Naver Shopping Brand-Store Search

This Actor enters product keywords into the search form of public brand.naver.com stores and extracts the product cards returned by the website. It supports one query or a batch of queries, bounded pagination, three website sort modes, and optional authorized proxies.

Search scope

Naver's general Shopping result origin returns HTTP 418 or maintenance responses from many data-center environments. This Actor therefore uses Naver's public autocomplete service to select a relevant brand store, then performs the requested search inside that store. Every record declares searchScope: "brand-store" and includes the selected store name and URL. To select the store yourself, provide one public https://brand.naver.com/<store> URL in startUrls.

This is not a claim of marketplace-wide coverage. If autocomplete cannot find a public brand store for a query, provide a brand-store URL or use a more brand-specific query.

Inputs

• query: one product keyword, such as 노트북.
• queries: additional product keywords for batch mode.
• startUrls: optional public Naver brand stores. One URL can be reused for all queries, or paired with queries by position.
• sortBy: relevance, newest, or reviews.
• maxItemsPerQuery: maximum products stored for each keyword (1–200).
• maxPages: maximum website pages opened per keyword (1–20).
• maxConcurrency: simultaneous query searches (1–10).
• maxRequestRetries: bounded retries for temporary browser or network errors.
• navigationTimeoutSecs and requestHandlerTimeoutSecs: timeouts.
• proxyConfiguration: Apify Proxy or user-provided proxy URLs. An explicit disabled setting is honored.

The legacy maxItems input is still accepted as an alias for maxItemsPerQuery.

Output

Records contain source-backed values only. Required identity, URL, pricing, store, query, position, source, and timestamp fields are always present. reviewCount and rating are omitted when the website does not show review data; hasReviews remains present so consumers can distinguish that case.

Important fields include:

• stable id and productId;
• canonical productUrl without expiring click signatures;
• title, imageUrl, price, priceNumeric, and currency;
• rating, reviewCount, and hasReviews;
• brandStoreName, brandStoreUrl, and categoryId;
• position, page, pagePosition, and resultTotal;
• searchQuery, sortBy, searchScope, and searchUrl;
• source, extractionMethod, and ISO-8601 scrapedAt.

Extraction behavior

The Actor first uses Naver's structured autocomplete JSON to discover a store, validating status and content type before parsing it. Naver does not expose a reliable structured response for the query-specific cards in the observed store-search flow, so the cards use a DOM fallback based on stable semantic and tracking attributes. Challenge or block pages are rejected and never stored.

The Actor uses persistent crawler sessions, a consistent desktop browser profile, bounded retries, and optional authorized proxies. It does not solve or bypass CAPTCHAs, login requirements, or access controls.

Example

{
  "query": "노트북",
  "queries": ["모니터"],
  "startUrls": [{ "url": "https://brand.naver.com/samsung" }],
  "sortBy": "relevance",
  "maxItemsPerQuery": 20,
  "maxPages": 2,
  "proxyConfiguration": { "useApifyProxy": false }
}

Run locally with:

$apify run --purge --input-file INPUT.json