Naver Map Scraper

Scrape Naver Map — South Korea's dominant local-search and maps platform. Search places by keyword or fetch specific places by URL, and get full business details: address, phone, hours, menus, photos, amenities, transit access, and review metrics. No login, no API key, and no paid proxy required.

What this actor does

• Two modes: search (keyword-based discovery) and byUrls (fetch specific places you already know)
• Broad category coverage: restaurants, cafes, bars, hotels, gyms, and general businesses, plus medical/dental clinics and hair/nail salons — Naver Map routes those verticals through separate endpoints, which this actor detects and handles automatically
• Rich detail enrichment: business hours by day, full menu with prices and photos, amenities, nearby subway/bus stations, photo gallery
• Review signals: visitor review count/score, blog & cafe review count, save count — all included by default (no extra request)
• Filters: category, minimum review score, minimum review count, open-now, has-booking
• Empty fields are omitted — every record only contains fields Naver Map actually returned for that place

Output per place

• placeId, name, category, businessCategory, categoryCodeList
• roadAddress, address, commonAddress, fullAddress
• x, y (coordinates), distance (from search origin, search mode only)
• phone, virtualPhone, hasMobilePhoneNumber
• talktalkUrl, chatBotUrl, routeUrl, placeUrl, homepageUrl (mode includeDetails=true)
• businessHoursStatus, businessHoursDescription — quick status (e.g. "Open now")
• businessHours[] — full weekly schedule (mode includeDetails=true)
• conveniences[], paymentInfo[], description, businessDescription (long-form owner write-up, mode includeDetails=true), microReview
• imageUrl, imageUrls[], imageCount, detailPhotos[], visitorImages[] (photos from visitor reviews), totalImageCount, panoramaUrl, staticMapUrl (mode includeDetails=true for the last four)
• visitorReviewCount, visitorReviewScore, blogCafeReviewCount, totalReviewCount, saveCount
• reviewImageCount, reviewAuthorCount, reviewThemes[], reviewKeywords[] — review-analysis breakdown (topics and voted tags with mention counts, mode includeDetails=true)
• hasBooking, bookingUrl, bookingBusinessId, bookingReviewCount, naverBookingCategory
• hasNPay, newOpening, hasWheelchairEntrance, promotionTitle, broadcasts[] (TV features), michelinGuide
• naverOrder — delivery/table-order/pre-order/pickup flags, only present when supported
• businessToolsUsing[] — active Naver business tools (e.g. naverBooking, smartOrder, smartCall), mode includeDetails=true
• coupons[], couponCount — active promotion titles and count
• posInsights — popular menu, average spend, and average stay time (search results)
• posSalesCount, posPopularDay, posAveragePrice (by meal period), posPopularMenus[] — richer POS breakdown (mode includeDetails=true, where the business shares POS data)
• menus[] — name, price, description, photo, recommended flag (mode includeDetails=true)
• facilities[], keywords[], parkingInfo, petFriendly, noKidsZone, subwayStations[], busStations[], relatedLinks[] (external listing sites) (mode includeDetails=true)
• recordType: "place", scrapedAt

Input

Field	Type	Default	Description
mode	string	search	search or byUrls
searchKeywords	array	["강남 맛집"]	Keywords to search (mode=search). Include a region/neighborhood name for local results.
placeUrls	array	–	Naver Map place URLs to fetch directly (mode=byUrls). Accepts full desktop, mobile, and naver.me short links.
category	string	–	Only keep places whose category contains this text (e.g. 카페, 한식)
maxResultsPerKeyword	int	20	Cap per keyword (1–100, Naver's real ceiling), mode=search
includeDetails	boolean	true	Fetch the place detail page for hours, menus, amenities, photos, transit
includeReviews	boolean	false	Emit embedded visitor reviews as separate records (mode=search only; see FAQ)
maxReviewsPerPlace	int	50	Cap on reviews fetched per place when includeReviews is on
minVisitorReviewScore	number	–	Only keep places at/above this review score (0–5)
minVisitorReviewCount	int	–	Only keep places with at least this many visitor reviews
openNow	boolean	false	Only keep places currently marked open
hasBooking	boolean	false	Only keep places with online booking
maxItems	int	100	Hard cap on total emitted records (1–5000)
proxyConfiguration	object	{"useApifyProxy": true}	Optional — Naver Map works without proxy for most traffic

