vs TLD
Kwaliti blong Translation — Machine vs human vs hybrid translation
International Targeting — Search Console configuration
Content Lokalisation — Beyond translation: cultural adaptation
RTL Support — Right-to-left languages (Arabic, Hebrew, Farsi)
Language Detection — Servem the right version automatically
Duplicate Content — Preventing cross-language cannibalization

1. Hreflang Implementa

Hreflang tags i talem ol search engine which URL i servem which language mo region. Ol i teknikali foundation blong multi-language SEO — mo ol i most commonly misconfigured element.

Basic hreflang syntax:

<link rel="alternate" hreflang="en" href="https://example.com/page" />
<link rel="alternate" hreflang="da" href="https://example.com/da/page" />
<link rel="alternate" hreflang="de" href="https://example.com/de/page" />
<link rel="alternate" hreflang="x-default" href="https://example.com/page" />

Critical rulz:

Evri page mas referem ALL alternate versions (includim hemsef)
x-default i designatem fallback (usially English or wan language-picker)
Hreflang tags mas be reciprocal (page A i link long B, B mas link bak long A)
Yusim ISO 639-1 language codes (en, da, de) no country codes
For region-specific content: en-us, en-gb, pt-br (language-region)
Maximum ~50 hreflang entries per page (performance limit)

Thri placement options:

| Method | Best For | Drawbacks | |--------|----------|-----------| | in | Small sites (< 10 languages) | Bloats HTML, slower parsing | | HTTP headers | Non-HTML files (PDFs, APIs) | Not widely supported | | XML sitemap | Large sites (10+ languages) | Slower discovery by crawlers |

Sitemap hreflang example:

<url>
  <loc>https://example.com/page</loc>
  <xhtml:link rel="alternate" hreflang="en" href="https://example.com/page" />
  <xhtml:link rel="alternate" hreflang="da" href="https://example.com/da/page" />
  <xhtml:link rel="alternate" hreflang="de" href="https://example.com/de/page" />
  <xhtml:link rel="alternate" hreflang="x-default" href="https://example.com/page" />
</url>

Common hreflang errors:

Missing self-referencing tag (page no includem hemsef)
Non-reciprocal tags (A references B, bat B no references A)
Wrong language codes (dk insted blong da for Danish)
Pointing long non-200 URL (redirects, 404s)
Mixing x-default with wan language-specific page

Quick win: Crawl yu saet mo eksportem ol hreflang tags. Checkem for non-reciprocal relationships — diswan i most common error mo i mekem Google i ignorem olgeta hreflang setup blong yu.

2. Locale Routing Strategies

Hao yu strukturim ol URL for difren languages i affektem SEO, user experience, mo teknikali complexity. Gat thri main approaches:

Subfolder (Recommended)

example.com/en/page
example.com/da/page
example.com/de/page

Pros: Single domain authority, isi for manage, wan SSL certificate, wan analytics property, wan Search Console property, best for most sites.

Cons: Les geo-targeting signal than ccTLDs.

Subdomain

en.example.com/page
da.example.com/page
de.example.com/page

Pros: Save host long difren servers/CDNs per region, separate crawl budgets.

Cons: Evri subdomain i bildem authority separately (link equity no flow automatically), multiple Search Console properties, more complex setup.

Country-code TLD (ccTLD)

example.com (English)
example.dk (Danish)
example.de (German)

Pros: Strongest geo-targeting signal, users trust local TLDs, evri domain i independent.

Cons: Expensive (bai 20+ domains), separate authority per domain, complex link building, separate analytics/console.

Oli recommendation: Subfolder routing for 90% blong sites. Hem i concentretim evri link authority long wan single domain while i givim clear locale signals. Yusim /{locale}/page format.

https://langr.org/page        (English, default)
https://langr.org/da/page     (Danish)
https://langr.org/de/page     (German)
https://langr.org/ja/page     (Japanese)

Quick win: Sipos yu i yusim subdomains mo i trabol long authority, konsidarem migret long subfolders. Sites we i migret long subdomains go long subfolders i yuusli lukim 20-50% organic traffic increase long 3-6 months olsem link equity i consolidate.

