Amazon Product Opportunity Gap Scorer

Amazon Product Opportunity Gap Scorer is an ingest-first Apify Actor for imported Amazon and product-research rows. It turns rows from existing Apify Amazon scrapers, Seller Central or Product Opportunity Explorer exports, research-tool exports, spreadsheets, and manual research into deterministic opportunity scores, risk flags, missing evidence, and next sourcing actions.

It is a scoring layer, not another Amazon crawler.

What it does

• Scores each imported product row from 0 to 100.
• Classifies the main gap type: high demand with low competition, quality gap, margin gap, data gap, risk blocked, crowded, or balanced.
• Normalizes common aliases such as asin, ASIN, productId, title, productTitle, price, buyBoxPrice, reviews, ratingsCount, bsr, sellerCount, estimatedMonthlySales, profitMargin, and productUrl.
• Dedupes duplicate rows by ASIN and marketplace, while keeping the same ASIN in different marketplaces separate.
• Emits fit signals, risk flags, missing evidence, a score breakdown, and a short next action for research reviews or sourcing dashboards.

What it does not do

• It does not scrape Amazon.
• It does not use Seller Central login.
• It does not call SP-API, Product Advertising API, or paid product-research APIs.
• It does not use proxies.
• It does not create official Amazon sales, revenue, fee, or profit estimates unless those fields are supplied by your input.
• It does not guarantee profitability, product-market fit, launch success, ad performance, sourcing success, compliance, IP safety, patent clearance, or legal safety. It does not guarantee legal clearance or replace professional legal review.

Input

Use items mode with a non-empty items array. Optionally include sellerProfile to tune target marketplace, categories, keywords, margin, review count, seller count, weight, and risk keywords.

{
  "mode": "items",
  "items": [
    {
      "asin": "B0GOOD123",
      "title": "Silicone drawer organizer set",
      "marketplace": "US",
      "category": "Kitchen",
      "price": 29.99,
      "reviewCount": 142,
      "rating": 4.2,
      "bestSellerRank": 1850,
      "sellerCount": 2,
      "monthlySales": 1200,
      "marginPercent": 38,
      "url": "https://www.amazon.com/dp/B0GOOD123"
    }
  ],
  "sellerProfile": {
    "targetMarketplace": "US",
    "targetCategories": ["Kitchen", "Home"],
    "targetKeywords": ["silicone", "organizer"],
    "minMarginPercent": 25,
    "maxReviewCount": 500,
    "maxSellers": 5,
    "riskKeywords": ["patent", "hazmat", "gated", "fragile", "seasonal"]
  },
  "scrapedAt": "2026-07-08T00:00:00.000Z"
}

Keyword and category matching is intentionally conservative. The Actor matches profile terms against product titles, categories, features, descriptions, complaint text, and risk fields. It does not infer keyword fit from source URLs alone.

Output

Each dataset item is one scored product opportunity row with normalized product fields, opportunityScore, opportunityBand, priority, gapType, fitSignals, riskFlags, missingEvidence, scoreBreakdown, nextAction, researchNote, confidence, warnings, and deterministic scrapedAt metadata.

Local usage

npm install
npm test
node src/main.js --input sample-input.json --output output.json

For the full factory verification, run the repository local quality gate against this Actor path.