French Property Intel · Value, Yield & Owner SIREN

Paste a French street address. Get one dossier per line: estimated value, deal score vs asking, DPE class and rental calendar, yield, flood/clay risks, and the SCI / property-company SIREN at that number.

No login. No API key. Open data only. No invented emails.

This is the address upsell after a city listing Actor. Those Actors find a verified house number. This one scores that number (DPE, hazards, value, yield). You are billed per dossier, not per cadastral row.

Already have a city list? Copy address from France Pool-Ready Plots, Garden Plots, Building Roofs, Building Heights, Industrial Buildings or Pools & Tennis.

Who is this for?

You are…	Typical goal	Suggested setup
Private investor	Am I overpaying this listing?	Number + street + postcode + city · paste asking price
Buy-to-let / yield hunter	Does this rent cover the price?	Add monthly rent of this dwelling · owner charges = 0 on a house
Agency / hunter	Qualify a lead before the visit	Address only · Investor view
Asset manager	Refresh a small portfolio	Extra addresses JSON · Max 25
Renovation / energy sales	Is this a passoire, and who owns the SCI?	Keep energy class + owner companies on
Data / ops	Feed a CRM from listings you already have	API run · Full dataset

Standby lookup (HTTP)

Console Start is unchanged: paste street, postcode and city, run, download the dataset.

Need one address dossier (or a few) in n8n, Make, or your own app? Use the Standby tab hostname. Authenticate with Authorization: Bearer <APIFY_TOKEN> (or ?token=). This Actor does not crawl listing websites — pass the French address (propertyUrl is treated as an address string).

Method	Path	Request	Response
GET / POST	/lookup	street + postalCode + city, or address / propertyUrl as an address string (max 3)	{ items, count, failed, budgetExhausted, cap }
GET	/	Header x-apify-container-server-readiness-probe	OK

Hard cap: 3 dossiers per request. Bulk stays on Start. Same PPE as a dataset row. First byte timeout: 5 minutes.

curl -H "Authorization: Bearer $APIFY_TOKEN" \
  "https://<user>--french-property-intel-hub.apify.actor/lookup?street=21%20rue%20Victor%20Hugo&postalCode=92600&city=Asnieres-sur-Seine"

Standby memory 512 MB, idle 60 s (Console → Actor → Standby).

What you get by default: normalized address, DPE class and surface, recency-weighted estimate (P25 / median / P75), street / neighborhood / commune €/m², commune rent check, flood/clay flags, copro id when published, SCI SIREN when a company sits at that number.

When to fill the listing fields: paste asking price to score the ad (not a multi-lot DVF deed). Paste monthly rent of this dwelling so yield is this unit. Paste owner charges per year (use 0 if none) for a net yield. Leave them empty for a market-only dossier.

How to score a listing

1. Open the Actor in Apify Console.
2. Keep the prefilled example, or paste your number and street, 5-digit postcode and city.
3. Optional: asking price, monthly rent of this dwelling, owner charges per year.
4. Click Start. Open the Investor dataset view — export JSON / CSV / Excel.

First run stays small (Max addresses = 25) so you can check one listing before a batch.

What it extracts

Category	What you get
Address	Normalized label, BAN id, GPS, INSEE city code
Energy	Latest DPE class, surface, consumption, F/G flag, Loi Climat rental calendar
Value	Recency-weighted estimate, P25–P75, similar-sale count, street / neighborhood / commune €/m², yearly trend
Deal	dealScore vs asking when provided (underpriced / in_line / overpriced) — never vs a package deed
Yield	This dwelling when you paste rent · commune ANIL as a check · net when charges are set (including 0)
Copro	Lots and syndic when the DPE has an id — charges are not in open data
Hazards	Flood, clay shrink-swell, earthquake and other public flags
Owner	SCI / foncière SIREN, legal form, public directors — emails always empty

Typical fill on a suburban house (Asnières-sur-Seine reference): DPE and estimate almost always present; phone/email never; copro register empty on a maison; SCI SIREN only when a company is registered at that number.

How much does it cost to score a French listing?

HTTP-only open data — default 512 MB, no browser, no residential proxy. Platform usage included. You pay per dossier, not per cadastral plot.

City listing Actors are ~$1.29 / 1,000 verified numbers. This hub is the next hop: ~$3.49–$5.99 / 1,000 dossiers (about 3–5× a listing row).

	All Apify plans
Actor start	$0.00005 / GB
Dossier with owner SIREN	$5.99 / 1,000
Dossier without owner SIREN	$3.49 / 1,000
1 listing, with SIREN	~$0.006
1,000 dossiers with SIREN	$5.99

Addresses that do not geocode are not billed. Dataset writes are included (not charged again). Free Apify accounts export 20 dossiers per run.

