Naver Blog Brand Monitor — sponsored vs organic posts

For brand marketers, PR teams and agencies in Korea who need to know what Naver Blog — the #1 place Koreans read product, restaurant and beauty reviews — says about a brand, and how much of it is paid (체험단 / 협찬) versus organic word of mouth, for $4 per 1,000 posts plus $1 per 1,000 judgments.

Enter a brand or product keyword and get every matching post with:

• Sponsored vs self-paid — 체험단 / 협찬 / 원고료 (sponsored) vs 내돈내산 (self-paid) vs unclear, with the evidence (disclosure sentence or review-platform banner)
• Affiliate flag — Coupang Partners, Naver Shopping Connect and similar commission links
• Sentiment — positive / neutral / negative toward the product or place
• Brand mentions — did the post mention each brand you care about (Korean or English spelling)?
• Post metadata — title, blog, publish time, snippet, image count, likes and comments, and full text on request

Use it to: monitor a brand on Naver Blog (네이버 블로그) · sponsored post detection for Korean influencer marketing · count 체험단 (experience-group) posts vs organic reviews · Korean social listening · Naver blog scraper with sentiment.

Output sample

Real rows from runs 4MpsWAuVQeJEw1AxJ (query "다이슨 에어랩 후기", brands 다이슨 / Dyson) and 68ansfQ5krx8lYz3J (query "라운드랩 선크림", brands 라운드랩 / Round Lab), both 2026-10-06. English in parentheses is added here for readers; the Actor returns the Korean title.

title (Korean)	sponsored.label (probability, method)	evidence	sentiment	brandMentions	likeCount / commentCount
다이슨 에어랩 아이디 내돈내산 후기｜8년 쓴 컴플리트와 비교… (Dyson Airwrap i.d. self-paid review, compared with my 8-year-old Complete)	self_paid (0.95, rule)	self_paid_phrase: "…내돈내산 후기…"	neutral	다이슨 true · Dyson true	0 / 0
차홍 에어스타일러 (Chahong air styler)	self_paid (0.95, rule)	self_paid_phrase: "…#내돈내산 다이슨에어랩을 사고…"	neutral	다이슨 true · Dyson true	0 / 0
다이슨 에어랩 코안다 2x 후기 (Dyson Airwrap Coanda 2x review)	unclear (1.00, model)	—	neutral	다이슨 true · Dyson true	4 / 0
바르는 코엔자임q10 추천, 하이퍼셀·이데베논 앰플… (Topical coenzyme Q10 ampoules)	sponsored (0.95, model)	—	positive	다이슨 false · Dyson false	1 / 0
[내돈내산] 벌써 2통째! 라운드랩 자작나무 수분선크림 푸른자차 실사용 후기 (Self-paid: my 2nd tube of Round Lab birch sunscreen)	self_paid (0.95, rule)	self_paid_phrase: "[내돈내산]…"	positive	라운드랩 true · Round Lab true	3 / 0

Every row also has url, blogName, publishedAt, snippet, imageCount, reactions (공감 breakdown), sponsored.affiliate and the probabilities behind each label — see Output below.

Input example

The form default — one keyword, the 8 most recent posts, one brand (about $0.04, 9 seconds):

{
  "queries": ["라운드랩 선크림"],
  "brandNames": ["라운드랩"],
  "maxPostsPerQuery": 8
}

For monitoring, raise maxPostsPerQuery (up to 1,000), add "dateFrom": "2026-09-01" and both spellings of your brand (["라운드랩", "Round Lab"]). Posts are deduplicated across queries. Sort by most recent (default) for monitoring, or by relevance.

Pricing

Pay only for results — no subscription.

Event	Price	When
post	$0.004	One blog post collected (metadata, snippet, image count; full text if requested).
post-judged	$0.001	Sponsored/self-paid, sentiment and brand-mention judgments for one post (only when judging is on).

That is $4 per 1,000 posts collected plus $1 per 1,000 posts judged (sponsorship, sentiment and brand mentions) — $5 per 1,000 posts with everything on. First run with the form defaults: about $0.04 (8 posts, 9 seconds).

Cost examples

Run	Collection only	With judgments
100 posts	$0.40	$0.50
1,000 posts	$4.00	$5.00
5,000 posts	$20.00	$25.00

With the free $5 monthly Apify credit you can collect and judge about 1,000 posts.

Posts that fail to load are reported with an error field and not charged. If you set a maximum cost per run the Actor stops cleanly; if the budget only covers collection, judging is turned off and collection continues.

Works with

• Apify API, JavaScript and Python clients — start a run and read the dataset like any Actor (leoworks/naver-blog-brand-monitor).
• Apify Schedules — a daily run with dateFrom keeps a rolling log of new posts; our own monitoring runs this Actor every morning.
• Claude, Cursor and Claude Code through the Apify MCP server — see the next section (verified 2026-10-06).
• Korean Review Classifier — the same complaint, sentiment and motive labels for any review dataset: leoworks/korean-review-classifier.

Use with Claude, Cursor or Claude Code (MCP)

Add the Apify MCP server with this Actor as a tool and ask your agent in plain language — for example "Find the latest 30 Naver Blog posts about 라운드랩 선크림 and tell me how many are sponsored (체험단) and what the self-paid ones say." The agent calls the tool leoworks--naver-blog-brand-monitor and reads the posts with get-dataset-items.

