Actor start# Construction Permit Leads — Trade-Scored Building Permit Monitor Get **new building permits, classified by trade and scored as sales leads**, delivered every morning. Built for roofers, HVAC and plumbing contractors, foundation companies, restoration crews, solar installers, and the suppliers who sell to them. A permit is a signed statement that someone is about to spend money on construction. This actor finds those statements the day they are filed. --- ## Why this instead of a raw permit dump Most permit tools hand you a spreadsheet of every record the city published and leave you to sort it. That is not a lead list, it is homework. This actor does four things to each permit before you see it: | | | |---|---| | **Classifies the trade** | Reads the work description and tags it `roofing`, `hvac`, `plumbing`, `electrical`, `foundation`, `remodel`, `new_construction`, `solar`, `pool`, `demolition`, `fence_deck`. Handles the messy real strings — `REPLACE ROOF STORM DAMAGE`, `RE-ROOF`, `ROOF OVER EXISTING` all resolve to roofing. | | **Scores the lead 0–100** | Weighted heavily toward **freshness**, because a three-day-old permit already has four competitors calling it. Job value, whether a contractor is already attached, and address quality also factor in. | | **Filters to your business** | Your trades, your ZIP codes, your minimum job size. | | **Remembers what it already sent you** | Run it daily and you get *only today's new permits* — never the same row twice. | Sort by `lead_score` descending and work top-down. That is the whole workflow. --- ## Coverage Every city below was queried live on 2026-07-28 and confirmed to be returning **current** data. A city is not listed until it passes that check. | City | Source | Freshness | Valuation | Contractor | Site ZIP | |---|---|---|---|---|---| | **Houston, TX** | Permit portal | Same/next day | Yes | — | Yes | | **Chicago, IL** | Open data | 1 day | Yes | Yes | **No** | | **New York City, NY** | Open data | 1 day | Yes | Yes | Yes | | **Austin, TX** | Open data | 1 day | **Estimated** | Yes | Yes | | **Seattle, WA** | Open data | 3 days | Yes | Sparse | Yes | | **Los Angeles, CA** | Open data | 3 days | Yes | — | Yes | Three caveats worth knowing before you buy: - **Austin publishes no valuation column at all.** Every Austin row carries a trade-median estimate flagged `valuation_is_estimated: true`. Your minimum-valuation filter never discards a row on an estimated value, so you do not lose real leads to a blank field. - **Chicago publishes no job-site ZIP.** It publishes contact ZIPs, which are the *contact's* address, not the job. Mapping those would silently mis-target your ZIP filter, so the field is left empty and Chicago rows are excluded when a ZIP filter is set. - **Houston has no data feed whatsoever.** The City stopped publishing Permit Activity Reports on 2025-12-01 and now exposes permits only through an interactive search. That is exactly why Houston data is hard to find anywhere else. A note on why this list is short: many well-known "open permit datasets" are frozen archives. Dallas's endpoint returns HTTP 200 and looks healthy — its newest record is from December 2019. San Francisco's building-permits feed is a contacts table with no addresses or descriptions. Neither is listed here, because a run that finds nothing is worse than a city that is missing. ## Output One row per permit: ```json { "lead_score": 78, "trade": "roofing", "issued_date": "2026-07-27", "valuation": 28000, "valuation_is_estimated": false, "address": "1234 Beechnut St, Houston TX 77072", "zip": "77072", "work_description": "REPLACE ROOF STORM DAMAGE", "permit_type": "Building Permit", "permit_number": "26-0891234", "contractor_name": "", "owner_name": "SMITH JOHN", "source_city": "houston_tx", "source_url": "https://www.houstonpermittingcenter.org/sold-permits-search", "scraped_at": "2026-07-27T13:04:11Z" } ``` Export to CSV or Excel, or push straight into your CRM via the Apify API. --- ## The call-today sheet Every run also writes **`hot-leads.txt`** — a plain-text call sheet containing only the leads that scored at or above your threshold (75 by default), best first. You do not have to open the dataset to use this