The Mine Works
Marketplace Data Beyond Amazon: Shopee Reviews, Tokopedia, Wildberries and Wallapop
← All posts
use-case November 20, 2026 · 8 min read Updated November 20, 2026

Marketplace Data Beyond Amazon: Shopee Reviews, Tokopedia, Wildberries and Wallapop

Pull reviews from Shopee, products and sold counts from Tokopedia, prices and stock from Wildberries, and second-hand listings from Wallapop, with real inputs, Python and exact prices.

Try the scraper

The actor referenced in this article. Pay only for results delivered.

View the scraper →

If you sell in Southeast Asia, Russia or Spain, most marketplace tooling you find was built around Amazon. The marketplaces your customers actually use each have their own quirks: Shopee refuses outside requests to its per product review feed, Tokopedia rounds its sold counts, Wildberries shut its old public API hosts in September 2026, and Wallapop is local and second-hand, so price depends on the city. A tool that ignores those quirks hands you empty runs or numbers that mean something other than what you think.

Try it live: Shopee Reviews Scraper. Pay per result delivered.

This post covers four actors, one per marketplace, what each returns, a real input for each, and where the numbers need care. None of them needs an account, cookies or an API key on the site.

Which actor fits which job

JobActorMarketInputOne row is
Read what buyers complain aboutShopee Reviews ScraperSG, MY, TH, PH, VN, TW, ID, BRProduct links, id pairs or shop linksOne review
Watch prices, discounts and sold countsTokopedia ScraperIndonesiaKeywords, category, shop or product linksOne product (optional review and shop rows)
Track sale prices, stock and sellersWildberries ScraperRussiaKeywords, category, brand, seller links or product idsOne product (optional review rows)
Price used goods by cityWallapop ScraperSpain, Italy, PortugalKeywords with a city and radius, or linksOne listing

Shopee: reviews from 8 markets

Shopee now answers outside requests to its per product review feed with a refusal unless the caller runs Shopee’s own signed security script. The Shopee Reviews Scraper reads the public review feed of the product’s shop instead, the one Shopee’s shop page uses, and keeps the reviews of your product. In tests on 30 Oct 2026 it returned reviews for 16 of 16 products across all 8 markets.

Here is a complaint mining input: only 1 and 2 star reviews that have text.

{
  "startUrls": ["https://shopee.sg/product/131198971/6646421738"],
  "stars": ["1", "2"],
  "withCommentOnly": true,
  "maxReviewsPerProduct": 300
}

With a star filter only those star lists are read, so the actor reaches older complaints than a mixed read would. In a test this returned 1 and 2 star reviews with text going back 16 months.

Each review row carries review_id, rating, comment, created_at, reviewer_name as Shopee shows it, variant, image_urls, video_urls, seller_reply and Shopee’s three sub ratings: product_quality, seller_service and delivery_service. Those three tell you whether a bad review is about the product, the shop or the courier. Each product also gets a free product_summary row with the reviews found per star and a coverage value: complete when every review in the feed was read, shopee_limit when the shop holds more than Shopee serves. Shopee serves the newest 1,000 reviews of a shop in each star rating, so for a very large shop you get part of the history (325 reviews back to January 2023 for one earbud in tests).

For a weekly monitor, set newerThan to "7 days" and maxReviewsPerProduct to 0; reading stops at the first older review.

Tokopedia: prices, sold counts and stock in Indonesia

The Tokopedia Scraper reads the same GraphQL API tokopedia.com calls itself, 200 products a request, with Tokopedia’s own filters. A price watch on one category, Mall stores only:

{
  "categoryUrls": ["https://www.tokopedia.com/p/olahraga/perlengkapan-lari/sepatu-lari-wanita"],
  "minPrice": 300000,
  "maxPrice": 1500000,
  "shopType": "official",
  "sortBy": "price_low",
  "maxItems": 1000
}

Rows carry price and original_price in rupiah, discount_percent, rating, sold_count, stock, shop_name, shop_city and shop_badge (official_store, power_shop or regular). Turn on includeDetails for variants with each size and colour’s own price and stock, plus description, specifications and rating_breakdown. includeReviews and includeShops add review and shop rows.

Two things to know. Search results give a rounded sold count, so sold_count there is a lower bound; the product page gives the exact one. And Tokopedia stops at 6,000 results per search, with some repeats, so a full read of “sepatu” gave 5,469 unique products in tests. Split a big search by price range or locations.

Wildberries: prices and stock in Russia

Since 22 Sep 2026 Wildberries answers direct calls to its old public API hosts with 403. The Wildberries Scraper opens one ordinary wildberries.ru page per run, lets the site finish its browser check, and reads the product lists a shopper sees, 100 products a call.

{
  "searchQueries": ["кроссовки мужские"],
  "maxItems": 500,
  "sort": "popular",
  "includeDetails": true
}

Rows carry price (the crossed out list price), sale_price (what a buyer pays before WB wallet and bank card discounts), discount_percent, rating, feedbacks, seller, seller_rating, stock_quantity, in_stock and sizes, each size with its own price and stock flag. includeDetails adds description, specifications, composition and vendor_code at no extra cost. Keywords work in Russian or English, and a pasted category, brand or seller link keeps the site’s own sort and filters.

Read stock_quantity carefully: Wildberries caps the public figure at 44, so 44 means 44 or more.