Input

Field	Default	Notes
street	21 rue Victor Hugo	Number and street only — no postcode here
postalCode	92600	5 digits
city	Asnières-sur-Seine	City name
askingPriceEur	—	Optional listing / offer. Deal score is vs this figure
actualRentEur	—	Monthly rent of this dwelling
chargesEurYear	—	Annual owner charges. Use 0 if none. Needed for net yield
places	—	Extra { street, postalCode, city, … } rows
maxItems	25	Dossier cap (1–500). Free plan 20 / run
collectDpe / collectDvf / collectLoyers / collectRnic / collectRisks / collectOwners	true	Turn a block off if you do not need it
ownerMatchMode	SCI / property companies	Or + syndics, or any company at the number

API-only (accepted in JSON, not shown on the Console form):

Field	Default	Notes
verboseLogs	false	Extra diagnostics in the run log (Debug: lines)

JSON example

{
  "street": "21 rue Victor Hugo",
  "postalCode": "92600",
  "city": "Asnières-sur-Seine",
  "askingPriceEur": 590000,
  "actualRentEur": 1800,
  "chargesEurYear": 0,
  "maxItems": 25
}

How it works

1. Resolve the address on the official French address database.
2. Pull energy class, official sales, commune rent, copro register, hazards and companies at that number.
3. Estimate the DPE dwelling from similar sales (same type, surface ±30%, recency-weighted).
4. Score asking vs that estimate. Compute yield from your rent when provided.
5. Export one dataset row as soon as the dossier is ready.

Output sample

Transformed values from a live Asnières-sur-Seine house (asking 590 000 €, rent 1 800 €/month, charges 0):

{
  "label": "21 Rue Victor Hugo 92600 Asnières-sur-Seine",
  "geocoded": true,
  "citycode": "92004",
  "surfaceM2": 90.5,
  "latestDpeClass": "C",
  "latestDpeDate": "2024-10-14",
  "rentalBanStatus": "none",
  "estimatedValueEur": 622188,
  "estimatedValueLowEur": 568159,
  "estimatedValueHighEur": 674135,
  "similarCompsUsed": 34,
  "similarBandPct": 30,
  "marketTrendPct": -9.5,
  "askingPriceEur": 590000,
  "askingVsEstimatePct": -5.2,
  "comparedTo": "asking",
  "dealScore": "in_line",
  "dealLabel": "Asking in line with estimated value",
  "actualRentEur": 1800,
  "chargesEurYear": 0,
  "yieldKind": "actual_net",
  "grossYieldVsEstimatePct": 3.5,
  "grossYieldVsAskingPct": 3.7,
  "netYieldVsEstimatePct": 3.5,
  "communeGrossYieldVsEstimatePct": 4.3,
  "latestSaleMixed": true,
  "latestSalePriceEur": 564382,
  "latestSaleDate": "2025-07-17"
}

One row = one address. dealScore in_line means asking is within ±8% of the dwelling estimate. yieldKind actual_net means headline yield uses this lease and charges — not the commune ANIL figure (kept as communeGrossYieldVsEstimatePct).

Dataset views: Investor (value / yield / DPE / owner) and Overview (wider reporting).

Limits

• France only.
• Owner match is street number + postcode, not a cadastral title. Several SCI can sit on the same building. Private individuals without a company are out of scope.
• A deed that mixes several dwellings or a shop is a package: no headline €/m², estimate stays the DPE dwelling.
• Neighborhood = 400 m around the point (the public sales file has no IRIS/quartier code). Recent sales weigh more (half-life 2 years).
• ANIL rent is a commune asking-rent indicator (charges included, unfurnished ads). Fragile communes are flagged rentReliable: false.
• Copro charges and unpaid fees are not in open data.
• DPE rental bans (Loi Climat): G since 1 Jan 2025, F from 1 Jan 2028, E from 1 Jan 2034. This is a calendar, not legal advice.

Is it legal to scrape French property data?

This Actor only reads data that France already publishes: API Adresse, ADEME DPE, geo-DVF, ANIL carte des loyers, ANAH RNIC, Géorisques, INSEE SIRENE.

As with any dataset that can contain personal or company information, ensure your use complies with GDPR and your own purpose. This is not a substitute for a notaire, diagnostician or legal opinion.

Local development

cd french-property-intel-hub
npm install
# storage/key_value_stores/default/INPUT.json must include street, postalCode, city
# (same keys as Console prefill), or:
npx apify run --input-file=./.actor/INPUT.json

On apify run, CLI input overwrites a root input.json when both exist. Local dataset lands in storage/. Set "verboseLogs": true in input for diagnostics.

Also available

Find the house number first (city → verified address, ~$1.29 / 1k):

• France Pool-Ready Plots · Builders by City — house + large garden, no OSM pool.
• France Garden Plots · Landscapers by City — garden ≥ X m².
• France Building Roofs · Size by City — roof plan ≥ X m².
• France Building Heights · Facades by City — height ≥ X m.
• France Industrial Buildings · Hangars by City — IGN hangars.
• Pools & Tennis · Cadastre Plots by City (FR · ES) — mapped pool or tennis/padel.

Same energy / owner family:

• French DPE F/G · Property Owner SIRENs — nationwide passoires → unique SCI file (not a single-address dossier).
• RGE France · Artisan Emails & Directors — qualified renovation contractors.
• French Companies · Search & SIREN Enrich — company search and SIREN enrich.
• French Syndics & Condos · ANAH Registry — syndic / copro directory.

Support

Contact corentin@outreacher.fr if you need a custom property workflow or CRM mapping. Use the Issues tab on the Actor page for bugs.