Skip to main content
Expand your e-commerce business into international markets with AI-generated content in multiple languages. Cernel creates native-quality product content localized for each language and market you serve.

Why Multi-Language Content?

The Opportunity: International e-commerce is growing rapidly, but language barriers prevent most stores from expanding globally. The Challenge: Manually translating thousands of product descriptions, meta tags, and attributes is expensive, slow, and requires native speakers for quality. The Solution: Cernel generates native-quality content directly in each target language—no translation needed.
Cernel doesn’t translate—it generates content natively in each language, ensuring cultural appropriateness and natural phrasing.

How Multi-Language Works

Site-Based Language Assignment

Multi-language content in Cernel is managed through Sites:
  • Each connected site has a language setting (e.g., English, French, German, Spanish)
  • When enriching, you can target specific sites or all sites at once
  • Content is generated directly in each site’s language
Example Setup:
Organization: GlobalFashion

Sites:
├── US Store (English)
├── French Store (French)
├── German Store (German)
└── Spanish Store (Spanish)
When you enrich a product, Cernel generates separate content for each site in its respective language.

Native Generation vs. Translation

How it works: The AI generates content directly in the target language using that language’s model.Advantages:
  • Natural phrasing and idioms
  • Culturally appropriate references
  • Native keyword selection for SEO
  • Accounts for regional differences (e.g., British vs. American English)
Example:
  • English: “This dress is perfect for summer weddings and garden parties”
  • French: “Cette robe est idéale pour les mariages estivaux et les réceptions en plein air”
(Not a direct translation—different phrasing, culturally appropriate)
Native generation produces significantly better quality for product content than translation, especially for marketing copy and SEO.

Setting Up Multi-Language

Step 1: Connect Multiple Sites

1

Go to Settings → Sites

Navigate to organization settings
2

Add New Site

Click “Add Site” for each language/market you want to serve
3

Configure Language

Set the language for each site (English, French, German, Spanish, Danish, etc.)
4

Connect Platform

Link each site to its corresponding Shopify store or upload separate feeds

Step 2: Sync Products

Once sites are connected, products sync:
  • Each site imports its product catalog
  • Products are matched across sites by SKU (if selling same items in multiple regions)
  • Or maintained as separate catalogs (if different products per region)

Step 3: Enrich in Multiple Languages

1

Select Products

Choose products to enrich (from any site or across all sites)
2

Open Enrichment Modal

Click “Enrich”
3

Choose Target Sites

In the modal, you’ll see all connected sites with their languages. Select which sites to enrich:
  • All sites (generates content in all languages)
  • Specific sites (e.g., only French and German)
4

Select Attributes

Choose which attributes to generate (descriptions, meta tags, etc.)
5

Start Job

Cernel creates separate enrichment jobs for each language
Each language job runs independently—if one fails, others continue. You can review and apply results per language.

Supported Languages

Cernel supports content generation in 20+ languages, including:

European

English, French, German, Spanish, Italian, Portuguese, Dutch, Danish, Swedish, Norwegian, Finnish

Asian

Japanese, Korean, Chinese (Simplified & Traditional), Thai, Vietnamese

Other

Arabic, Hebrew, Russian, Polish, Czech, Turkish

Coming Soon

Request additional languages via support
Language quality varies by AI model capabilities. English, French, German, Spanish, and Japanese have the highest quality. Reach out to support for guidance on specific languages.

Viewing Multi-Language Content

Site Selector

In the product detail view, use the site selector to switch between languages:
  1. Open any product
  2. Click the site dropdown in the header
  3. Select a site/language
  4. View product data in that language
This lets you compare content across languages and ensure consistency.

