Daiso Korea Ranking Scraper

Pick a Daiso Korea ranking (best sellers: rising now, daily or weekly; overall or by category) and get one flat JSON row per product: rank, rank change since your previous run, price, units sold, online stock quantity, sold-out flag, review count and rating. Optional: Daiso barcode and wishlist count. A new-arrivals ranking and keyword search are included.

Daiso Korea (daisomall.co.kr) is South Korea's largest fixed-price variety store chain; most items cost 1,000-5,000 KRW. Its online rankings show which household, beauty and stationery items Korean shoppers are buying. This Actor is part of our K-retail ranking series together with the Olive Young Ranking Scraper and uses the same rank-tracking fields.

Live example: today's Daiso best-seller Top 100 is rebuilt every morning from this Actor's output, with rank changes since the previous day.

What you get

One dataset item per product:

• Ranking position: rank (position in the list), overallRank (Daiso's own rank number), previousRank (Daiso's previous rank, rising-now list only)
• Rank tracking across your runs: lastRunRank, rankChange (positive = moved up), isNewEntry, compared with your last complete run of the same list
• Product ID, Korean product name, brand (when Daiso lists one), price in KRW
• Units sold (totalSold, Daiso's cumulative order quantity) and online stock (onlineStockQty)
• Flags: sold out, new product, online only, store pickup, delivery, same-day delivery, bulk order
• Review count and rating, up to 3 category levels (Korean)
• Product URL, image URL, ranking page URL, approximate ranking refresh time
• With includeDetails: barcode, wishlistCount, brandEn, maxOrderQty
• Run summary (SUMMARY record, not charged): status and stop reason for every ranking or keyword

No login, no API key and no browser needed. Review text and reviewer data are not collected.

Use cases

• Product sourcing for resellers - schedule a daily run and see which Daiso items climb, enter the ranking or sell out
• Stock watching - online stock quantity and sold-out flag for every ranked item
• Supplier and competitor tracking - follow rank and units sold of products in your category (13 category rankings)
• K-beauty and "Daiso dupe" research - the beauty/hygiene ranking and the ranking of products launched in the last 7 days
• AI agents - one call with no required input returns a current, structured Daiso best-seller list

Input

All fields are optional. With no input you get the top 100 of the overall rising-now best-seller ranking.

Field	Description
mode	bestsellers (default), newArrivals (sales ranking of products launched in the last 7 days) or search (keyword search).
period	Ranking tab: risingNow (default, "급상승"), daily ("일간") or weekly ("주간"). Aliases R, D, W also work. Ignored in search mode.
categories	beautyHygiene, kitchen, cleaningBath, storage, stationery, interiorGarden, toolsDigital, food, sportsHobby, fashion, pets, babyToys, seasonal. Empty or all = overall ranking. One ranking per category. Ignored in search mode.
keywords	Search mode only. One search per keyword, up to 50. Korean keywords work best: 텀블러 (tumbler), 수납함 (storage box), 립밤 (lip balm).
maxItems	Products per ranking/category or keyword, 1-1000. Default 100.
includeDetails	Adds barcode, wishlistCount, brandEn, maxOrderQty, and fills onlineStockQty in the daily and weekly lists. One extra request per product; charged as a separate event, only when it succeeds. Default false.
excludeBoxProducts	Skip box (case pack) listings in search results. Default true.
historyStoreName	Key-value store that remembers previous lists for rank tracking. Default daiso-ranking-history. Use different names for separate watchlists.
proxyConfiguration	Keep the default (Apify datacenter proxy).

