Yelp Scraper

Scrape Yelp business listings by search term and location, or collect public details from business URLs you already have. Export Yelp business data to CSV, JSON, or a spreadsheet, including contact information, ratings, opening hours, categories, amenities, and photos. This is a Yelp API alternative for public business profiles; it does not require a business API key.

Each dataset item represents one business. The scraper combines search discovery with the structured data on each business page. It can search several terms across several locations, process direct business or search URLs in the same run, and stop at a total result limit. Duplicate business URLs found by multiple searches are processed once.

What data can you extract from Yelp?

Field	Description	Example
name	Displayed business name	Homeslice Neal's Yard
bizId	Stable Yelp business identifier	3AqrHgpIbcMXj0q-p_oPzg
directUrl	Canonical business page	https://www.yelp.com/biz/homeslice-neals-yard-london
alternateNames	Other published names	[]
address	Street lines, city, region, postal code, country	{"city":"London","country":"GB"}
latitude, longitude	Coordinates when present in the page	51.514438, -0.126427
phone	Displayed business number	020 3151 7488
website	Business website URL	http://www.homeslicepizza.co.uk/
aggregatedRating	Overall rating	4.4
reviewCount	Published review total	291
priceRange	Displayed price indicator	££
type	Broad category	Restaurants
cuisine	Restaurant category, when applicable	Pizza
categories	All displayed category labels	["Restaurants","Pizza"]
operationHours	Daily regular hours and exceptions	Monday, noon to 11 PM
upcomingSpecialHours	Published future exceptions	[]
amenitiesAndMore	Active services and amenities	["Offers take-out","Vegan options"]
businessHighlights	Business supplied highlights	[]
aboutTheBusiness	Specialties, history, year established	{"yearEstablished":"2011"}
healthScore	Published inspection score, where offered	null
claimed	Whether the business profile is claimed	true
verified	Whether the owner is verified	false
advertiser	Whether the business advertises on Yelp	false
isClosed	Closed business flag	false
primaryPhoto	Main photo URL	A public image URL
images	Photo URLs with captions	[{"link":"…","alt":"Salami, rocket, parmesan"}]
summary	Brief public description	null
menuUrl	External menu URL, when supplied	http://www.homeslicepizza.co.uk/menu/
scrapeStartedAt, scrapeFinishedAt	UTC processing timestamps	ISO 8601 timestamps

The address object includes addressLine1, addressLine2, addressLine3, city, regionCode, postalCode, and country. Every profile uses the same set of top-level keys. Missing optional facts appear as null or an empty list, making exports easier to join and compare.

No individual review text is collected. aggregatedRating and reviewCount describe the business as a whole. The images array contains photo links and captions, not downloaded image files. Location coordinates may be absent for profiles that do not expose a public point on the map.

How to scrape Yelp

1. Enter one or more search terms, such as Pizza or Plumbers, and one or more locations. Every term is paired with every location. You can also paste direct business or search URLs.
2. Set Results per search to the number of businesses you want from each pair. Set Total businesses when you need a cap across the whole run.
3. Choose Photos per business. The default is 20 links. Set it to zero if photo links are not needed.
4. Run the scraper and open its dataset. Preview the results or download them as CSV or JSON.

For example, two terms and three locations create six separate searches. If a business appears in more than one of those searches, it is output only once. Direct business URLs are processed as well, so you can add a known competitor to a broader search run. A direct search URL must include both find_desc and find_loc query parameters.

Input

Parameter	Type	Default	Description
searchTerms	List of strings	Empty	Business names, products, or categories to search.
locations	List of strings	Empty	Cities or areas searched for every term.
searchLimit	Integer	10	Maximum profiles processed for each term and location pair, up to 240.
directUrls	List of strings	Empty	Business or search URLs to process alongside keyword searches.
maxImages	Integer	20	Photo links returned per business. Zero skips photos.
maxItems	Integer	0	Maximum dataset items across the whole run; zero adds no overall cap.
proxyConfiguration	Object	Enabled	Network routing for public page requests.

Provide at least one complete term and location pair, one direct URL, or both. Search terms without locations are skipped with a warning. If no valid source remains, the run fails with a clear input error. Direct URLs can point to regional Yelp domains; the scraper resolves the same public business data and outputs a canonical yelp.com business URL.

This input collected 35 distinct public profiles in a completed verification run:

{
  "searchTerms": ["Pizza"],
  "locations": ["London"],
  "searchLimit": 40,
  "maxItems": 35,
  "maxImages": 20
}

The result count can be lower than the search limit. Search result counts change, duplicates are removed, and a profile can become unavailable between discovery and collection. maxItems caps emitted businesses, while searchLimit caps the number of candidates to process for each search pair. Set searchLimit above your required final count when you expect duplicates or unavailable profiles.

