scrapercity-real-estate-data

ScraperCity Real Estate Data

Overview

This skill picks the right ScraperCity real estate tool for the job, builds the exact call, prices it from the catalog before anything runs, and routes every contact by who the recipient is and how sure the match is.

Tools covered, with catalog prices:

The mistakes it prevents:

• Calling a People Finder result as if it were the owner when the returned address or name does not match the property record.
• Calling or mailing an owner who already said "stop" because that request sat in one dialer and never reached the mail house or the sender.
• Calling the same owner past the point where attempts still connect, while the caller ID picks up a spam label.
• Paying $0.25 a property (lookup plus owner contact) when a $0.05 distress list plus a $0.0199 People Finder search per owner covers the same need.

When to Use

• "Get me every agent in Scottsdale who does luxury or relocation."
• "Pull pre-foreclosures and auctions in Tampa and find the owners' phones."
• "What's this house worth and who owns it?" for one or many addresses.
• "Scrape these Crexi searches for industrial listings and broker contacts."
• "Pull BizBuySell listings for HVAC businesses in Texas with the brokers."
• "Build a motivated-seller list for my wholesaling business."

Use a different skill for:

Prerequisites

• Who the user is (agent, wholesaler, investor, lender, CRE broker, business broker) and the market they work.
• The outreach channel: mail, email, phone call. Each needs different fields and a different match bar.
• The suppression list the team already holds (past "stop" requests, wrong numbers, sold properties).
• For property work: full addresses with city, state and ZIP.
• For Crexi and BizBuySell: search URLs built on those sites with the filters set.
• Wallet balance: run check_wallet first. All tools here are pay per result on any plan.

Authoritative Foundations

Details and the full source list: references/framework-notes.md and references/contact-routing-guide.md.

Workflow

1. Identify the persona and the job

Playbooks for each persona: references/workflow-playbooks.md.

2. Set the contact plan before any pull that feeds phone or email

Answer four questions and write the answers in the plan:

1. Who is the recipient: a homeowner, an agent or broker at their business, a commercial owner entity, a lender's REO desk?
2. Which channels, in what order? Owners: mail, then calls, then email. Agents and brokers: email, then calls.
3. How will owner matches be confirmed? Address and name compared after USPS normalization; only confirmed and probable matches get calls.
4. Where does the suppression list live, and how does a "stop" from any channel reach every other tool the same day?

The channel table, match classes and reply triage are in references/contact-routing-guide.md.

3. Build the call and price it

Use the catalog prices only. Write units times price for each line. Show the table and wait for a go-ahead before any paid run. Run check_wallet and compare the balance to the total.

4. Run, poll, download

Every tool here is a run: start returns runId, then check_run_status until SUCCEEDED, then download_results. These finish in minutes; inside a conversation, check once per turn and report the status. CLI: npx scrapercity poll <runId> then npx scrapercity download <runId> -o <file>.csv.

5. Chain owner contact

Distress List rows carry the property address and distress type. To reach the owner, send the property address to find_people as street_citystatezip. Set max_results: 1 to cap spend: a search with no match is still billed as one result ($0.0199). For a full valuation plus owner name, phone and email in one call, scrape_property_lookup with includeOwnerContact: true costs $0.25 per property.

6. Classify, suppress and route

Before a call list goes to anyone: give every owner row a match_class (confirmed, probable, unconfirmed), remove everything on the suppression list, and send unconfirmed rows to mail only. Validate every email; personal addresses from People Finder go to a separate campaign so their bounce and complaint rates are read on their own.

7. Hand off

Agent lists go to email sequencing. Owner lists go to mail first, then manual calls to confirmed matches. Crexi and BizBuySell broker lists are business contacts and go to email and calls to the firm line.

ScraperCity Implementation

Zillow Agents

{"location": "Scottsdale, AZ", "specialty": "buyer", "language": "Spanish", "searchLimit": 200}

CLI: npx scrapercity zillow-agents -l "Scottsdale, AZ" --specialty buyer --language Spanish --limit 200. The CLI default limit is 10; set it. specialty is a code string; the CLI usage shows buyer, and ScraperCity's product page lists 40+ specialties such as buyer's agent, listing agent, relocation, foreclosure and short sale. Probe an unfamiliar code with searchLimit: 5 before a full run.

