Yelp Business Search & Review Scraper

Yelp Business Profiles & Review History Extractor

Yelp packs a lot into each business page: star ratings, price signals, weekly hours, photo galleries, contact details, and years of customer reviews. This actor turns that material into structured business records, with the review feed paged well beyond the handful of reviews Yelp shows by default.

What this actor does

• Finds businesses by keyword and location, or by a Yelp search URL that already contains filters.
• Looks up exact businesses from a Yelp business URL, a bare alias such as prince-street-pizza-new-york-2, or a 22-character Yelp business ID.
• Pages each business's review feed up to your chosen limit, including reviewer names, Elite years, photos, reactions, and owner replies.
• Adds search metadata so you can tell organic ranking from sponsored placement and remove duplicate cards.
• Omits empty fields at every level. A business that does not publish a website, price level, or owner reply simply has no such key in its record.
• Writes an explicit status record when a business is missing, a page is blocked, or Yelp changes the data contract, so an empty dataset is never mistaken for a successful scrape.

Output per business

Each business is one dataset row. Reviews are nested under reviews.

Field	Description
recordType	business for a business row, status for a run status row
businessId	Yelp's 22-character business ID
encId	The same stable Yelp business ID, kept as a separate field for joins
alias	Yelp URL slug, e.g. prince-street-pizza-new-york-2
name	Business name
categories	Array of Yelp category titles
address	Object with streetAddress, addressLocality, addressRegion, postalCode, addressCountry, and formatted when supplied
city	City name
state	State or region code
zip	Postal code
country	Two-letter country code
lat	Latitude
lng	Longitude
rating	Average star rating (number)
reviewCount	Review total shown on the business profile
reviewsTotal	Reviews available in the public review feed, when it differs from reviewCount
price	Price level such as $ or $$
phone	Business phone number
website	Business website URL
hours	Weekly hours keyed by short day name, each value an array of time ranges
photos	Business photo URLs, including the primary photo and gallery
url	Yelp business page URL
isClosed	Whether Yelp marks the business as closed
isClaimed	Whether the business profile has been claimed
description	Short business summary from Yelp
reviews	Array of review objects, omitted when reviews are off or none are available
searchRank	Organic position in the search results, when found through search
sponsored	true when the business first appeared as a sponsored card
searchQuery	Keyword that found the business
searchLocation	Location used for that search
searchCategory	Category filter used for that search
sourceUrl	Search page or business URL the record came from
scrapedAt	ISO timestamp for the record

Review objects contain:

Field	Description
id	Review ID
rating	Star rating (1–5)
text	Full review text
author	Reviewer display name
authorId	Reviewer ID
authorUrl	Reviewer profile URL
authorLocation	Reviewer location as shown by Yelp
authorReviewCount	Number of reviews on the reviewer's profile
date	Review date in UTC
dateLocal	Review date in the business's local time
photos	Review photo URLs
ownerReply	Object with text, author, and date when the business replied
elite	Whether the reviewer has an Elite year
eliteYear	The Elite year Yelp displays
language	Review language code
isFirstReview	Whether Yelp marks this as the reviewer's first review
checkInCount	Check-in count attached to the review
reactions	Object with useful, funny, and cool counts

Status records

When something prevents a business or search page from being collected, the dataset receives a row with recordType: "status":

• status is not_found, blocked, api_drift, no_results, or invalid_input.
• message explains what happened.
• sourceUrl and businessId are included when known.
• scrapedAt records when the status was written.

Legacy compatibility keys are also present when they have values: directUrl (same as url), bizId (same as alias), aggregatedRating (rating as text), priceRange (same as price), and images (same as photos). Review rows also keep isUsefulCount, isFunnyCount, and isCoolCount beside the reactions object.

Status rows are separate from business rows, so filter on recordType before loading the business records.

