Homely.com.au Scraper - Buy, Rent, Sold, Agents, Suburb Reviews & Q&A

Scrape everything that matters on Homely.com.au, Australia's community-driven property portal - in one actor:

• 🏠 Buy - properties for sale
• 🔑 Rent - properties for rent
• 💰 Sold - sold/completed listings with sold dates
• 👤 Agent Finder - agents and offices/agencies, with contact details
• 🌟 Suburb Review - community suburb ratings, pros/cons and full reviews
• ❓ Questions - community Q&A for any suburb

Search by suburb name, or paste exact Homely URLs. Datacenter-friendly, with optional one-click export into Notion, Linear, Airtable or Apify via MCP connectors.

---

What you get

Mode	Records	Highlights
Buy / Rent / Sold	one per listing	price (+ parsed numeric value), beds/baths/cars, property type, full address + GPS, agent names & phones, office, all photos, inspections, auction, days on market; Rent adds bond + available date; Sold adds sold date + sold price
Agents	one per agent	name, profile URL, email + mobile + office phone, bio, agency, tier, follower count, review summary, listing counts (buy/rent/sold/leased), transactions
Agents → Offices	one per office	office name, agency/company, review summary, listing counts, transactions
Suburb	one per suburb	score, rating, ranking, "great for"/"not great for", who lives here, pros/cons, all reviews inline (title, body, rating, pros/cons, helpful count, author, comments), best/other nearby suburbs, popular questions
Questions	one per question	title, author, reply count, follower count, suburb, question URL

Turn on Fetch full details to enrich listings with agent email, full description, council rates and floor plans, and agents with job title, website, areas of expertise and performance stats.

---

Input

The input has two independent questions. First how you supply the input, then what you want back:

Field	Applies to	Description
1. Input method	all	search (tell us a suburb) or url (paste Homely links). In URL mode the link decides what is scraped, so the dataset picker and every search filter are ignored
2. What to scrape	search	buy, rent, sold, agents, offices, suburb, or questions
Location	search	Suburb to scrape, e.g. Bondi NSW 2026, Bondi, NSW, or a slug bondi-nsw-2026. State + postcode disambiguate same-named suburbs
Min/Max price, beds, baths, car spaces	buy/rent/sold	Client-side listing filters (price = weekly rent for Rent)
Property types	buy/rent/sold	house, unit, apartment, townhouse, villa, land, etc.
Required features	buy/rent/sold	Keep only listings whose tags include all of these
Sort agents by	agents/offices	Agent finder ordering: rating, reviews, sales, listings. Listings are returned in Homely's default order
Agent specialty	agents/offices	sales or rentals agents
Office ID	agents	List all agents in a specific office
Homely URLs	url	Paste for-sale / for-rent / sold / suburb-profile / questions / listing / agent URLs
Fetch full details	all	Add a detail request per record for extra fields
Add suburb insights to each listing	buy/rent/sold	Attach the suburb's rating, score, pros/cons and "great for" tags to every property. One extra request per suburb, not per listing. Off by default
Max records	all	Run cap (default 20, 0 = unlimited)
Max pages per location	all	Optional per-location page limit
Resume from a previous run	all	Continue ONE interrupted crawl (run ID or dataset ID); skips records already collected there
Incremental changes for scheduled runs	all	Daily/recurring monitoring of the SAME search — see below
Proxy	all	Default datacenter works on all plans
MCP connectors	all	Optionally pipe a summary of each record into Notion / Linear / Airtable / Apify

Example - properties for sale in a suburb

{
  "mode": "search",
  "dataset": "buy",
  "location": "Bondi NSW 2026",
  "bedroomsMin": 2,
  "priceMax": 5000000,
  "maxItems": 100,
  "proxy": { "useApifyProxy": true }
}

Example - agent finder with contact details

{
  "mode": "search",
  "dataset": "agents",
  "location": "Surry Hills NSW 2010",
  "agentSpecialty": "sale",
  "detail": true,
  "maxItems": 50
}

Example - suburb reviews

{ "mode": "search", "dataset": "suburb", "location": "Bondi NSW 2026" }

---

Resume & recurring updates

Two distinct features, both off by default:

• Resume from a previous run — continue ONE specific interrupted crawl. Paste the run ID (or dataset ID) and this run skips every record it already collected, so a large pull can be split across runs without duplicates or double charges.
• Incremental mode — turn this on when you schedule the same search to run again and again (daily, weekly). The actor remembers the previous run's records itself (no run ID to paste) and adds changeType (NEW / UPDATED / UNCHANGED / REAPPEARED / EXPIRED), changedFields, firstSeenAt and lastSeenAt to every record. By default only NEW / UPDATED / REAPPEARED records are returned (and billed) — turn on Emit unchanged or Emit expired to also get those (extra billed rows). State is scoped separately per search/filter/detail-toggle combination; set State key to name or deliberately share a monitoring campaign. EXPIRED is only produced after a run that fully scanned the tracked search — never when Max records capped it or when Resume was used, since a partial scan cannot tell "gone" apart from "not reached yet".

Excluded from change detection (kept in the output, just never drives changeType): suburbInsights on Buy/Rent/Sold records is a suburb-level aggregate shared across every listing in that suburb, not the listing's own data — including it would flip every listing in a suburb to UPDATED whenever the aggregate ticks (and it can also go briefly unavailable on a single fetch, which would otherwise look like a spurious change). daysOnMarket, daysOnHomely and isNew are excluded too: Homely's own "days on market" counters advance by one every single day regardless of whether anything about the listing changed, so keeping them in the fingerprint would mark every listing UPDATED on every day's incremental run and suppression would never fire.

Notes

• A suburb name with state + postcode (e.g. Bondi NSW 2026) resolves most reliably; bare names pick the best match by state/postcode.
• Listing filters are applied to the fetched results, so a tightly filtered search may read several pages to fill Max records.
• Leased rentals are not exposed publicly by Homely; use Sold for sales history.
• This actor returns publicly visible information only. You are responsible for how you use scraped data, including any personal data of agents (GDPR/CCPA and Homely's terms).

If you have a saved task from before July 2026

The input used to be a single Mode field holding buy, rent, sold, agents, suburb, questions or url. That form is no longer accepted. Update saved tasks like this:

old	new
{"mode": "buy"}	{"mode": "search", "dataset": "buy"}
{"mode": "rent"}	{"mode": "search", "dataset": "rent"}
{"mode": "sold"}	{"mode": "search", "dataset": "sold"}
{"mode": "agents"}	{"mode": "search", "dataset": "agents"}
{"mode": "agents", "agentSubMode": "offices"}	{"mode": "search", "dataset": "offices"}
{"mode": "suburb"}	{"mode": "search", "dataset": "suburb"}
{"mode": "questions"}	{"mode": "search", "dataset": "questions"}
{"mode": "url"}	{"mode": "url"} (unchanged)

Everything else - location, filters, limits, proxy - is unchanged.