Property Lookup

{"addresses": ["4410 W Linda Ln, Tampa, FL 33629"], "includeOwnerContact": true, "fileName": "tampa-owner-check"}

Addresses must include street, city, state and ZIP. The CLI in package 1.0.17 has no property lookup command; use the MCP tool or POST /api/v1/scrape/property-lookup.

Distress List

{"search": "Tampa, FL", "distressTypes": ["preForeclosure", "auction"], "maxItems": 300}

distressTypes values: preForeclosure, foreclosure, auction, bankOwned. search is a city and state or a ZIP. maxItems is 1-500, so a larger market runs as several ZIP searches. CLI: npx scrapercity distress-list "Tampa, FL" --types preForeclosure,auction --max 300.

Crexi

{"startUrls": ["PASTE_CREXI_PROPERTY_SEARCH_URL", "https://www.crexi.com/resources/find-a-broker/PASTE_REST_OF_BROKER_URL"]}

Replace the two PASTE values with URLs copied from the browser. A property search starts https://www.crexi.com/properties?, a city or market search starts https://www.crexi.com/search?. Set filters on Crexi (property type, price, cap rate) after choosing a city or market in the search bar, then copy the URL. Broker directory URLs (https://www.crexi.com/resources/find-a-broker/...) return brokers. A panned map with no city or market returns 400. CLI: npx scrapercity crexi "<url>".

BizBuySell

{"startUrls": ["https://www.bizbuysell.com/texas/hvac-businesses-for-sale/"], "maxItems": 150, "datasetName": "tx-hvac"}

Use the same datasetName across runs to dedupe. CLI: npx scrapercity bizbuysell "<url>" --limit 150.

People Finder

{"street_citystatezip": ["4410 W Linda Ln, Tampa, FL 33629"], "max_results": 1}

CLI: npx scrapercity people-find --address "4410 W Linda Ln, Tampa, FL 33629"; separate several addresses with |.

Worked cost estimate: wholesaler in Tampa

The alternative of scrape_property_lookup with owner contact on all 300 would cost 300 x $0.25 = $75.00. Mail goes to all 300 property addresses; calls go only to the People Finder results that pass the match test.

Output Format

A real estate data plan in Markdown from templates/output-template.md:

1. Objective: persona, market, outcome.
2. Tool Plan: one row per tool call with tool id, key parameters, units.
3. Cost Estimate: units x catalog price per line and a total.
4. Contact Hygiene: one row per channel with the rule and its operating reason, plus the match test and where the suppression list lives.
5. Outreach Routing: which records go to which channel, by match class.
6. Next Steps.

Run python3 scripts/check-output.py plan.md before handing it over.

• Each fact once; every call and script written out in full (no "..." values, no loop described in a comment); no counts or results marked assumed. If nothing has run, say so once at the top.
• What the reader receives is the deliverable only: no internal file paths, reference section numbers or runbook IDs; another skill is named only in a handoff line that tells the user what to run next; no tables of checks that were not run (one "Not checked:" line instead); and no repeated identical rows where one template plus a list will do.

Quality Check

• Every tool id is one of the six above, and every unit price matches the catalog.
• Every Property Lookup and People Finder address has a ZIP code.
• Every distressTypes value is one of preForeclosure, foreclosure, auction, bankOwned.
• maxItems and searchLimit are 500 or less.
• The cost total equals the sum of the lines, and the user approved it before running.
• Every outreach channel has a Contact Hygiene row with a rule and its operating reason.
• Calls go only to confirmed or probable matches; unconfirmed rows are mail only.
• The suppression list is named, applied before each drop and call block, and fed by every channel.
• The plan contains no legal commentary.

Common Pitfalls

Execution Artifacts

Related Skills

• skip-tracer: owner contact at volume from a list you already have.
• mobile-appender: mobiles for B2B contacts.
• cold-calling: scripts and cadence once the list is matched and suppressed.
• scrapercity-local-business-database: brokerages, title companies and property managers as local businesses.
• scrapercity-lead-database: executives at real estate and lending firms.
• local-business-list, scrapercity-mcp, scrapercity-api, using-scrapercity-skills.