Input

Field	Type	Default	Description
mode	string	search	search discovers businesses by keyword, category, or Yelp search URL; direct scrapes business URLs, aliases, or business IDs
searchQuery	string	pizza	Keyword for mode=search
location	string	New York, NY	City, neighborhood, or ZIP for mode=search
startUrls	array	[]	Yelp search URLs in mode=search, or business URLs, aliases, and IDs in mode=direct
categories	array	[]	Yelp category aliases or names used to narrow a search
sortBy	string	recommended	Search ordering: recommended, rating, or reviewCount
price	array	[]	Price levels $, $$, $$$, $$$$; empty keeps all
minRating	number	0	Drop businesses rated below this value (0–5)
maxItems	integer	50	Maximum business records (1–500)
maxReviewsPerBusiness	integer	20	Maximum reviews per business (1–1000)
includeReviews	boolean	true	Turn off to skip review collection
reviewSort	string	DATE_DESC	DATE_DESC, DATE_ASC, RATING_DESC, RATING_ASC, RELEVANCE_DESC, or ELITES_DESC
reviewRatings	array	["5","4","3","2","1"]	Star ratings to include
reviewLanguage	string	en	Review language code, or all to page every language Yelp reports
proxyConfiguration	object	Apify Proxy AUTO	Optional routing choice; the actor automatically uses a working route when Yelp blocks a request

mode is the only required field. mode=search needs a query, category, or search URL. mode=direct needs at least one business URL, alias, or ID.

Example: keyword search

{
  "mode": "search",
  "searchQuery": "ramen",
  "location": "Seattle, WA",
  "price": ["$$", "$$$"],
  "minRating": 4,
  "maxItems": 20,
  "maxReviewsPerBusiness": 30,
  "reviewSort": "DATE_DESC",
  "reviewLanguage": "en"
}

Example: specific businesses by URL

{
  "mode": "direct",
  "startUrls": [
    "https://www.yelp.com/biz/prince-street-pizza-new-york-2",
    "https://www.yelp.com/biz/bar-lento-new-york"
  ],
  "maxItems": 2,
  "maxReviewsPerBusiness": 100,
  "reviewSort": "RATING_DESC",
  "reviewRatings": ["5", "4"]
}

Example: businesses and reviews in another language

{
  "mode": "direct",
  "startUrls": ["https://www.yelp.com/biz/bar-lento-new-york"],
  "maxReviewsPerBusiness": 50,
  "reviewLanguage": "all"
}

Example: profiles only

{
  "mode": "search",
  "searchQuery": "coffee",
  "location": "Portland, OR",
  "categories": ["coffee"],
  "includeReviews": false,
  "maxItems": 10
}

Use cases

• Competitive benchmarking — compare ratings, review volume, price levels, and review themes across a set of local businesses.
• Review monitoring — schedule runs against your own locations and watch new reviews, owner replies, and rating changes over time.
• Lead research — collect phone numbers, websites, cities, and categories for businesses in a target market.
• Voice-of-customer analysis — export hundreds of reviews per business with ratings, dates, languages, and reviewer context.
• Location planning — map business density and quality across ZIP codes, neighborhoods, or categories.
• Data enrichment — join Yelp business IDs, aliases, websites, and coordinates to an existing CRM or dataset.

FAQ

Do I need to supply a proxy? No. The actor starts with a direct request and automatically routes through Apify Proxy when Yelp blocks it. Leave proxyConfiguration at its default, or supply a country or custom proxy only when you need a specific route.

How many businesses can one search return? Yelp caps a single search at 240 businesses, displayed 10 at a time. maxItems can be as high as 500 across multiple queries or direct URLs, but each individual search stops at Yelp's 240-result ceiling.

How many reviews can I collect per business? Up to 1,000. The review feed is paged at 49 reviews per request, so a deep run uses several requests per business. Set maxReviewsPerBusiness lower for faster, cheaper runs.

Why is reviewsTotal sometimes lower than reviewCount? Yelp's profile count and its public review feed are not identical. The feed leaves out reviews Yelp does not recommend, so reviewsTotal can be slightly lower than the number shown on the page.

What does reviewLanguage: "all" do? The actor reads the language counts Yelp reports for the business and pages each language until the review limit is reached, de-duplicating by review ID. It can use several requests per business.

Why is a business or review field missing? Yelp does not publish every field for every business. Missing values are omitted rather than written as null, empty strings, or empty arrays.

What happens when Yelp changes its data format? The actor writes an api_drift status row with a message and the source URL, and the run status message reports the problem. It does not silently return an empty dataset.

Can I use a Yelp search URL with filters already in it? Yes. Put it in startUrls with mode=search; the actor keeps the URL's query parameters and pages from its existing start position.

Does this actor collect the reviews Yelp does not recommend? No. It collects the public review feed. Reviews outside that feed are not exposed, which is why the feed total can differ from the profile count.

Is this affiliated with Yelp? No. This is an independent third-party actor and is not endorsed by or affiliated with Yelp.