Naver SERP Search Scraper

Turn Korean search queries into structured Naver SERP dataset rows across web, news, image, video, and shopping. Paste one keyword or a batch, pick a vertical, and the Actor returns ranked titles, links, snippets, press names, image URLs, channel labels, and shopping prices when the public results page exposes them. No NAVER API HUB Client ID or Secret.

This Actor is built for Korea SEO, brand-monitoring, and agent workflows that need SERP coverage including video and shopping cards from public search HTML.

Best fit for this Actor

• Monitor Korean keywords across news, web, image, video, and shopping in one schema.
• Feed MCP or CRM pipelines with one predictable row per result (resultType, position, title, link).
• Sample shopping cards from nexearch product blocks when you need price-aware SERP snapshots.

For official HUB Search API rows with your own Client ID/Secret, continue with Search Naver Open API. For deeper product catalogs after a shopping SERP sample, continue with Naver Product Shopping Scraper. For place and map enrichment, use Naver Map Scraper.

Practical scenario

A Korea SEO analyst tracks 서울 맛집 in the news vertical. They paste the query, set where to news, keep maxResultsPerQuery at 5, and leave Residential KR proxy on. The run returns five ranked news rows with title, link, press name, and snippet. The same input pattern works for 무선이어폰 with where=shopping when they need nexearch product cards before deciding which listings to enrich next.

Quick start input

{
  "query": "서울 맛집",
  "where": "news",
  "maxResultsPerQuery": 5
}

query and queries can be combined; duplicates are collapsed. where accepts web, news, image, video, shopping, or nexearch. Prefer Korean keywords and Apify Residential with country KR.

Input reference

Field	Type	What it controls
query	string	Single search keyword. Example: 서울 맛집.
queries	array	Batch keywords merged with query. Max 100 unique values.
where	enum	Vertical: web, news, image, video, shopping, nexearch. Default web.
maxResultsPerQuery	integer	Rows kept per query after parse. Default 30, max 300. Prefill uses 5.
proxyConfiguration	object	Prefer Apify Residential + KR for stable public SERP HTML.

What data you receive

One dataset item is one SERP result row.

{
  "resultType": "news",
  "query": "서울 맛집",
  "where": "news",
  "position": 1,
  "title": "Example headline",
  "link": "https://n.news.naver.com/article/example",
  "snippet": "Short snippet…",
  "pressName": "Example Press",
  "newsDate": "1시간 전",
  "collectedAt": "2026-09-22T12:00:00.000Z"
}

Field	Meaning
resultType / where	Vertical that produced the row
position	1-based rank within the query
title + link + snippet	Public SERP card text and destination URL
pressName / newsDate	News vertical fields when present
imageUrl / channelName / price	Image, video, or shopping fields when present

OUTPUT and RUN_SUMMARY in the default key-value store hold outcome, itemsPushed, and charged event counts. Outcomes include COMPLETE, PARTIAL, VALID_EMPTY, INVALID_INPUT, and UPSTREAM_FAILED. Download the dataset as JSON, CSV, Excel, or HTML.

Use through the API

curl -X POST "https://api.apify.com/v2/acts/khadinakbar~naver-serp-search-scraper/runs" \
  -H "Authorization: Bearer $APIFY_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"query":"서울 맛집","where":"news","maxResultsPerQuery":5}'

CLI alternative:

$apify call khadinakbar/naver-serp-search-scraper --input='{"query":"서울 맛집","where":"news","maxResultsPerQuery":5}'

Use with AI agents through Apify MCP

Scrape Naver news SERP for 서울 맛집. Return title, link, pressName, position, and snippet. Read OUTPUT.outcome and itemsPushed after the run. Budget: about $0.005 per accepted search-result plus platform usage. Scope: public Naver search HTML only. Provenance: collectedAt on each row.

Connect via https://mcp.apify.com. MCP tool: apify--naver-serp-search-scraper. Keep maxResultsPerQuery small on the first call while confirming yield.

Connect the workflow

• After a shopping SERP sample, enrich product pages with Naver Product Shopping Scraper.
• When you need official HUB Search API quotas with BYOK keys, use Search Naver Open API.
• For place and map context around a keyword, continue with Naver Map Scraper.

Pricing

This Actor uses Pay per event plus Apify platform usage. The live Pricing tab is the current source of truth for billing details.

• apify-actor-start: $0.00005 per run
• search-result: $0.005 per accepted SERP result row

A five-result news prefill is about $0.025 in result events plus platform usage. Empty-match and invalid-input runs have no search-result charge.

How it works

1. Normalize and deduplicate query / queries, then map where to the matching Naver search URL.
2. Fetch public SERP HTML with Cheerio and Apify Residential KR by default.
3. Parse SDS/Fender cards and embedded JSON into one row per result.
4. Charge search-result for each accepted row, then write the dataset item plus OUTPUT / RUN_SUMMARY.

Best results

• Prefer Korean keywords that match the vertical intent (news, shopping, video).
• Keep Apify Residential with country KR enabled for stable public HTML.
• Start with maxResultsPerQuery near 5 while confirming yield, then raise the cap.
• An empty or whitespace-only query finishes SUCCEEDED with INVALID_INPUT and no result charge.
• A nonsense query that returns no cards finishes SUCCEEDED with VALID_EMPTY and no result charge.

Builder's note

I found that public Naver SERP HTML on search.naver.com already exposes web, news, image, video, and nexearch shopping cards without Open API keys, which is enough for a lean multi-vertical scraper. Dedicated shopping hosts can return access challenges from non-KR routes, so this build keeps shopping on nexearch product cards and couples Actor.charge('search-result') to each validated row write so billing stays aligned with useful dataset items.

Legal and responsible use

Use this Actor on public Naver search pages you are authorized to process, follow applicable law and Naver's terms, and keep the output in your own compliance workflow. Naver is a trademark of Naver Corp. This independent Actor is not affiliated with, associated with, or endorsed by Naver.

Issues and feature requests: use the Actor Issues tab on Apify.