Language Badges

The product table shows badges indicating which languages have enriched content:
  • 🇺🇸 EN - English content available
  • 🇫🇷 FR - French content available
  • 🇩🇪 DE - German content available
  • 🇪🇸 ES - Spanish content available

Best Practices

Perfect your prompts and content quality in one language before scaling to others. Once you’re satisfied, enable additional languages.
Have native speakers review content quality for each language—especially for your first enrichments. They’ll catch cultural issues that might not be obvious.
French for France vs. French for Canada can differ. Spanish for Spain vs. Mexico is different. If serving multiple regions with the same language, test content with representatives from each.
SEO keywords vary by region. A product might be searched as “trainers” in UK English but “sneakers” in US English. Cernel handles this, but validate with SEO tools.
Brand names, product names, and technical specs should generally remain consistent across languages. Configure prompts to preserve these elements.
Color meanings, sizing standards, and style preferences vary by culture. Cernel’s AI accounts for this, but review carefully for culturally sensitive products.

Customizing Prompts Per Language

You can customize prompts per language/site for maximum relevance: Example: Product Description Prompt English Site:
Tone: Professional yet approachable
Length: 150-200 words
Emphasis: Features, benefits, versatility
French Site:
Tone: Elegant and sophisticated
Length: 100-150 words (French tends to be more concise)
Emphasis: Style, craftsmanship, heritage
German Site:
Tone: Precise and factual
Length: 150-200 words
Emphasis: Technical specs, quality, durability
Learn how to customize prompts →

Multi-Language SEO

Generating content in multiple languages dramatically improves international SEO:

Benefits

  • Rank in local search engines: Google.fr, Google.de, etc.
  • Local keyword optimization: Keywords natives actually search for
  • Hreflang compatibility: Content ready for proper hreflang implementation
  • Reduced bounce rate: Visitors stay longer when content is in their language

SEO Best Practices

Each language should have its own URL structure:
  • Subdomain: fr.yourstore.com
  • Subdirectory: yourstore.com/fr/
  • ccTLD: yourstore.fr
Ensure your e-commerce platform supports this.
Tell Google which language version to show to which users. Most e-commerce platforms handle this automatically when you have multi-language content.
Don’t just translate—let Cernel generate native content. This avoids duplicate content penalties and improves rankings.
Meta titles and descriptions should be optimized for local search behavior, not direct translations.
Learn more about SEO optimization →

Common Multi-Language Scenarios

Scenario 1: Global Brand, Multiple Regions

Setup:
  • One product catalog
  • Same products sold in US, UK, France, Germany, Spain
  • 5 sites connected (1 per region)
Workflow:
  1. Enrich all products for all sites simultaneously
  2. Review content per language with native speakers
  3. Apply to all sites
  4. Monitor SEO performance per region

Scenario 2: Different Catalogs Per Region

Setup:
  • Different products in different regions
  • European store has different inventory than US store
  • 2 sites connected
Workflow:
  1. Sync each site independently
  2. Enrich each site’s catalog separately
  3. Use site-specific prompts for regional preferences

Scenario 3: Gradual Expansion

Setup:
  • Start with English only
  • Expand to French, then German, then Spanish
Workflow:
  1. Perfect English content first
  2. Add one language at a time
  3. Validate quality before adding next language
  4. Reuse successful prompt configurations across languages

Pricing for Multi-Language

Multi-language enrichment costs are calculated per language:
  • Enriching 100 products x 5 attributes in 3 languages = 1,500 attribute generations
  • Each language job is billed independently
  • Token usage scales with number of languages
Start with your highest-revenue markets first to maximize ROI from multi-language investment.

Troubleshooting

Issue: English content is great, but other languages feel off.Solution: Different languages may need different prompt configurations. Work with native speakers to refine prompts for each language.
Issue: Enrichment succeeded for some languages but not others.Solution: Check job details for specific errors. Language models occasionally have temporary issues—retry usually resolves it.
Issue: Content is grammatically correct but culturally odd.Solution: Add cultural guidelines to your prompts (e.g., “Avoid American idioms for UK audience”).
Issue: Product names translating when they shouldn’t.Solution: Configure prompts to preserve specific elements (brand names, proper nouns, technical terms).

What’s Next?


Next: Learn about Bulk Operations to efficiently enrich thousands of products at once.