Naver Map & Place Scraper 네이버 지도·플레이스 - Korean Business Data
Pricing
from $3.99 / 1,000 places
Naver Map & Place Scraper 네이버 지도·플레이스 - Korean Business Data
네이버 지도·플레이스 수집: 주소, 전화번호, 영업시간, 메뉴와 가격, 평점, 리뷰, 사진까지. 검색어 하나로 300곳, 지역 전체를 훑으면 그 이상. Scrape Naver Map and Naver Place for Korean local business data: addresses, phone, opening hours, menus with prices, ratings, reviews and photos.
Pricing
from $3.99 / 1,000 places
Rating
0.0
(0)
Developer
Zen Studio
Maintained by CommunityActor stats
1
Bookmarked
90
Total users
54
Monthly active users
5 days ago
Last modified
Categories
Share
Naver Map & Place Scraper (네이버 지도·플레이스) | Korean Local Business Data, Reviews & Photos (2026)
From Zen Studio, creators of Apple Maps, Yandex Maps and Naver Shopping. Proven experience collecting local business and Korean marketplace data.
Why choose this actor?
Go beyond the 300-place search limit.
Go beyond Naver’s 300-result search limit with area-wide collection. Select a district to collect thousands of businesses. Existing runs have returned 6,000 places, giving you broader coverage for prospecting and market research.
Phone numbers, menu prices and proper opening hours.
Export business contacts, street addresses and coordinates alongside menus and per-day hours, including breaks and last-order times where published. Build usable business records, not just a list of names.
Add the detail your research needs.
Collect visitor reviews, blog write-ups, photos and bookable services through optional settings. Reviews are exported to a separate dataset, ready for analysis.
Quick start
Start with 50 restaurants in Gangnam-gu. Increase Maximum places to collect a larger area dataset.
Open actor input to choose your search and area. Export JSON, CSV or Excel; use JSON to preserve nested menus and opening hours.
Area coverage takes longer than a standard search. Detail and optional reviews, photos and booking data carry additional charges. Fields depend on what each business publishes; Naver forwarding numbers are identified separately from business phone numbers.
| Zen Studio · Naver Suite | ||||
 Naver PlaceYou are here |  Cafe, Blog & NewsPosts & articles |  Naver SearchSearch & AI answers |  Naver ShoppingProducts & sellers |  Naver ReviewsBuyer reviews |
Copy to your AI assistant
Key Features
- 19 categories, one Actor. Restaurants, cafes, hospitals and clinics, beauty, hair salons, hotels and stays, attractions, pharmacies, pet services, convenience stores, supermarkets, schools, cinemas, gas stations, hiking trails, pop-up stores, golf courses, car services, plus All places, which searches across everything.
- Two kinds of review, not one. Short visitor ratings and the long blog and cafe write-ups Koreans actually read before choosing a place.
reviewSourcelabels every rowvisitor,receipt,payment,booking,blogorcafe. - A whole district, not just the first page. A single search term returns up to 300 places (up to 399 for restaurants, hair salons, pharmacies, attractions, pet services and gas stations). Set an Area and the Actor works through the entire region instead: a measured district sweep returned 45,940 of 46,248 places, 99.3%.
- Every review a place has. Not a sample: one place was drained of 1,149 of its 1,149 reviews. Reviews arrive with full text, star rating, visit date, reviewer tags, photos and the owner's reply.
- Opening hours you can actually use. Per day, with break times, last-order times and closed days, not one "Open now" string.
- Menus with prices, plus description, keywords, nearby subway and bus stops, the star-rating breakdown, and Naver blog posts written about the place.
- Photos from three sources: the owner's own, visitors' review photos, and stills from video clips, all as full-resolution URLs.
- Live bookable inventory. For places that take Naver reservations, the Actor returns the bookable services and current availability. 10 of 12 Gangnam restaurants in one test had bookable services; for golf courses this includes real tee times with prices.
- Korean or English output. Korean is the original data. English returns Naver's own translation of names, categories and addresses.
- 137 fields on a place row with every layer switched on, and nothing the source returns is thrown away.
How to Scrape Naver Map and Naver Place
- Open the Actor and type a search term into Search terms. Korean works best, e.g.
강남 맛집(Gangnam restaurants). You can also paste Naver Map place links. - Pick a Category.
All placessearches everything; a specific category narrows the search and unlocks that category's filters. - Set Maximum places, the run's total. With several search terms, set Maximum places per search term too, so every term contributes. To go past 300 on one term, set an Area as well.
- Press Start, then download the results as JSON, CSV or Excel, or pull them from the API.