Output example

The following item is from the completed London pizza run. Opening hours and photos are shortened here; the dataset item contains the full lists:

{
  "name": "Homeslice Neal's Yard",
  "bizId": "3AqrHgpIbcMXj0q-p_oPzg",
  "directUrl": "https://www.yelp.com/biz/homeslice-neals-yard-london",
  "alternateNames": [],
  "address": {
    "addressLine1": "13 Neal's Yard",
    "addressLine2": "Covent Garden",
    "addressLine3": "",
    "city": "London",
    "regionCode": "XGL",
    "postalCode": "WC2H 9DP",
    "country": "GB"
  },
  "latitude": 51.514438,
  "longitude": -0.126427,
  "phone": "020 3151 7488",
  "website": "http://www.homeslicepizza.co.uk/",
  "aggregatedRating": 4.4,
  "reviewCount": 291,
  "priceRange": "££",
  "type": "Restaurants",
  "cuisine": "Pizza",
  "categories": ["Restaurants", "Pizza"],
  "operationHours": [
    {
      "hours": ["12:00 PM - 11:00 PM"],
      "regularHours": ["12:00 PM - 11:00 PM"],
      "regularHoursRaw": [[720, 1380]],
      "hasSpecialHours": false,
      "dayOfWeekShort": "Mon"
    }
  ],
  "upcomingSpecialHours": [],
  "amenitiesAndMore": ["Offers take-out", "Vegan options", "Wheelchair accessible"],
  "businessHighlights": [],
  "aboutTheBusiness": {
    "specialties": null,
    "history": "Homeslice began with a hand-built wood fire oven...",
    "yearEstablished": "2011"
  },
  "healthScore": null,
  "claimed": true,
  "verified": false,
  "advertiser": false,
  "isClosed": false,
  "primaryPhoto": "https://s3-media0.fl.yelpcdn.com/bphoto/G2yVWPacPsc3qkxlDHkaMg/l.jpg",
  "images": [
    {
      "link": "https://s3-media0.fl.yelpcdn.com/bphoto/G2yVWPacPsc3qkxlDHkaMg/o.jpg",
      "alt": "Salami, rocket, parmesan"
    }
  ],
  "summary": null,
  "menuUrl": "http://www.homeslicepizza.co.uk/menu/",
  "scrapeStartedAt": "2026-09-26T11:54:31.804Z",
  "scrapeFinishedAt": "2026-09-26T11:54:33.733Z"
}

The timestamps identify when that business was fetched, not when Yelp last edited its listing. Ratings, reviews, prices, hours, and contact details can change later. Use bizId or directUrl when comparing two runs; business names alone are not unique.

Use cases

• Local market research: Compare the number, locations, categories, prices, and ratings of businesses in a neighborhood. Search the same category across several cities to build a consistent comparison.
• Sales prospecting: Export public business websites, phone numbers, and addresses for businesses in an area. Review each record before outreach and follow the rules that apply to your use.
• Competitive monitoring: Schedule the same input regularly, then compare reviewCount, aggregatedRating, hours, closed status, and advertised amenities by bizId.
• Location planning: Map public latitude and longitude values where present and combine them with category and price data to spot clusters or gaps.
• Directory enrichment: Refresh a directory of known businesses with direct URLs, contact links, opening hours, menu URLs, and current categories.
• Hospitality research: Compare restaurant cuisine, service options, photo coverage, and customer rating totals across neighborhoods without opening every profile manually.

How much does it cost to scrape Yelp?

The current result price is $0.001 per business, or $1.00 per 1,000 dataset items. The event fires when a business is written to the dataset. Search candidates that do not become output items do not create a result event. Platform usage is included in the result price; there is no separate photo event in this actor.

Example	Calculation	Result charge
100 businesses	100 × $0.001	$0.10
1,000 businesses	1,000 × $0.001	$1.00
5,000 businesses	5,000 × $0.001	$5.00

These examples describe output pricing, not a guarantee that one search can find that many businesses. The completed verification run emitted 35 businesses, corresponding to $0.035 in result events at this price. A 5,000-result project needs many distinct term and location pairs or direct URLs because Yelp exposes a finite number of results per search.

Raising maxImages above the photos carried on the business page adds gallery requests and may make the run take longer. It does not add a separate event charge. When keeping cost and run time predictable matters most, begin with one term, one location, a modest searchLimit, and the default photo setting. Check the actual dataset size before scaling up.

Tips for best results

Use specific terms that match how businesses describe themselves. A broad term such as Restaurants can produce a different mix from Pizza, even in the same city. Search terms and locations are combined as a Cartesian product, so large lists can create many page requests and overlapping results. Start narrow, inspect the first run, then add terms or nearby locations to cover more ground.