Wallapop: second-hand listings in Spain, Italy and Portugal

The Wallapop Scraper searches es, it or pt.wallapop.com centred on a city or coordinates, with an optional radius:

{
  "searchQueries": ["bicicleta de montaña", "sofa"],
  "country": "ES",
  "location": "Valencia",
  "radiusKm": 50,
  "minPrice": 50,
  "sortBy": "newest",
  "maxItems": 100
}

In a test run this returned 200 listings in 10 seconds, all within 48.5 km of Valencia. Rows carry price, original_price when the seller cut it, city, postal_code, distance_km, shipping_available and posted_at. By default each listing’s record adds condition, brand, model, attributes, views, favorites and conversations, and each seller’s seller_rating, seller_reviews and seller_sold, all at the same price. Private sellers’ names are left out as personal data.

Python: sort Shopee complaints by cause

This runs the complaint input above and splits the bad reviews by which sub rating the buyer marked down.

pip install apify-client
from collections import Counter
from apify_client import ApifyClient

client = ApifyClient("YOUR_APIFY_TOKEN")

run = client.actor("themineworks/shopee-reviews-scraper").call(run_input={
    "startUrls": ["https://shopee.sg/product/131198971/6646421738"],
    "stars": ["1", "2"],
    "withCommentOnly": True,
    "maxReviewsPerProduct": 300,
})

causes = Counter()
examples = {}
for row in client.dataset(run["defaultDatasetId"]).iterate_items():
    if row.get("_type"):
        if row["_type"] == "product_summary":
            print("coverage:", row.get("coverage"), row.get("coverage_note", ""))
        continue
    low = [name for name in ("product_quality", "seller_service", "delivery_service")
           if (row.get(name) or 5) <= 2]
    for name in low or ["no sub rating marked down"]:
        causes[name] += 1
        examples.setdefault(name, row.get("comment", "")[:120])

for cause, n in causes.most_common():
    print(f"{n:4}  {cause}  e.g. {examples[cause]!r}")

If product_quality dominates, the product is the problem; if delivery_service does, it is the courier. The same split works on any product link from the 8 markets.

What each run costs

All four bill per row delivered, plus a small start fee per run that is charged only once rows have come back. Free is the Apify free plan; Gold applies to Gold and above.

ActorEventFree planGoldStart fee
Shopee ReviewsReview$0.00359$0.00269$0.002
TokopediaProduct from search$0.00135$0.00081$0.003
TokopediaProduct with page details$0.0036$0.00216
TokopediaReview$0.00045$0.00027
TokopediaShop$0.00135$0.00081
WildberriesProduct$0.0027$0.00162$0.003
WildberriesReview$0.0018$0.00108
WallapopListing$0.000225$0.000207$0.001

So 300 Shopee complaints cost $1.077 plus $0.002 on the free plan, and 1,000 Wallapop listings with details cost $0.225 plus $0.001. A Tokopedia product is charged as one of its two product events, never both. In all four, summary, info and error rows and duplicates within a run are never charged, and neither are Shopee reviews or Tokopedia products that your filters drop.

Limits worth knowing

  • Shopee lists reviews newest first only. There is no “most helpful” order, and likes is rarely filled. Shop name links and shp.ee short links are not read; paste a product link instead.
  • Tokopedia’s shop data can show 0 items sold for shops opened since October 2024, even when their products show sales. Product sold counts are the better measure.
  • Wildberries prices are for its default delivery region. Another city is not supported yet, and a sold out product has no price.
  • Wallapop’s brand and model exist only when the seller picked them (1,126 and 852 of 1,561 test rows). Condition was filled on 1,559 of 1,561.
  • Reviews carry names as the site shows them. Shopee masks anonymous buyers and Wildberries shows a first name only, but they can still be personal data under GDPR, LGPD, the PDPA laws or 152-FZ. Store only what you need.

For Japan and Turkey, see Mercari Japan sold price comps and Trendyol price, buy box and review tracking.

Related Actor

Explore the scraper referenced in this article: inputs, outputs, and pricing, then run it on Apify.

Frequently asked questions

Which Shopee markets does the reviews actor cover? +

Singapore, Malaysia, Thailand, the Philippines, Vietnam, Taiwan, Indonesia and Brazil. The market comes from each product link, so one run can mix them. For bare id pairs, or for shopId with itemId, set country to SG, MY, TH, PH, VN, TW, ID or BR.

Is Tokopedia's sold count exact? +

Not from search results. Tokopedia shows a rounded label there, such as 100rb+ terjual, and the actor turns it into the lower bound, 100000. Turn on includeDetails and the product page gives the exact count, charged as a product with details instead of a plain product.

Can I get Wildberries prices for a specific city? +

Not yet. Rows show the prices and stock Wildberries returns for its default delivery region. The public stock number also stops at 44, so a row with 44 in stock means 44 units or more, while lower numbers are exact.

Does the Wallapop actor return private sellers' names? +

No. A private seller's name is personal data, so it is left out. Every seller still comes with an id, private or professional, rating, review count, items sold and member since. Shops and car dealers come with the business name, phone, website and address they publish on Wallapop.

Apify Store

Find a ready-made scraper for your job

Apify has over 80,000 scrapers and automations, ours included. Start free with $5 of platform credit every month.