Basic: restaurants in a neighbourhood
Every place in a whole district
coverageArea takes a Naver region code (0968 is Gangnam-gu, 09 is all of Seoul), or a map box as minLng,minLat,maxLng,maxLat. It covers the area's rectangle on the map, so places just across a district's border come back too (for Gangnam-gu, about half the places are in neighbouring districts); every row's address names its district, so you can keep one. This is the thorough mode and takes noticeably longer than a plain search.
With reviews and photos
Reviews land in a separate reviews dataset on the same run, one row per review, so a place row stays a place row.
Specific places by link
Golf tee times
Input Parameters
| Parameter | Type | Description |
|---|---|---|
searchQueries | array | What to look for. Each term is searched separately. Korean returns the richest results. |
placeUrls | array | Naver Map or Naver Place links, including short naver.me links, to collect specific places instead of or alongside searching. A link to a saved list cannot be opened (Naver shows those only to a signed-in account), and the run log says so. |
businessType | select | One of 19 categories, or All places. |
maxPlaces | integer | Total places to collect across every term and link, up to 200,000. Set 0 for every place available. |
maxPlacesPerQuery | integer | Places to take from each search term, so every term contributes. Leave empty to share maxPlaces across the terms in order; a term reached after the total is not searched, and the run log names it. |
coverageArea | string | Naver region code or map box, to sweep an entire area instead of taking one page of results. Covers the area's map rectangle, border places of neighbouring districts included. |
sortingOrder | select | Best match, Most visited, Most saved, Most reviewed, Nearest. |
menuKeyword | string | Keep only places offering this, e.g. 피자. Restaurants only; on any other category it is not applied and the run says so. |
| Filters | boolean | Open now, Open 24 hours, Takes reservations, Delivers, Takeaway, Naver Pay, Coupons, Newly opened, Step-free entrance. Category-specific, see below. |
includeDetail | boolean | Hours by day, menus, description, keywords, nearby stations, rating breakdown. On by default. |
includeReviews / maxReviewsPerPlace | boolean / integer | Visitor reviews into the separate reviews dataset. Set the number to 0 for every review. |
includeBlogReviews | boolean | Blog and cafe posts written about the place, into the same reviews dataset. reviewSource separates them. |
deepBlogSearch / maxBlogReviewsPerPlace | boolean / integer | Adds posts found by searching the place name, many more than the exact ones. blogSource marks each row place or search. |
includePhotos / maxPhotosPerPlace | boolean / integer | Full-resolution photo URLs. |
includeBooking | boolean | Bookable services and live availability. |
language | select | Korean (original) or English (Naver's translation). |
A note on filters
Naver applies these filters itself, and it only implements each one for certain categories. The Actor knows which, and skips a filter that does not apply to your chosen category, saying so in the run log, rather than sending it and returning nothing.
| Filter | Field | Works on |
|---|---|---|
| Open now | openNow | Restaurants |
| Open 24 hours | open24h | All places, Beauty, Cafes, Hotels & stays, Pet services, Pharmacies, Restaurants |
| Delivers | delivery | Restaurants |
| Takeaway | takeout | Restaurants |
| Takes reservations | booking | Restaurants |
| Accepts Naver Pay | npay | All places, Beauty, Cafes, Hair salons, Hotels & stays, Restaurants |
| Has coupons | coupon | All places, Attractions, Cafes, Hair salons, Restaurants |
| Newly opened | newlyOpened | All places, Cafes, Hair salons, Restaurants |
| Step-free entrance | wheelchair | All places, Attractions, Beauty, Cafes, Hair salons, Hospitals & clinics, Hotels & stays, Pet services, Pharmacies, Restaurants |
What Data Can You Extract from Naver Map?
One row per place. Counts arrive from Naver as display strings ("3,712"); the matching ...Num field is the same value as a number, and both are kept. saveCount is Naver's rounded figure, so saveCountNum is a bound: "84,000+" gives 84000 (at least that many) and "~100" gives 100 (at most that many). phone is the business's own line and virtualPhone is Naver's forwarding number; many places publish only the latter. hasDetail / hasPhotos / hasReviews / hasBookingData tell you which optional layers came back for that row.
The example below is built from real rows with every layer switched on. Only two things are shortened: arrays are cut to their first item or two, and single strings longer than 200 characters are middle-truncated. Every key is shown, at every level. Where one item lacks a key that others of its kind carry, the value shown comes from another real item of the same kind, so every key appears. Some keys belong to one kind of place and are null elsewhere: michelinGuide to restaurants in the Michelin Guide (isBib marks a Bib Gourmand), hotelMinPrice and motelMinStayPrice to hotels and stays, and the short description (an admission price, for example) to attractions.