product. Open one text file and start dialing. ``` ============================================================================== CALL TODAY — permits scoring 75+ generated 2026-07-28 ============================================================================== 53 lead(s) worth a call, out of 166 total this run. Cities: New York City, NY (34), Chicago, IL (19) Trades: Remodel (22), HVAC (7), New Construction (5), Roofing (3) ... ------------------------------------------------------------------------------ #1 SCORE 90 NEW CONSTRUCTION Chicago, IL ------------------------------------------------------------------------------ ADDRESS : 310 W HURON ST CONTRACTOR: 310 HURON OWNER LLC VALUE : $300,000 ISSUED : 2026-07-27 WORK : SPR SPECIAL CONSTRUCTION - INSTALL SKIP HOIST FROM SIDEWALK LEVEL TO LEVEL 9 AS PER PLANS. PERMIT : PERMIT - NEW CONSTRUCTION #101084332 WHY : large job $300,000 ``` The **WHY** line explains the score rather than asking you to trust it — whether no contractor is attached yet, how big the job is, and whether the value was estimated because the city published none. Find it in the run's **Storage → Key-value store → hot-leads.txt**. Set the threshold to 0 to skip the file entirely. ## Reading `lead_score` | Score | Meaning | |---|---| | **75–100** | Call today. Fresh, real money, often no contractor attached yet. | | **50–74** | Worth a call or a mailer this week. | | **25–49** | Marketing list, not a phone list. | | **0–24** | Old, tiny, or missing an address. | **`contractor_name` is empty** means the homeowner pulled the permit themselves and has not hired anyone. If you sell to homeowners, these are your best rows and the score reflects that. If you sell *to contractors*, flip on **Only permits with a named contractor** to invert it. --- ## Recommended setup **Roofing contractor in southwest Houston:** ``` cities: houston_tx lookbackDays: 2 trades: roofing minValuation: 8000 zipWhitelist: 77072, 77036, 77081, 77074, 77099 requireContractor: false onlyNew: true ``` Then **Schedule** it daily at 6:00 AM. You get that morning's new roofing permits in your inbox before you leave the yard. --- ## Pricing Pay per event. You are charged for **new leads actually delivered** — not for rows scraped, not for duplicates you already received, not for permits your filters excluded. **Worked example.** A Houston roofer running daily with the setup above typically sees 8–25 qualifying new roofing permits per day. At the per-lead event price shown on this listing, a full month of daily runs bills for roughly 240–750 delivered leads. If your average roof job is $14,000, the arithmetic is not close. Set a **max cost per run** in the Console and the actor stops cleanly when it reaches your cap — it will never quietly overrun your budget. --- ## FAQ **Where does this data come from?** Public municipal permit records. Building permits are public records in Texas by statute. **Is this legal?** Yes. The actor reads only public, logged-out pages and never creates an account, never logs in, and never bypasses any access control. It also rate-limits itself so it does not burden city servers. **Why is Houston slower than Austin?** Austin is a direct data feed. Houston has no feed at all — the actor has to drive the city's search portal in a real browser. Expect a Houston run to take a few minutes. That is also why Houston data is hard to get anywhere else. **Some valuations look like round numbers.** Those are estimates, and they are marked `valuation_is_estimated: true`. Austin publishes no valuation field, and many cities leave it blank on trade permits. Rather than drop those leads or show `$0`, the actor fills in a trade median. Your minimum-valuation filter never discards a row on an estimated value, so you do not lose real leads to a blank field. **A permit is classified under the wrong trade.** Send the exact `work_description` and the correct trade. Classification is rule-based and deterministic, so a reported miss becomes a fix in the next release rather than a shrug. **Can you add my city?** Ask. Cities with a live open-data feed can be added quickly. Cities with only a search portal take longer but are usually the ones worth the most, because nobody else bothers.