3. Kwaliti blong Translation vs Machine Translation

Kwaliti blong translation i direkli impaktem ranking. Google i save detectem machine translation mo i save devaluem poorly translated pages. Bat long 2026 AI translation i dramatically better than the rule-of-thumb from 2020.

The translation quality spectrum:

| Level | Method | Quality | Cost | Best For | |-------|--------|---------|------|----------| | 1 | Raw GPT/DeepL output | 60-75% | ~$0 | Internal use, drafts | | 2 | AI + post-editing prompts | 75-90% | ~$0 | Blog content, non-critical pages | | 3 | AI + human review | 90-97% | $0.03-0.08/word | Product pages, key landing pages | | 4 | Professional native translator | 97-100% | $0.10-0.25/word | Legal, medical, brand-critical |

Key insight for 2026: Level 2 (AI with quality prompts) i nao suffisent for most web content. Google's duplicate content algorithms no moa penalizem well-done machine translation — ol i penalizem bad translation we no providem value.

Signs of translation that hurts SEO:

Untranslated segments mixed with translated content
UI elements (buttons, labels) still long source language
Locale-specific formatting no adapted (dates, currencies, phone numbers)
Cultural references we no translate (idioms, jokes, examples)
Same meta title/description across ol languages

Hao for validate translation quality:

Checkem se ALL visible text i translated (includim navigation, footer, forms)
Verifai locale-specific formatting (DD/MM/YYYY vs MM/DD/YYYY)
Testem CTAs — i sound natural long target language?
Checkem meta tags — title mo description mas be independently written per language
Verifai no raw i18n keys i exposed (e.g., nav.home insted blong "Home")

Quick win: Checkem yu translated pages for mixed-language content. Sipos eni buttons, labels, or UI elements i still long yu source language, fixem ol i mediatli — mixed-language pages i signalem low quality long Google.

4. International Targeting long Search Console

Google Search Console's international targeting settings i helpem Google long undastandem which pages mas rank long which countries.

For subfolder setups:

Yu no save set country targeting per subfolder (only per domain/subdomain)
Insteada, rely long hreflang + content language + user signals
Submitim wan per-locale sitemap: sitemap-en.xml, sitemap-da.xml

For ccTLDs:

.dk i automatically targeted long Denmark
.de i automatically targeted long Germany
No manual configuration needed

For generic TLDs (.com, .org, .net):

Set "International Targeting" per property long Search Console
Yusim hreflang olsem primary signal

Practical steps:

Verifyem yu saet long Search Console (wan property for subfolder setup)
Submitim yu sitemap wetem hreflang annotations
Checkem "International Targeting" report for errors
Monitorim "Coverage" report per language (yusim URL parameter filtering)
Checkem for "Duplicate without user-selected canonical" warnings — diswan i often i showem hreflang problems

Quick win: Go long Search Console > Performance > Filter by country. Checkem se ol users long Germany i landing long yu English pages insted blong German. Sipos yes, yu hreflang configuration i gat errors.

5. Content Lokalisation (No Jus Translation)

Lokalisation i go beyond word-for-word translation. Hem i adaptim content for cultural context, local search behavior, mo market-specific needs.

What for lokalise:

Currency mo pricing: Showem local currency (€ long Germany, kr long Denmark, ¥ long Japan)
Date/time formats: 25/06/2026 (EU) vs 06/25/2026 (US) vs 2026/06/25 (ISO/Japan)
Phone numbers: Local format wetem country code for international
Addresses: Match local postal format
Social proof: Local customer names, local companies, local case studies
CTAs: Adaptim tone (formal long German/Japanese, casual long English/Danish)
Images: Lokalise text in images, yusim culturally appropriate visuals
Legal: GDPR for EU, difren cookie consent requirements per country
Examples: Local brands, local websites, local references

Content we no should be directly translated:

Blog posts about local topics (raetem unique per market)
Case studies (yusim local businesses)
Pricing pages (difren pricing per market i valid)
News content (regional relevance varies)

Keyword lokalisation: No translate ol keywords — researchem ol natively. "Car insurance" long English i save kam "bilforsikring" long Danish, bat ol search volume leader i save kam "forsikring bil" (difren word order). Yusim local keyword research tools.

