Bulk Processing - Managing Large Catalogs

Bulk Processing for Large Catalogs

Translating thousands of products efficiently requires the right strategy. Here's how to handle catalogs with 1,000+ SKUs.

System Limits

Plan	Max Products/Upload	Max File Size	Processing Time
Free	100	5 MB	~5 min
Starter	500	10 MB	~10 min
Growth	2,000	50 MB	~30 min
Scale	10,000	50 MB	~2 hours

Strategy 1: Prioritized Batches

Best for: 5,000+ product catalogs

Step 1: Sort by Priority

Divide your catalog into tiers:

• Tier 1 (Top 20%): Best sellers, high margin products → Translate first
• Tier 2 (Next 30%): Moderate sellers → Translate second
• Tier 3 (Bottom 50%): Slow movers → Translate last or skip

Step 2: Upload in Batches

1. Upload Tier 1 products (up to 2,000 per batch)
2. Wait for completion (30-60 minutes)
3. Review results and upload Tier 2
4. Continue until complete

Strategy 2: Language Staging

Best for: Testing multiple languages before full rollout

Phase 1: Single Language Test

1. Translate top 100 products to Hindi only
2. Upload to marketplace and monitor for 2 weeks
3. Check conversion rates and customer feedback

Phase 2: Expand to 3 Languages

1. If Hindi test succeeds, add Tamil and Telugu
2. Translate same 100 products
3. Monitor regional performance

Phase 3: Full Catalog Rollout

1. Translate entire catalog to all 12 languages
2. Upload in batches of 2,000 products
3. Track revenue impact by language

Strategy 3: Category-Based Processing

Best for: Sellers with diverse product categories

Organize by Category:

1. Category 1 (e.g., Electronics): Formal tone, all languages
2. Category 2 (e.g., Fashion): Youth tone, Hindi + Tamil only
3. Category 3 (e.g., Home & Kitchen): Bargain tone, Hindi + Bengali

This allows you to customize tone and languages per category!

Optimizing File Preparation

Before Upload Checklist:

• ✅ Remove duplicate SKUs
• ✅ Delete out-of-stock products
• ✅ Clean up formatting (remove extra spaces, line breaks)
• ✅ Ensure all required columns are present
• ✅ Add brand terms to glossary FIRST
• ✅ Split large files (>2,000 rows) into multiple files

File Naming Convention:

catalog_batch1_electronics_20260128.csv
catalog_batch2_fashion_20260128.csv
catalog_batch3_home_20260128.csv
    

Parallel Processing (Scale Plan)

Scale plan users can run up to 5 concurrent jobs:

1. Upload 5 separate files simultaneously
2. Each file processes independently
3. All complete in ~30 minutes instead of 2.5 hours

Progress Monitoring

Dashboard Indicators:

• Queue Position: Your place in line
• Processing Status: % complete (updates every 30 seconds)
• Words Translated: Real-time credit usage
• Errors: Failed products (downloadable error log)

Handling Errors in Bulk Jobs

Common Issues:

1. Malformed HTML: Unclosed tags break translation
2. Character limit exceeded: Amazon has 2,000 char limit for descriptions
3. Missing required fields: SKU or title empty

Error Recovery:

1. Download error log from job details page
2. Fix errors in original file
3. Create new file with only error rows
4. Re-upload and translate
5. Merge with successfully translated file

Cost Optimization Tips

API Automation (Scale Plan)

For 10,000+ SKU catalogs, use API automation:

# Python example
import requests

files = ["batch1.csv", "batch2.csv", "batch3.csv"]

for file in files:
    response = requests.post(
        "https://api.shabdly.online/v1/translate",
        headers={"X-API-Key": "your_api_key"},
        files={"file": open(file, "rb")},
        data={"target_languages": ["hi", "ta", "te"]}
    )
    print(f"Job {file}: {response.json()['job_id']}")
    

Need Help with Large Catalogs?

Email us at heyshabdly@gmail.com with:

• Total SKU count
• Target languages
• Monthly translation volume

We'll create a custom processing plan and recommend the best approach!