Output Example: a place
Output Example: a review
Reviews go to the separate reviews dataset, one row per review. userIdno is Naver's own id for the reviewer, the same on every review they write, so it groups one person's reviews across places. Where one item lacks a key that others of its kind carry, the value shown comes from another real item of the same kind, so every key appears.
Output Example: a blog or cafe post
This is a post tied to the place (blogSource: "place"); the ranking signals (isOfficial, isRepresentative, rank, bySmartEditor3, hasNaverReservation) appear only on these. Posts found by name search carry the same core fields plus the blog's own, shown in the next example.
Blog and cafe posts are charged at the same rate as a review, and land in the same reviews dataset as the visitor reviews; reviewSource is blog or cafe. The full post text is in body (and its original key contents). buyWithMyMoneyType is true where Naver shows its 내돈내산 ("bought with my own money") badge on the post, and cafe posts carry a bypassToken, Naver's access token for opening the post from the place page. publishedAt is the post's date even where date is relative, as on recent posts ("2일 전"). Long text and the thumbnail list are trimmed here. Where one item lacks a key that others of its kind carry, the value shown comes from another real item of the same kind, so every key appears.
Output Example: a post found by name search
A post found by searching the place name (blogSource: "search", with Also search for posts by place name on). It carries the core fields of a place-linked post plus the blog's own: blogName, domainIdOrBlogId and logNo (which together address the post), addDate, noTagTitle, profileImgUrl and the thumbnail list. Long text and the thumbnail list are trimmed here. Where one item lacks a key that others of its kind carry, the value shown comes from another real item of the same kind, so every key appears.
How Much Does It Cost to Scrape Naver Map?
Pay per event: you pay for what the Actor actually returns, and nothing for a layer that came back empty.
| Event | Price |
|---|---|
| Place | $3.99 / 1,000 |
| Place detail | $1.99 / 1,000 |
| Review, or a blog or cafe post | $0.99 / 1,000 |
| Photo | $0.10 / 1,000 |
| Bookable services (per place) | $1.99 / 1,000 |
| Actor start | $0.005 per run |
A place whose detail did not come back is not charged for detail, and a place with no bookable inventory is not charged for booking.
Advanced Usage
Go past 300 places on one term
Set Area. A single term returns up to 300 places (399 for restaurants, hair salons, pharmacies, attractions, pet services and gas stations); an area sweep covers the whole region instead, section by section, so even a dense district comes back essentially complete.
Find places the default ranking buries
Sort by Nearest. It reorders the entire result set rather than nudging it, so it surfaces a different slice of the corpus than Best match does.
Track openings and closures
Schedule a district sweep weekly and diff placeId between runs. New ids are openings; ids that stop appearing are closures.
Monitor competitor ratings
Run a short list of placeUrls daily and track visitorReviewScoreNum and visitorReviewCountNum over time. Switch on Include reviews as well for the full rating breakdown (ratingAverage, ratingCount, ratingAuthorCount, ratingDistribution).
Build a menu price index
menus carries item names with prices. Sweep a category across a district and you have a priced menu dataset for the area.
Scheduling and Integrations
Schedule this Actor from the Apify Console: weekly suits district sweeps and openings tracking, daily suits rating and review monitoring on a fixed shortlist. Results push straight into Google Sheets, Airtable, Make, Zapier, Slack or your own endpoint via webhooks, and every run is available through the API in JSON, CSV or Excel.
New to scheduling on Apify? Apify's own walkthroughs cover it:
FAQ
Is Naver Place the same as Naver Map?
They are two views of the same data. A map.naver.com link and a pcmap.place.naver.com link both point at the same place record, and this Actor accepts either.
Why should I use this instead of a Google Maps scraper? Coverage. Google Maps is thin in Korea: small businesses often have no listing, no hours and no reviews there, while the same business on Naver has hundreds of reviews, a photographed menu with prices and bookable tables. For Korean local data, Naver is the primary source and Google is the copy.
How many places can I get from one search term? Up to 300 per term, and up to 399 for restaurants, hair salons, pharmacies, attractions, pet services and gas stations. With several terms, Maximum places is the run's total, so set Maximum places per search term to take some from every term. To collect more, set an Area: the Actor then covers the whole region, and a measured district sweep reached 99.3% of the district's places.
Does Naver Place have an API? Naver publishes developer APIs for search and maps, but they are keyed, quota'd, and do not return the things that make a place listing useful: the per-day opening hours with break times, the priced menu, the review text, the photo set, or the bookable inventory. This Actor returns all of them, in one row per place, with no key to obtain and no quota to manage.
Do I need a proxy? No. Everything needed to collect Korean listings is handled for you, and there is nothing to configure.
Can I get reviews for a specific place?
