Turkish E-Commerce Review Aggregator - Trendyol, Hepsiburada, N11 Reviews

Extract product reviews from the biggest Turkish marketplaces into one normalized dataset with built-in sentiment tags.

Pricing: $3 per 1,000 reviews.

Why teams use it

• Monitor customer feedback across Trendyol, Hepsiburada, and N11 in one pipeline.
• Run basic Turkish sentiment analysis without building a custom tagging layer first.
• Track competitor complaints, product quality issues, and praise trends over time.
• Stream review-level records directly into dashboards, BI tools, or AI workflows.

Works great with...

• ../n11-product-scraper/README.md for product discovery and catalog context.
• ../seller-intelligence/README.md for correlating review quality with seller trust signals.

Input example

{
  "productUrls": [
    "https://www.trendyol.com/spigen/ciel-by-cyrill-iphone-15-pro-kilif-cecile-flower-garden-acs06760-p-758714142",
    "https://www.hepsiburada.com/spigen-20w-usb-c-mini-hizli-sarj-aleti-sarj-isisini-dusurur-gan-destekli-akim-korumali-guc-adaptoru-iphone-android-ipad-type-c-white-ach02071-p-HBCV000008SWTT",
    "https://www.n11.com/urun/logitech-mk270-kablosuz-usb-turkce-q-klavye-mouse-seti-61465"
  ],
  "searchQuery": "kablosuz klavye",
  "platforms": ["trendyol", "hepsiburada", "n11"],
  "maxReviewsPerProduct": 100,
  "minRating": null,
  "sortBy": "recent",
  "proxyConfig": {
    "useApifyProxy": true,
    "apifyProxyGroups": ["RESIDENTIAL"],
    "apifyProxyCountry": "TR"
  }
}

At least one of productUrls or searchQuery must be provided.

Output Schema

Each review is emitted as its own dataset item. At the end of the run, the actor also emits a RUN_SUMMARY record.

{
  "scrapedAt": "2026-04-04T11:02:15.512Z",
  "platform": "n11",
  "sourceUrl": "https://www.n11.com/urun/logitech-mk270-kablosuz-usb-turkce-q-klavye-mouse-seti-61465",
  "dataVersion": "product-review/v1",
  "productId": "61465",
  "productTitle": "Logitech MK270 Kablosuz USB Turkce Q Klavye Mouse Seti",
  "productUrl": "https://www.n11.com/urun/logitech-mk270-kablosuz-usb-turkce-q-klavye-mouse-seti-61465",
  "reviewId": "n11-61465-1",
  "reviewerName": "A*** K***",
  "rating": 5,
  "title": "Bekledigimden iyi",
  "body": "Kaliteli paketleme ve hizli teslimat. Tavsiye ederim.",
  "reviewDate": "2026-03-29T12:00:00.000Z",
  "isVerifiedPurchase": true,
  "helpfulCount": 3,
  "reviewImages": [],
  "sentimentTag": "positive",
  "sellerName": "GTI-Bilisim",
  "variantInfo": null
}

Notes

• Ratings are normalized to a 1-5 scale across all supported marketplaces.
• Review records are pushed as they are collected, so you get streaming output instead of one large batch at the end.
• Progress logs include sentiment totals such as Reviews collected: 230 (78 positive, 45 negative, 107 neutral).
• The dataset ends with a RUN_SUMMARY record that captures totals, duration, success rate, and platform breakdown.

FAQ

Is each dataset item a product or a review?

Each dataset item is a single review. productUrl and productTitle are included on every row for context.

Can I search for products instead of passing URLs?

Yes. Use searchQuery and the actor will discover top products on the selected platforms before scraping reviews.

How is sentiment assigned?

The actor uses Turkish positive and negative keyword matching first, then falls back to rating-based tagging when the text is ambiguous.

What happens if one product page or review endpoint fails?

The actor retries when appropriate, logs the failure, and continues processing the rest of the run.

Changelog

v1.0.0

• Added stronger input validation for product URLs, platform selection, and review limits.
• Added streaming progress reporting, partial completion handling, and final RUN_SUMMARY records.
• Added publication-ready README, actor metadata, and smoke-test checklist script.