For direct collection, paste the complete URL of each known business. The scraper follows the business alias and uses the canonical page URL in output. You can mix direct URLs with searches to cover important businesses that a given search ranking does not show. Repeated URLs in the same run are deduplicated. When joining runs, deduplicate by bizId.

Use maxItems when you need a strict overall dataset cap. Use searchLimit when you need a per-query cap; it does not force Yelp to return that many listings. Set maxImages to zero when you need only business metadata, or keep it at 20 for a representative visual sample. More photos require additional gallery pages and can increase elapsed time for each business.

For periodic monitoring, save your original input and run it on the same schedule each time. Compare results using IDs, and keep the run timestamp alongside each snapshot. A missing business in a later search may have moved in ranking rather than closed; use a direct URL for profiles that must always be checked.

Limits

Yelp reports up to roughly 240 results for a single search query in the tested search response. This actor caps pagination at 24 pages of ten results. Query wording and location strongly affect which businesses appear inside that window. For broader coverage, split a large area into smaller locations or use related but distinct search terms. The result limit is a cap on processed candidates, not a promise of that many complete profiles.

Some search responses are incomplete or include sponsored results before ordinary listings. The scraper reads complete business records that are present and skips links that are not direct business pages. If a search page or business page is blocked or unavailable, the run can return fewer items than requested. Access behavior can vary by location and over time.

Optional fields reflect what the business profile publishes. A website, menu, history, inspection score, special hours, or highlights may be absent. Some regions do not display public health inspection data. claimed is null when the page does not provide a claim status. verified describes the displayed owner verification status and is separate from license information. Coordinates are included when present in the page data or map link.

Opening hours and amenities are published business information and may be outdated. A photo caption can be empty, and the number of usable photos can be lower than maxImages if the gallery has fewer images or mixes photos with videos. This actor returns business summaries and rating totals, not individual customer reviews or reviewer details.

The completed verification run searched Pizza in London, requested up to 40 profiles, and returned 35 unique businesses. All 35 had a name, ID, direct URL, address object, rating, review count, categories, and at least one photo link. That observation demonstrates the tested input, not a permanent success rate for every city or query.

Use in automated workflows

The input and output are structured for repeatable collection. Two example requests for an automated workflow are:

Find up to 40 pizza businesses in London. Return their names, ratings, review totals, public websites, and direct URLs.

Recheck these five Yelp business URLs weekly. Flag any change in opening hours, closed status, price indicator, or review count.

The first request maps to searchTerms, locations, and searchLimit; the second maps to directUrls and a recurring schedule. Keep business IDs in your downstream table so changed names and URLs do not create duplicate records.

FAQ

Is it legal to scrape Yelp business data?

The actor collects public business profile information. Review Yelp's terms and the rules that apply to your location and intended use. If you combine the data with personal information, assess privacy obligations before collecting or sharing it. This actor does not collect individual review text or reviewer profiles.

Do I need to provide proxies?

The default input has proxy routing enabled. You can change proxyConfiguration if your run requires different routing. Access restrictions can still affect a particular query or page; a proxy setting cannot guarantee access to every profile.

How fast is a run?

Time depends on search page count, the number and size of business profiles, photo limits, and temporary blocks. In the completed London verification, 35 business items were delivered in about two minutes. Treat that as an observed example rather than a fixed speed for other inputs.

Can I schedule and monitor repeated runs?

Yes. Save the input, run it on a schedule, and compare dataset items by bizId. Keep scrapeStartedAt or the run time with each snapshot. For important known businesses, use direct URLs so a shift in search ranking does not hide them.

Can I export to a spreadsheet?

Yes. The dataset can be downloaded as CSV or JSON. The address, operationHours, aboutTheBusiness, and images fields contain nested data; flatten those objects if your spreadsheet workflow requires one value per column.

Why is a field empty?

Yelp may not publish that fact for the business or region. For example, inspection scores are regional, menu links depend on the business, and owner-written history may never have been added. A missing value is preserved as null or an empty list; it is not guessed from another source.

Why did a search return fewer businesses than searchLimit?

The limit is a maximum. Yelp may expose fewer matches, repeat a business in several searches, or refuse access to a page temporarily. Some response pages include only the complete records available before the response ends. Try a more specific location, another relevant term, or direct URLs for profiles you must include.

Integrations

Use the dataset through the platform API or connect it to webhooks, scheduled jobs, spreadsheets, and workflow automation. The stable bizId and directUrl fields make it practical to join results with an existing directory. For change tracking, store one row per business and collection date, then compare rating totals, contact details, hours, and closed status over time.

Support

Open an issue on the Issues tab; we reply within 24h and add fields on request.