Weekly best sellers of two categories:

{
  "mode": "bestsellers",
  "period": "weekly",
  "categories": ["beautyHygiene", "storage"],
  "maxItems": 100
}

Keyword search (텀블러 = tumbler, 수납함 = storage box):

{
  "mode": "search",
  "keywords": ["텀블러", "수납함"],
  "maxItems": 200
}

Output

A full row from the weekly beauty/hygiene best-seller ranking with includeDetails on (real run, 2026-10-01, second run on the same history store):

{
  "schemaVersion": 1,
  "source": "daiso",
  "mode": "bestsellers",
  "period": "weekly",
  "rankingCategory": "beautyHygiene",
  "rankingCategoryId": "CTGR_01050",
  "keyword": null,
  "rank": 1,
  "overallRank": 1,
  "previousRank": null,
  "productId": "1030480",
  "name": "헬로 데일리 미용티슈 280매",
  "brand": null,
  "brandEn": null,
  "price": 1000,
  "currency": "KRW",
  "reviewCount": 21906,
  "rating": 4.9,
  "totalSold": 988800,
  "onlineStockQty": 12069,
  "isSoldOut": false,
  "isNew": false,
  "isOnlineOnly": false,
  "pickupAvailable": true,
  "deliveryAvailable": true,
  "bulkOrderAvailable": false,
  "todayDeliveryAvailable": true,
  "isBoxProduct": false,
  "category": "뷰티/위생",
  "categoryId": "CTGR_01050",
  "subCategory": null,
  "leafCategory": null,
  "barcode": "8819910304809",
  "gtin": "8819910304809",
  "wishlistCount": 17867,
  "maxOrderQty": 6,
  "detailsFetched": true,
  "url": "https://www.daisomall.co.kr/pd/pdr/SCR_PDR_0001?pdNo=1030480",
  "imageUrl": "https://cdn.daisomall.co.kr/file/PD/20260402/ujGdhVM7dPVFWG90s0g31030480_00_00ujGdhVM7dPVFWG90s0g3.jpg",
  "rankingUrl": "https://www.daisomall.co.kr/ds/rank/C105",
  "rankingUpdatedAt": "2026-10-01T14:42:00.000Z",
  "lastRunRank": 1,
  "rankChange": 0,
  "isNewEntry": false,
  "comparisonStatus": "compared",
  "qualityStatus": "ok",
  "previousRunId": "RtgXJxVzu3CwpTEQk",
  "previousScrapedAt": "2026-10-01T15:16:10.125Z",
  "runId": "PnHW59bnllais1oVN",
  "scrapedAt": "2026-10-01T15:22:50.134Z"
}

Selected fields of a row from the overall rising-now ranking, from a run 30 minutes after the previous one (real run, includeDetails off). The product moved from position 68 to position 6 in the list, so rankChange is 62:

{
  "mode": "bestsellers",
  "period": "risingNow",
  "rankingCategory": "all",
  "rank": 6,
  "overallRank": 25,
  "previousRank": 42,
  "productId": "1077128",
  "name": "논슬립 목 늘어남 방지 옷걸이 5개입",
  "price": 1000,
  "totalSold": 9651,
  "onlineStockQty": 1724,
  "subCategory": "옷걸이",
  "leafCategory": "플라스틱옷걸이",
  "barcode": null,
  "detailsFetched": null,
  "rankingUpdatedAt": "2026-10-01T14:45:00.000Z",
  "lastRunRank": 68,
  "rankChange": 62,
  "isNewEntry": false,
  "comparisonStatus": "compared",
  "previousRunId": "bfkcSxilpI91yWBIp",
  "previousScrapedAt": "2026-10-01T15:15:27.000Z",
  "runId": "YsjCoYDVQY8pkkRSu",
  "scrapedAt": "2026-10-01T15:45:18.972Z"
}

Every row has the same fields. A value Daiso does not provide is null, never left out. Numbers are numbers. Product, brand and category names are in Korean as on the site.

Rank tracking

Two different "previous" values, on purpose:

• previousRank - Daiso's previous rank number, given only in the rising-now list
• lastRunRank, rankChange, isNewEntry - your own history: the product's rank in your last complete run of the same list

Each run saves the list as a snapshot in the key-value store historyStoreName. A list is the combination of mode, category or keyword, period and maxItems (changing maxItems starts a new history). The next run fills lastRunRank, rankChange and isNewEntry and names the compared run in previousRunId / previousScrapedAt. comparisonStatus is compared, noBaseline (first run), skippedIncomplete (this list looked incomplete) or baselineSmaller (the compared list was clearly shorter, so isNewEntry is null).

