Most e-commerce catalog work is brutal. You have hundreds or thousands of products with descriptions that are either copied from a supplier, machine-translated from Chinese, or just the product name repeated three times. Rewriting them manually doesn't scale. Hiring a copywriter for every SKU isn't viable.
The E-commerce Product AI Enricher solves this. You give it a list of product page URLs. It scrapes each page, extracts all available product data, and runs it through Claude AI to generate: an SEO-optimized title, a 150-word description with natural keyword integration, a structured spec table, competitive positioning, cross-sell angles, and a target customer persona.
This tutorial walks through exactly how to use it.
What You'll Build
A pipeline that takes product page URLs and returns:
- SEO title optimized for search intent (not just the product name)
- 150-word description that reads like a human wrote it
- Structured specs extracted from unstructured page content
- Competitive positioning: what makes this product worth buying
- Cross-sell and upsell angles
- Target customer persona
A 10-product run finishes in 90 seconds. A 50-product run costs under $1.00 total.
What You Need
- An Apify account — $5 free compute to start
- An Anthropic API key — Claude Haiku; fractions of a cent per product
- A list of product page URLs to enrich
The URLs can be your own store, competitor pages, or supplier pages — any publicly accessible product URL works.
Step 1: Collect Your Product URLs
For your own store, pull a CSV export from Shopify, WooCommerce, or your platform of choice. The product URL column is what you need.
For competitor research, find their product category pages and collect URLs manually or via a link scraper.
For dropshippers: go to your supplier's catalog pages (AliExpress, CJdropshipping, etc.) and collect the URLs of products you're listing. The actor will turn generic supplier descriptions into differentiated copy.
Step 2: Open the Actor
Go to E-commerce Product AI Enricher on Apify and click Try for free.
Step 3: Configure Your Run
Here's a sample input for enriching ergonomic office furniture:
{
"productUrls": [
"https://www.amazon.com/dp/B08N5WRWNW",
"https://somestore.com/products/ergochair-pro",
"https://supplier.com/products/mesh-office-chair-lumbar"
],
"anthropicApiKey": "sk-ant-api03-...",
"targetKeyword": "ergonomic office chair",
"outputFormat": "full",
"maxConcurrency": 5
}
productUrls — any publicly accessible product page. Amazon, Shopify, WooCommerce, Magento, custom sites all work.
targetKeyword — optional, but use it. Claude will optimize the title and description to rank for this term. Use your primary keyword — the one your category targets in Google.
outputFormat — full gives you everything. copy-only gives just the title and description if you only need the text.
maxConcurrency — how many pages to scrape in parallel. Default 5 is fine for most runs. Raise to 10 for large batches.
Step 4: Run It
Click Start. The actor visits each URL, extracts the product content, and sends it to Claude AI for enrichment. A 10-product run finishes in 90 seconds.
Step 5: Understand the Output
Here's what a full enrichment record looks like:
{
"url": "https://www.amazon.com/dp/B08N5WRWNW",
"originalTitle": "ErgoChair Pro Office Chair",
"enrichment": {
"seoTitle": "Ergonomic Office Chair with Lumbar Support — Adjustable Armrests, Mesh Back, 8h Comfort",
"description": "Built for people who sit 8 hours a day. The ErgoChair Pro combines a breathable mesh back with a 4-way adjustable lumbar support system that actually holds its position. The armrests adjust in height and angle independently — unusual at this price point. Suitable for desks 70–80cm high. Weight capacity 120kg. Available in black and grey.",
"specs": {
"Material": "Mesh back, foam seat",
"Max weight": "120kg",
"Adjustable height": "42–52cm",
"Armrests": "4D adjustable",
"Warranty": "2 years"
},
"competitivePositioning": "Priced below Humanscale and Herman Miller alternatives while offering comparable lumbar adjustment. Lacks the build quality of premium options but outperforms generic mesh chairs at this price range.",
"targetPersona": "Remote workers spending 6+ hours seated daily, prioritizing back support over aesthetics",
"crossSellAngles": ["monitor arms", "standing desk converters", "desk pads"]
}
}
seoTitle — leads with the primary keyword, adds specific differentiators ("Adjustable Armrests, Mesh Back") that people actually search for. Better than the original "ErgoChair Pro Office Chair" which doesn't explain what makes it worth buying.
description — specific, not generic. It mentions the 4-way armrest adjustment being unusual at this price point. That's the kind of detail that converts.
specs — structured from whatever was on the page, even if it was buried in paragraph text. Ready to paste into your product spec table.
competitivePositioning — most useful for competitive research. Run a competitor URL and this field tells you where their product is genuinely weaker.
crossSellAngles — use these for your "frequently bought with" recommendations or bundles.
Step 6: Export and Import
Export the Apify dataset as CSV → import into Shopify, WooCommerce, or wherever you manage your catalog. The fields map cleanly to standard product fields.
For large catalogs, use Apify's webhook or API integration to pipe results directly into your catalog management system as each product completes.
Use Cases
Dropshippers — supplier descriptions are almost always generic and keyword-stuffed. This actor turns "High Quality Ergonomic Mesh Chair With Lumbar Support Good For Office Use" into copy that actually explains the product. Differentiation at scale.
Amazon sellers — run your competitor listings through the enricher. The competitive positioning field will tell you where their copy is weak, what they're not saying, and how to write a better listing for the same product.
Agencies — if you manage e-commerce clients, this is a service you can sell. Catalog enrichment at scale is labor-intensive; this makes it fast. Run 100 products, edit the top 10% manually for hero products, ship the rest as-is.
Price comparison sites — if you aggregate product data from multiple retailers, the enricher normalizes it into a consistent schema. Specs, descriptions, and titles are standardized across sources.
Tips
Quality is higher when you provide targetKeyword. Claude generates more focused copy when it knows what term you're optimizing for. Don't skip it.
Run your worst listings first. Sort your catalog by conversion rate ascending and start with the bottom 10%. Those are the products where better copy will have the most immediate revenue impact.
Use the spec output for structured data. Google uses structured product data for rich results in shopping. Clean spec tables are also what customers look for on mobile — the description matters less than the bullet points.
For non-English pages: Claude AI is multilingual and can extract data from French, German, Spanish, Dutch, and other European-language pages reliably. Set targetKeyword in English for the output copy even if the source is in another language.
Pricing
| Component | Cost |
|---|---|
| Apify compute | ~$0.01–$0.02 per product URL |
| Anthropic (Claude Haiku) | ~$0.003–$0.005 per product enriched |
| 50 products | Under $1.00 total |
A 500-product catalog enrichment costs under $10. Compare that to a copywriter at $5–$15 per product description.
Try It
The actor is live: E-commerce Product AI Enricher on Apify
Start with 10 of your worst-converting product URLs and a target keyword. Review the output — you'll see immediately which fields are ready to publish and which need a quick edit.
Questions or feedback? Drop a comment below.
Top comments (0)