Yes. Paste its link into Place links, switch on Include reviews, and set Maximum reviews per place to 0 for all of them. Reviews arrive in the separate reviews dataset. Switch on Include blog and cafe posts as well to add the longer write-ups bloggers publish about the place.
Why did I get about 100 blog posts when the place shows thousands?
Those are two different numbers. The figure on the place page is how many posts Naver associates with the business; what Naver actually publishes for that place is a curated shortlist of up to about 100 to 150, and that is everything there is to collect. Each one provably belongs to that exact branch, which is what makes them worth having. placeBlogReviewsAvailable on each row tells you how many Naver lists for the place and placeBlogReviewsCollected how many actually arrived, so you are never guessing; with name search on, the collected figure also counts the posts it found and can be the larger one. To go wider, switch on Also search for posts by place name, which finds many more. Those are matched by name rather than tied to the business, so a branch whose name is nearly identical to another can pick up a few of its posts: measured at 3 to 6 percent on the hardest case, two branches differing only by 1호점 and 2호점, and zero where the name is distinctive. blogSource marks every row place or search, so you can keep only the exact ones whenever precision matters more than volume.
A place is missing a field I expected. Is that the real answer?
Yes, whenever hasDetail is true on that row: the full record arrived, so a field absent from it is absent at the source. Many Korean listings leave optional fields blank. hasDetail false means the record itself could not be read, which is almost always a removed or merged place. A place you supplied by link has nothing else to carry, so it is dropped rather than delivered as a bare row; one found by search keeps its search details and is marked hasDetail false. You are only charged for places whose record arrived.
Why is phone empty on some places?
Many Korean businesses publish only a Naver forwarding number. That is in virtualPhone, and it reaches the business.
Why did my filter get skipped? Naver implements each filter only for certain categories. The Actor skips one that does not apply to your category and notes it in the run log, rather than sending it and returning an empty run. See the filter table above.
Can I get the data in English? Yes, set Output language to English. Names, categories and addresses come back in Naver's own translation. Korean remains the original and the more precise of the two.
Is it legal to scrape Naver Place? This Actor collects only publicly visible data, nothing behind a login. You are responsible for complying with Naver's terms of service and with applicable data protection law: GDPR, CCPA, and Korea's Personal Information Protection Act (PIPA). Reviews contain personal data such as reviewer nicknames and photos, so treat them accordingly.
More Zen Studio scrapers for Asian platforms
🛒 E-commerce
TikTok Shop
Lazada
- Naver Search 네이버 검색
AliExpress 全球速卖通
- Naver Blog & Cafe 네이버 블로그·카페
- Naver Shopping 네이버쇼핑
Coupang 쿠팡
Taobao 淘宝
JD.com 京东
1688 阿里巴巴
Weidian 微店
Goofish 闲鱼
Dewu 得物
🎬 Short-video & social
Douban 豆瓣
Weibo 微博
Kwai (Brazil)
Kuaishou 快手
Douyin 抖音
RedNote 小红书
Xigua 西瓜视频
Bilibili 哔哩哔哩
🏠 Real estate & autos
-
Naver 부동산 -
Anjuke 安居客 -
58.com 58同城 -
Autohome 汽车之家- Autohome (Che168) Used-Car Scraper 🗺️ Maps & local
-
Kakao Map 카카오맵
Support
Found a bug, or need a field that is not here yet? Open an issue on the Actor's Issues tab. Questions about a specific run are answered fastest with the run ID.
Legal Compliance
This Actor accesses only publicly available pages. It does not log in, does not bypass access controls, and does not collect private data. You remain responsible for how you use the data, including compliance with Naver's terms of service, GDPR, CCPA and Korea's PIPA. Personal data appearing in reviews should be handled under the applicable regime.
Naver Place (네이버 플레이스) local business data for Korea: names, addresses, coordinates, phone numbers, opening hours, menus with prices, ratings, reviews with photos, and live bookable inventory, by search term, by place link, or swept across a whole district.
![Europages [$6💰] B2B Company Data + Emails & VAT IDs avatar](https://images.apifyusercontent.com/wz09jINlk8Vak8_v_-pJg3T5XihmtQuLw4MkXX2tPyw/rs:fill:76:76/cb:1/aHR0cHM6Ly9hcGlmeS1pbWFnZS11cGxvYWRzLXByb2QuczMudXMtZWFzdC0xLmFtYXpvbmF3cy5jb20vbmduOHZSRVNUa3NleVdFNnAtYWN0b3ItZUtUWWlhVU9tbzFrZEQ0N0QtcHpFVENCNVpzWi1ldXJvcGFnZXMucG5n.webp){kind=avatar}