The baseline only moves to a new run when all rows of that list were saved and the list looked complete (no big drop in size, few duplicates, few missing prices). A failed save, a run stopped by your maximum charge or an incomplete list leaves the old baseline in place. If two runs share a history store at the same time, the newer list wins.

On the first run the tracking fields are null. Schedule the Actor (daily or weekly) to build a rank history.

Pricing events

Pay per event:

• product - one product row saved to the dataset
• product-details - one product whose detail data (barcode, wishlist count, stock) was fetched, only with includeDetails and only on success

The SUMMARY record, failed requests and failed lists are not charged. When your maximum charge per run is reached the Actor stops, keeps what was saved and does not move the rank-tracking baseline for the list that was cut.

Run summary

The default key-value store record SUMMARY lists every ranking or keyword with status (succeeded, partial, failed, notStarted), stopReason (completed, noResults, chargeLimitReached, requestLimitReached, blocked, unexpectedResponse, emptyRanking, datasetWriteFailed), product counts, the number of products in the list on the site (totalOnSite), quality issues and whether this run became the new baseline (baselineUpdate). It also reports how many detail requests succeeded.

The run fails when every list failed, or when Daiso answered in an unexpected format (a sign the site changed - please report it in the Issues tab). Otherwise rows from successful lists are kept even if some lists failed.

Tips

• For rank tracking, prefer daily or weekly. The rising-now list is ordered by how fast sales are rising and turns over heavily: in our test, 83 of the top 100 products were new entries 30 minutes later. Use risingNow to discover products, daily / weekly to follow them.
• Need stock in the daily or weekly list? Turn on includeDetails; it fills onlineStockQty from the product detail data (50 of 50 products in our weekly test run).
• Keep maxItems and historyStoreName the same between scheduled runs, otherwise a new history starts and rankChange is null for that run.
• productId is stable; use it to join rows across runs, lists and keywords.
• Product names are Korean, so use Korean keywords in search mode.
• Runs with includeDetails are slower: about 3.5-4 seconds per product in our test runs, because the Actor waits 1-2 seconds between requests.

Notes

• Online stock only. onlineStockQty is the Daiso Mall online stock. Stock of individual offline stores needs a Daiso login and is not collected.
• Daily and weekly best-seller lists have fewer fields. Daiso does not report stock or sub-categories there, so onlineStockQty, subCategory, leafCategory and previousRank are null, and bulkOrderAvailable can be null. includeDetails fills onlineStockQty only. The rising-now list, new arrivals and search have these fields without details.
• Rising-now changes a lot between refreshes (see Tips). rankingUpdatedAt is approximate (Daiso shows "updated N minutes ago"), and we have not measured the refresh interval: two runs 2 minutes apart returned the same list, a run 30 minutes later a different one.
• rank is the position in the list Daiso returns. The Daiso website hides sold-out items in the best-seller tab, so rank can differ from the on-screen position; overallRank is Daiso's own number. In the rising-now list the order is Daiso's "fastest rising" order, not overallRank.
• totalSold is Daiso's cumulative order quantity for the product, not sales in the selected period. It is not guaranteed to only go up, and lists can disagree: in our tests one product showed 988,800 in the weekly list and 986,867 in the daily list half an hour later. Treat differences between runs as approximate.
• Search results come in Daiso's default relevance order only. Sorting by price, sales or date is not supported.
• Names are Korean only. brand is null for many items (25 of 50 in the example run); brandEn is filled only with includeDetails and only when Daiso has an English brand name.
• barcode is the barcode value in Daiso's product data. In the 50 products of the example run every value was 13 digits made of a Daiso prefix (04099 or 88199), the product ID and a check digit - a Daiso-assigned code, so do not expect it to match a manufacturer's GTIN in other catalogs. gtin repeats barcode when the check digit is valid.
• rankingUrl is the ranking page of the mode; category rankings are tabs on the same page.
• This Actor reads the same unofficial data endpoints the Daiso Mall website uses. Daiso can change them without notice; when the format changes the run fails with unexpectedResponse instead of returning wrong data.
• The Actor waits 1-2 seconds between requests and stops at 1,500 requests per run.
• Collects publicly visible product data only. No login, no reviews, no personal data.