Local Business Phone Scraper

Give it a trade and a city — plumber, Chicago, IL — and get back local businesses with phone, address, website, opening hours, category and a Better Business Bureau letter grade. Twenty-five columns on every row.

You are billed only for rows that carry a phone number or a website, and only for emails that were actually found. A row with no way to reach the business is free. A website that turns out to have no email address is free.

The last verification run returned 108 Chicago plumbers and dentists in 43 seconds — every one with a phone number, 17 of them with an email as well.

What it costs

Plus $0.00005 to start a run — five thousandths of a cent. No per-page fee, no proxy cost, no charge for a search that finds nothing.

The email pass is off by default and costs nothing until you switch it on. When it is on, only a found address is billed: on the measured runs about four websites in ten yield one, and the other six cost you nothing at all. Compare that with paying per row and receiving an email on some unstated fraction of them.

What one row looks like

Real output, plumber in Chicago, IL, with the email pass on:

{
  "name": "Power Plumbing & Sewer Contractor, Inc.",
  "phone": "+17732489574",
  "additionalPhones": "",
  "email": "power@powerplumbinginc.com",
  "emailsFound": "power@powerplumbinginc.com,service@powerplumbinginc.com",
  "website": "https://powerplumbinginc.com",
  "socialProfiles": "https://facebook.com/PowerPlumbingChicago,https://linkedin.com/company/power-plumbing",
  "address": "3840 N Ashland Ave",
  "city": "Chicago",
  "state": "IL",
  "postalCode": "60613-2708",
  "country": "US",
  "latitude": "41.95162582397461",
  "longitude": "-87.66901397705078",
  "category": "Plumber",
  "rating": "A+",
  "reviewCount": 218,
  "openingHours": "Mon 24h; Tue 24h; Wed 24h; Thu 24h; Fri 24h; Sat 24h; Sun 24h",
  "accredited": true,
  "sourceName": "bbb,duckduckgo",
  "sourceUrl": "https://www.bbb.org/us/il/chicago/profile/plumber/power-plumbing-sewer-contractor",
  "sourceId": "1001869",
  "searchTerm": "plumber",
  "searchLocation": "Chicago, IL"
}

The 25 columns

Every row carries all 25 in the same order, so the CSV has no ragged edges. A field the source did not publish comes back empty — never guessed, never filled with N/A.

email is the single best address; emailsFound is every address on the page, in case you want them. The best one prefers the business's own domain over a free mailbox, and a role box like info@ over a named person — info@ is answered by whoever is on duty, john@ bounces the day John leaves.

Field coverage, measured

Over 1,821 businesses: 11 trades — plumber, dentist, HVAC, roofing, electrician, law firm, landscaping, pest control, locksmith, moving, accountant — across Chicago, Austin, Phoenix, Denver, Miami, Seattle, Atlanta and Boston.

Every trade and every city in that sample returned a contact on 100% of rows. The businesses missing a street line are overwhelmingly home-based contractors who publish a phone and a service area but no premises.

Why some columns are around half. Two directories are searched, and they return largely different businesses rather than the same ones twice — measured overlap is 2.8%. So a row usually comes from one of them: a Better Business Bureau row carries the letter grade and the accreditation flag; a places-index row carries the website, the hours and the review count. Name, phone, address, city, state, postcode and coordinates come from both. Where a business appears in both, the row is merged and has everything, as in the sample above.

Because the email pass reads the website, it can only run on rows that have one — about four in ten. Select one source if you would rather have consistent columns than maximum coverage.

Input

{
  "searchTerms": ["plumber", "hvac contractor"],
  "locations": ["Chicago, IL", "Austin, TX"],
  "sources": ["bbb", "duckduckgo"],
  "maxResultsPerSearch": 100,
  "findEmails": true,
  "onlyWithContact": false
}

Terms are crossed with locations, so two terms and two cities is four searches.

A location is required, and it has to be a real place. Results are geolocated from the address the request comes from, so a search with no city in it would return businesses near the server rather than near you. The Actor refuses that input instead of handing you a tidy table of the wrong city.

What this does not return

• No personal contact details. Published business information only.
• No verification that a number is in service or an inbox is monitored. Addresses are read from the company's own website, not tested for delivery.
• No more than about 246 businesses per search — see below.
• No email for a business with no website, which is roughly six rows in ten.

Limits, stated up front

• Better Business Bureau returns at most 225 businesses per search. Its own paging stops at 15 pages of 15, even on a query that reports 4,565 matches.
• The places index returns at most 21 per search, and has no paging at all.
• So one search is not a whole city. To cover a large market, list neighbourhoods as separate locations — Lakeview, Chicago, IL, Pilsen, Chicago, IL, Wicker Park, Chicago, IL. Measured: seven neighbourhood searches returned 67 distinct plumbers where one city-wide search returned 21.
• United States and Canada are well covered. Other countries are thin.
• Coverage is what the directories publish: good for trades and professional services, thinner for very new or very small businesses.

Free plan

Runs on the Apify free plan, and the trial covers your first 100 results — enough to look at the data before paying for any of it. Every field is included in the trial, the email pass included; nothing is held back for paying users.

Free-plan runs get less memory, so keep a first run to a handful of searches.

Speed and cost of a run

108 businesses in 22 to 28 seconds over 8 requests, measured on the platform at the default 1 GB. With the email pass on, the same run took 43 seconds and read 40 websites. Across the 1,821-business sample the directory half came to 3.2 KB per business, with no directory refusing a single request.

Requests to the directories are deliberately spaced — they throttle parallel callers and forgive patient ones. Business websites are unrelated to each other and are read several at a time, which is why the email pass costs seconds rather than minutes. There is no proxy option because none is needed.

Troubleshooting

I asked for 500 and got 90. You hit the per-search caps above. Split the city into neighbourhoods, or add related trades — plumber,

emergency plumber

website is empty on many rows. Those rows came from the directory that does not publish a website. The overlap between the two sources is about 2.8%, so most rows carry one source's fields rather than both.

email is empty on many rows. Two reasons, and both are free to you: the business has no website to read, or the website does not publish an address. Turn on Also open one contact page per site to add about ten percentage points.

An email looks like a person, not the company. The picker prefers role boxes, but some businesses only publish a named address. Every address found on the page is in emailsFound so you can choose differently.

The run succeeded but returned very few rows. The trade may be thin in that town, or spelled differently there. Try the plain trade word before a long phrase. Rows that came back with no contact were not charged for.

A search in a non-US city returned little. Expected — coverage outside the United States and Canada is thin, and this page says so rather than letting you find out on a bill.

Output views

Five ready-made views: Contacts (name, phones, email, website, address, grade), Hours & reviews, Map data (coordinates), Outreach (email, socials, phone) and All fields. Export any of them to CSV, JSON or Excel, or feed them straight into another Actor.

Other scrapers we maintain

Same engine, same pricing model, same promise: flat rows, stated limits, and no charge for a row that came back empty.