Quick win: Checkem yu pricing page across ol languages. I showem the correct local currency? Ol yu CTAs culturally appropriate? Wan "Get Started Free" CTA i save nidim for "Jetzt kostenlos testen" long German — no juswan literal translation, bat wat German users i expect for lukim.

6. RTL Support

Right-to-left languages (Arabic, Hebrew, Farsi/Persian, Urdu, Pashto) i require significant layout adaptation. Servem RTL content wetem wan left-to-right layout i mekem yu saet i no usable for ~500 million native speakers.

Teknical implementa:

<!-- Detect mo applyem direction -->
<html lang="ar" dir="rtl">

What mas flip long RTL:

Text alignment (right-aligned body text)
Layout direction (sidebars i move from left long right)
Navigation order (reverse)
Icons wetem directional meaning (arrows, progress bars)
Padding mo margin (swap left/right values)
Border radius (swap corner values)
CSS flexbox/grid direction

What mas NO flip:

Phone numbers mo mathematical expressions
Left-to-right brand names mo logos
Audio/video player controls
Horizontal scroll indicators
Embedded code blocks

CSS approach (modern):

/* Yusim logical properties */
.card {
  margin-inline-start: 1rem;  /* replaces margin-left */
  padding-inline-end: 0.5rem; /* replaces padding-right */
  border-start-start-radius: 8px; /* top-left long LTR, top-right long RTL */
}

Testing RTL:

Add dir="rtl" long mo checkem evri page
Verifai Arabic/Hebrew text i readable (no garbled Unicode)
Testem form inputs (text entry direction)
Checkem se numbers i display correctly within RTL text

Quick win: Sipos yu supportem Arabic or Hebrew, add dir="rtl" long yu HTML element for diswan locale mo yusim CSS logical properties (margin-inline-start insted blong margin-left). Diswan single change i fixim 80% blong RTL layout issues.

7. Language Detection mo Routing

Hao yu decide which language version blong show wan user i affektem bof long UX mo SEO.

Best practice: URL-based wetem preference cookie

First visit: Showem content based long URL (e.g., /da/page = Danish)
Root URL (/): Redirect based long Accept-Language header OR showem default (English)
Manual switch: Sipos user i selektem language, setem wan cookie mo respectem hem long future visits
Never: Auto-redirect away from wan language-specific URL

What for avoid:

IP-based redirects (Google crawls from US IPs → only indexes English)
JavaScript-only language detection (search engines can't execute JS reliably)
Redirecting /de/page to /en/page for English users (breaks hreflang)
Cloaking (showing difren content based long user-agent)

Correct redirect behavior:

User visits: /             → 302 redirect to /{detected-locale}/
User visits: /da/page      → Serve Danish content (never redirect away)
User visits: /nonexistent  → 404 (don't redirect to default language)

SEO-critical rule: Evri language URL mas be directly accessible by Googlebot without redirects. Sipos Google crawls /da/page mo i get redirected to /en/page, hem no ever i go long index yu Danish content.

Quick win: Verifai se Googlebot i save access ol yu language URLs directly. Long Search Console, yusim URL Inspection tool long wan non-English URL. Sipos hem i showem wan redirect, fixem yu routing logic.

8. Duplicate Content Across Languages

Multi-language sites i fesem wan unique duplicate content challenge: similar pages long difren languages i save compete wetim each oda long search results.

