Daft.ie Scraper — Ireland Property Listings

Extract property listings from Daft.ie, Ireland's #1 property portal — for rent, for sale, sharing, and new homes. Returns clean structured data (price, beds, baths, BER, agent, geo) ready for CSV / Excel / JSON / API.

Built for real estate investors, proptech developers, market analysts, and agencies who need Irish property data at scale.

What you get per listing

Field	Example
address	Griffith Wood, Griffith Avenue, Drumcondra, Dublin 9
price / priceUnit	3141 / month
beds / bedsMin / bedsMax	1 / 1 / 2 (ranges like "1 & 2 Bed" preserved)
propertyType	Private Rental Sector
ber	A2
sellerName / sellerType	The Carriageworks / UNBRANDED_AGENT
lat / lng	53.35 / -6.23
publishDate	2026-05-13T…
url	https://www.daft.ie/for-rent/…/6306467

Detail-only fields (baths, sizeSqm, psra, priceHistory) are returned as null from the search listing and populate once the detail crawl lands (see Roadmap).

Input

{
  "categories": ["rent"],
  "counties": ["dublin-city", "cork"],
  "priceMax": 3000,
  "bedsMin": 2,
  "maxItems": 1000,
  "proxyConfiguration": { "useApifyProxy": true, "apifyProxyGroups": ["RESIDENTIAL"], "apifyProxyCountry": "IE" }
}

• categories: rent | sale | share | new-home
• counties: Daft location slugs exactly as in the site URL — dublin-city, dublin, cork, cork-city, galway, …
• priceMin/Max, bedsMin/Max: optional filters (price is monthly for rent/share, total for sale)
• maxItems: hard stop
• Each category × county is one search; the actor reads the total, then fans out all pages.

Run locally

npm install
npm start            # builds TS then runs; uses ./storage for input/output

Put test input in storage/key_value_stores/default/INPUT.json. Output lands in storage/datasets/default/.

Deploy to Apify

npm i -g apify-cli
apify login
apify push

Then in the Apify console set Monetization → Pay per result (suggested $2 / 1,000 results; you keep 80%). To use pay-per-event instead, switch the actor to PPE and uncomment the Actor.charge(...) line in src/main.ts.

How it works

Daft is a Next.js site — listing data is embedded in the page as JSON (classic #__NEXT_DATA__ or app-router self.__next_f streaming chunks), so the actor parses JSON directly (fast, cheap, no headless browser). A DOM-parsing fallback covers markup changes.

First run: confirm listings populate from JSON (source: "json" in output). If you see source: "dom" warnings, Daft changed its data shape — update the array detection in extractEmbedded/findListingsArray.

Legal / etiquette

Scrapes public listing pages only. Does not touch api.daft.ie (gated, ToS-restricted). Respects robots by avoiding blocked bot UAs; rate-limited via Apify proxy. Photos are linked, not rehosted. Public agent/business listing data → low PII exposure.

Roadmap

• Detail crawl — visit each listing for guaranteed BER, geo, floor area, full description, all photos.
• Monitor mode — diff against last run → emit new / price_drop / delisted (recurring revenue).
• Sister actors — clone skeleton to MyHome.ie (uncovered #2 IE portal), then PropertyPal (NI).