Claude Desktop or Cursor (mcp.json):

{
  "mcpServers": {
    "apify": {
      "url": "https://mcp.apify.com?tools=leoworks/naver-blog-brand-monitor",
      "headers": { "Authorization": "Bearer YOUR_APIFY_TOKEN" }
    }
  }
}

Claude Code: claude mcp add --transport http apify "https://mcp.apify.com?tools=leoworks/naver-blog-brand-monitor" --header "Authorization: Bearer YOUR_APIFY_TOKEN". Leave out the header to sign in with OAuth in the browser instead. Your Apify token is in Console → Settings → API & Integrations. We verified this setup with the Apify MCP server (v0.17.2) on 2026-10-06.

Output (one row per post)

{
  "query": "라운드랩 선크림",
  "url": "https://m.blog.naver.com/…/224428209119",
  "title": "눈시림없고 백탁없는 라운드랩 자작나무 수분 선크림",
  "blogName": "솝비의 나날들",
  "publishedAt": "2026-10-01T04:53:26.388Z",
  "snippet": "안녕하세요 솝비입니다! 요즘 외출할 때 선크림은…",
  "likeCount": 132,
  "reactions": { "like": 128, "impressive": 4 },
  "commentCount": 81,
  "imageCount": 6,
  "sponsored": { "label": "sponsored", "probability": 0.99, "method": "rule", "evidence": [{ "type": "platform_banner", "source": "revu.net" }], "affiliate": false },
  "sentiment": { "label": "positive", "probability": 0.97, "confidence": 0.95 },
  "brandMentions": [{ "brand": "라운드랩", "mentioned": true, "probability": 1, "method": "exact" }]
}

likeCount is the number shown under the post (all reactions — 공감, 감사, 칭찬 …; breakdown in reactions) and commentCount the number of comments, both read without login. If the counts cannot be loaded they are null; the post is still returned.

How sponsorship is decided

1. Rules first — Korean FTC-style disclosure sentences (원고료, 제품을 제공받아, 협찬, 체험단을 통해…) and banners/links of review-experience platforms (Revu, ReviewNote, Dinner Queen and others) → sponsored; explicit 내돈내산 / 직접 구매 → self_paid. Evidence is returned.
2. Model second — posts without explicit evidence are judged by Jev, TypeSafe's decision model, and stay unclear unless it is confident.
3. Affiliate commission disclosures ("…수수료를 제공받습니다") are reported as affiliate: true, not as sponsorship.

Accuracy: 92% on 50 hand-labelled posts (beauty, restaurants, camping, hair tools). Hidden sponsorship that is never disclosed cannot be detected by anyone — unclear is an honest answer.

Summary by keyword and brand (REPORT)

Each run also saves a REPORT record (Output tab → Summary by keyword and brand) at no extra charge — for all posts, for each keyword and for each brand you listed: the sponsored / self-paid / unclear split, affiliate-link share, sentiment of all posts and of self-paid posts only (the closest thing to organic customer voice), posts per month, and the 3 latest negative posts that are not sponsored.

{
  "brand": "라운드랩",
  "posts": 42,
  "judged": 42,
  "sponsorship": { "sponsored": 0.548, "self_paid": 0.333, "unclear": 0.119 },
  "affiliateShare": 0.071,
  "sentiment": { "positive": 0.857, "neutral": 0.095, "negative": 0.048 },
  "selfPaidSentiment": { "positive": 0.714, "neutral": 0.143, "negative": 0.143 },
  "postsByMonth": { "2026-09": 25, "2026-10": 17 },
  "latestNegativePosts": [{ "title": "…", "url": "https://blog.naver.com/…", "publishedAt": "2026-10-02T…", "sponsorship": "self_paid" }]
}

Limits

Item	Limit
Posts per query	Up to 1,000 (what Naver search returns)
Brands to check	Up to 10 per run
Speed	About 998 posts with judgments in 6 minutes
Source	Naver Blog (blog.naver.com) posts written with Naver's SmartEditor
Personal data	Blog name and nickname only — no real names, emails or phone numbers

Optional: use your own NAVER API Hub key

By default posts are discovered through Naver web search. For very large daily volumes you can plug in your own NAVER Cloud key: create a NAVER Cloud account → NAVER API Hub → create an application with Search → copy the key ID and key into the two Advanced fields (stored as secrets). The search API is free up to Naver's daily limit at the time of writing.

FAQ

Which AI makes the judgments? Jev, TypeSafe's decision model (version jev-1.13.0, pinned). Jev answers each label with a calibrated probability instead of generated text, so the same input gets the same answer from run to run. We re-measure accuracy on our labelled test set before we change the model version. Only the post title and text (first 2,500 and last 1,500 characters) and your brand names are sent to Jev; blog names and nicknames are not.

Is this the official Naver API? Post discovery uses Naver's public mobile search by default (or your own NAVER API Hub key); post content is read from public mobile blog pages. No Naver login is used.

Does it generate summaries? No — it collects posts and adds labels with probabilities.

Disclaimer

Independent tool — not affiliated with, endorsed by or sponsored by NAVER Corp. "Naver" is a trademark of its owner and is used only to describe the data source.

Reviews and support

If this Actor saved you time, a short review on Apify Store helps others find it. Questions or a query that does not work? Open an issue in the Issues tab — we answer within a day.

Changelog

See the Changelog tab.