Example: search a neighborhood for top-rated cafes

{
  "mode": "search",
  "searchKeywords": ["이태원 카페"],
  "maxResultsPerKeyword": 30,
  "category": "카페",
  "minVisitorReviewScore": 4.5,
  "includeDetails": true
}

Example: fetch specific places you already know

{
  "mode": "byUrls",
  "placeUrls": [
    "https://pcmap.place.naver.com/restaurant/2020647783/home",
    "https://naver.me/xxxxxx"
  ],
  "includeDetails": true
}

Example: fast bulk discovery without detail pages

{
  "mode": "search",
  "searchKeywords": ["강남 맛집", "홍대 맛집", "부산 해운대 맛집"],
  "maxResultsPerKeyword": 50,
  "includeDetails": false,
  "maxItems": 150
}

Example: only currently-open bookable restaurants

{
  "mode": "search",
  "searchKeywords": ["신촌 맛집"],
  "openNow": true,
  "hasBooking": true
}

Scope / Limitations

• Hair/nail salon results don't include a native category value from Naver (that vertical's data simply doesn't carry one) — the actor fills in 미용실/네일샵 respectively so the category filter still works for them, but a few other business-metadata fields common on restaurants (e.g. posInsights, saveCount) aren't available for this vertical because Naver doesn't return them there.
• includeReviews returns only a few reviews per place (typically 2–3), sourced from the same search response the rest of the record comes from. Naver's dedicated deep review-pagination endpoint applies bot-detection to repeated automated requests, so this actor intentionally does not use it — review counts and scores remain fully reliable regardless.
• category is a substring filter, not a structured taxonomy lookup — it matches against whatever category string Naver assigned the place (which can be compound, e.g. 카페,디저트).
• No standalone "browse by category" mode. Naver Map's category browsing requires either a keyword or map bounds; use a keyword plus the category filter to approximate this.
• 100 results per keyword is the real, reliable ceiling. Naver's list page only exposes deeper pagination through an endpoint that applies bot-detection to automated requests; use multiple/narrower keywords (e.g. by neighborhood) for broader coverage instead of expecting one keyword to return thousands of results.
• Reviews (mode=search only). includeReviews produces recordType: "review" records in the same dataset, each with placeId, sourceUrl, reviewId, content. Not available in byUrls mode.

Use cases

• Local SEO & marketing agencies — audit business listings, hours, and review scores for clients across Korea
• Restaurant/cafe discovery apps — bulk-import venue data with menus, photos, and transit access
• Market research — analyze review volume and ratings across a neighborhood or category
• Real estate & site selection — evaluate foot-traffic proxies (review counts) and nearby transit for a location
• Travel content — build curated guides with real business hours, menus, and coupon data

FAQ

Is this affiliated with Naver? No. This is an independent, third-party actor that reads Naver Map's public search and place pages. It is not affiliated with or endorsed by Naver Corporation.

Do I need a Naver account or API key? No. The actor reads the same public pages your browser loads when you search Naver Map — no login, cookies, or API key required.

Why are some fields missing on some places? Naver Map only returns the fields a business has actually filled in. A place without a booking system won't have bookingUrl; a place without a menu photo won't have menus[].images. Empty fields are omitted rather than sent as null.

What does includeDetails=false skip? Full weekly business hours, menus, amenities/facilities, nearby transit, and the extended photo gallery. Core fields (address, phone, review counts, coupons, business-hours status) are still included from the search results.

How many reviews does includeReviews return per place? Up to a few (usually 2–3) real visitor reviews per place, sourced from the same search response — no extra request needed, so it's always reliable. Naver's separate, deeper review-pagination endpoint applies bot-detection to repeated automated requests, so this actor deliberately does not use it; review counts and scores (via the core fields, always included) remain the reliable way to gauge review volume beyond what individual review text is available for.

How does category filtering work? It's a substring match against the place's Naver-assigned category (e.g. 카페, 한식, 요리주점). Leave it blank to get every category returned by your keyword.

Can I search in English? Yes, but results are far better with Korean keywords, since Naver Map is primarily indexed in Korean. Including a Korean neighborhood/region name in the keyword significantly improves relevance.

Is a proxy required? No. Naver Map's search and detail pages work over the default (free) Apify proxy configuration. The proxyConfiguration input is available if you want to route traffic differently, but it's optional.