Competitor Price Intelligence

Collect product offers from Shopify stores and product pages, automatically match the same product across sellers, and report the price spread between matched offers — each group annotated with a matching confidence so you can trust (or discount) the comparison.

The Actor is deliberately conservative: it never combines different products. Two offers are only grouped when they share a hard identifier (SKU/GTIN) or their names match with confidence ≥ 0.8. Because matching caps confidence when identifiers conflict, similarly-named but distinct products stay in separate groups.

What does this Actor do?

1. Collect offers — from each Shopify store via the public products.json endpoint, and from each product/store URL via embedded schema.org (JSON-LD) Product data.
2. Normalize — every offer is reduced to name, brand, sku, seller, price, currency, availability, url, timestamp.
3. Group same-product offers — SKU/identifier grouping first, then name-similarity matching (shared.matching.match_records).
4. Compare prices — within each multi-offer group, annotate the difference versus the cheapest offer, flag the cheapest seller, and record the average pairwise matching confidence.

Input

Field	Type	Description
shopifyStores	array	Shopify store URLs (collected via products.json).
productUrls	array	Product/store page URLs (parsed for JSON-LD product data).
searchKeyword	string	Optional. Keep only offers whose name contains every keyword token.
maxProductsPerStore	integer	Max products per Shopify store (default 100).
proxyConfiguration	object	Proxy settings (Apify Proxy recommended).

Provide at least one of shopifyStores or productUrls.

Input example

{
  "shopifyStores": ["https://store-a.myshopify.com", "https://store-b.myshopify.com"],
  "searchKeyword": "desk lamp",
  "maxProductsPerStore": 100
}

Output

One record per offer. Same-product offers share a groupId; multi-offer groups carry price-comparison fields.

Field	Description
name, brand, sku, seller	Product identity and the store/source the offer came from.
price, currency, availability, url	The offer itself.
timestamp	ISO-8601 UTC collection time.
groupId, offersInGroup	Which same-product group the offer belongs to and its size.
matchingConfidence	Average pairwise match confidence within the group (null for singletons).
priceDifference, priceDifferencePct	Difference versus the cheapest offer in the group.
isCheapest	True for the cheapest offer in the group.
cheapestOffer	The cheapest offer's {seller, price, currency, url}.

Output example

{
  "name": "Aurora Desk Lamp",
  "brand": "Aurora",
  "sku": "LAMP-100",
  "seller": "https://store-b.myshopify.com",
  "price": 44.99,
  "currency": "USD",
  "groupId": 0,
  "offersInGroup": 2,
  "matchingConfidence": 0.98,
  "priceDifference": 5.0,
  "priceDifferencePct": 12.5,
  "isCheapest": false,
  "cheapestOffer": { "seller": "https://store-a.myshopify.com", "price": 39.99, "currency": "USD", "url": "https://store-a.myshopify.com/products/aurora-desk-lamp" }
}

You can download the dataset in JSON, CSV, Excel or HTML.

How matching works (and why it is trustworthy)

• Identifiers first — offers sharing a SKU/GTIN are the same product with confidence ≈ 0.98.
• Then names — offers are grouped when normalized-name similarity (plus brand/model-token agreement) reaches 0.8. Marketing filler is stripped before comparison.
• Conflicts cap confidence — when two offers carry different hard identifiers, confidence is capped low, so they are never merged. Different products are not combined.