Sipos duplicate content i kam wan problem:

Pages we i 90%+ identical across languages (untranslated content)
Same URL accessible wetem mo without locale prefix (/page mo /en/page)
Missing canonical tags allowing bof versions for be indexed
Hreflang errors causing Google for pick the "wrong" version

Solutions:

| Problem | Solution | |---------|----------| | Untranslated pages | Use noindex until translated, or show English wetem clear language indicator | | Double URLs (/page + /en/page) | 301 redirect wan to the other | | Google indexing rong language | Fix hreflang reciprocity, verifai long Search Console | | Low-quality translations indexed | Improve translation quality or consolidate to fewer languages |

Canonical tag strategy:

<!-- Evri language page i own canonical -->
<!-- /en/page -->
<link rel="canonical" href="https://example.com/en/page" />

<!-- /da/page -->
<link rel="canonical" href="https://example.com/da/page" />

Never point canonical from wan language go long narafala language (e.g., Danish canonical pointing to English) — dat i talem Google for ignorem the Danish version entirely.

Quick win: Search site:yourdomain.com "your page title" long Google. Sipos yu lukim bof English mo translated versions i appearing for the same query, yu gat wan duplicate content or hreflang issue.

The Multi-language SEO Checklist

Run through diswan for evri internationalised site:

[ ] Hreflang tags long evri pages, includim self-referencing mo x-default
[ ] Ol hreflang relationships i reciprocal (checked wetem crawler)
[ ] Correct locale routing (subfolder recommended): /{locale}/page
[ ] No automatic redirects from language-specific URLs
[ ] Kwaliti blong translation validated (no mixed-language pages)
[ ] Meta title mo description unique per language (not duplicated)
[ ] Currency, dates, mo formats lokalised per market
[ ] RTL support implemented for Arabic/Hebrew/Farsi (sipos applicable)
[ ] Per-locale sitemap submitted to Search Console
[ ] Evri language page i gat hem own canonical URL (not pointing to another language)
[ ] No raw i18n keys visible long eni page
[ ] Language switcher accessible long evri pages (linked to equivalent pages, not homepage)

How LANGR Checks Multi-language SEO

LANGR i gat tu dedicated modules for multi-language SEO:

i18n-checker: Crawls up to 5 locale variants blong yu pages mo checks:

Hreflang tag completeness mo reciprocity
Inaccessible locale URLs (returning 404 or redirecting)
Hardcoded/untranslated strings across locales (fallback detection)
Raw i18n keys exposed as visible text
Translation coverage percentage

Translation scanner: AI-powered quality assessment:

Evaluates translation naturalness long a 0-100 scale
Detects machine translation artifacts
Identifies untranslated segments within otherwise translated pages
Checks UI elements (buttons, labels, navigation) separately from body content

Combined, diswan modules i check yu multi-language implementation from bof the technical (hreflang, routing, canonicals) mo quality (translation, lokalisation) perspectives — tufala blong LANGR's 13 SEO disciplines.

Common Multi-language Mistakes (Ranked by Impact)

Non-reciprocal hreflang — Google i ignorem the entire setup
Auto-redirecting based long IP — Prevents Googlebot from indexing non-English versions
Same meta tags across languages — Wastes the ranking potential blong translated pages
Mixed-language pages — Buttons long English, content long German = low quality signal
No x-default — Google no save determine yu fallback version
Translating URLs literally — /about-us → /uber-uns i fine, bat keepem consistent
Ignoring RTL — Broken layout for 500M+ native speakers
Canonical pointing to another language — Kills the translated version long Google's index

What's Next?

Step 11: B2B Lead Discovery — Turnim SEO data long qualified leads wetem automated prospecting, domain-based lead scoring, mo outreach powered by SEO findings.

Diswan gaid i part blong LANGR's 13-step SEO series. Run a free audit for lukim wea yu saet i stand across ol 13 disciplines.

SEO Gaid Stet 10: Multi-language SEO — Reachim Ol Global Audience Wetem Nogo Blong Dilutem Yu Ranking