Houzz Scraper

Collect home improvement professionals by trade and location. One clean row per business with phone, full postal address, areas served, social profiles and website. No login and no API key.

What it does

• Contact details ready to use. Phone number as displayed, plus a digits-only version for joining and deduplication.
• Full postal address split into fields: street, city, state, ZIP and country, plus one combined line.
• Website and listing page kept apart. A business's page in the directory is not their own website, so it is returned separately as profileUrl and never passed off as one.
• Social profiles as their own list, separate from any other links.
• Deduplicated across every trade and location, so a business listed under two searches is delivered once and charged once.
• Many trades and locations in one run, all returning the same fields.

Common uses: lead lists for suppliers and software vendors, local market research, competitor mapping, recruitment, and building a trades directory.

Input

Field	Type	Description
trades	array	Trade slugs as they appear in the site address. See the list below.
locations	array	Locations in the site's own form, for example Austin--TX.
maxResultsPerSearch	integer	Cap per trade and location pair. Default 100.
proxyConfiguration	object	Optional. Enable to spread requests across IPs.

Trades confirmed to work: general-contractor, architect, interior-designer, landscape-architect, roofing-and-gutters, painter, plumber, electrician. Others may exist; a trade that is not recognised is reported in the log and skipped rather than returning unrelated results.

Example input

{
  "trades": ["general-contractor", "architect"],
  "locations": ["Austin--TX", "New-York--NY"],
  "maxResultsPerSearch": 300
}

Output

Each item is one professional.

{
  "name": "Acme Design & Build",
  "trade": "general-contractor",
  "phone": "+1 555-0100",
  "phoneDigits": "15550100",
  "website": "https://example.com",
  "profileUrl": "https://www.houzz.com/professionals/general-contractors/example-pf~1000000000",
  "street": "1 Example Street",
  "city": "Austin",
  "region": "TX",
  "postalCode": "78701",
  "country": "US",
  "address": "1 Example Street, Austin, TX, 78701",
  "areaServed": "Austin",
  "searchLocation": "Austin--TX",
  "socialProfiles": ["https://www.facebook.com/example"],
  "otherLinks": [],
  "imageUrl": "https://example.com/photo.jpg"
}

Notes

• The listing reorders between requests. Asking for the same page twice does not return quite the same set, so this actor does not assume a fixed number of pages. It keeps requesting until several consecutive requests turn up nothing new, and deduplicates as it goes. That is why the run log shows more requests than a simple page count would suggest.
• Because of that reordering, a second run of the same search can surface a few businesses the first run did not. Raising maxResultsPerSearch gives the run more room to find them.
• website is filled only when a business publishes its own address. Many list only social profiles, so this field is often null while profileUrl is always present.
• Some entries list a bare social handle rather than a full address. The platform cannot be determined from a handle alone, so those are left out instead of being filed under a guess.
• street is missing for a minority of businesses that do not publish one. Those rows keep every other field.
• Results for a location include nearby areas, so a search can return businesses in neighbouring states. The region field shows exactly where each one is.

Privacy

To improve our actors we collect anonymized usage telemetry (run stats and input patterns). No personal account data is collected.