Data sources & credits

sqmly is built on public, free, authoritative datasets. Every score we show traces back to one of these sources. If a source has a limitation we know about, we say so on the dimension it powers.

Disclaimer: sqmly scores are independently calculated and are not official ratings issued by any government body. Use them as one input among many, not as a substitute for a formal survey, EPC, or local-area report.

MHCLG Get energy performance of buildings data

What: Domestic Energy Performance Certificate (EPC) records for every UK property that's been let, sold or surveyed since 2008. Includes certified measured floor area, EPC band, current energy efficiency score, and assessor-estimated annual heating / lighting / hot-water cost.

How we use it: (1) Cross-validates the OCR'd floor area on Space & Size — when the OCR and EPC agree we mark the dimension as register-verified; when they diverge by more than 15% we use the EPC value. (2) Drives the Bills dimension — a renter-relevant score on the actual energy bill they will pay.

Licence: Open Government Licence v3.0

data.police.uk

What: Per-month street-level recorded crime within an arbitrary polygon, with category and approximate location.

How we use it: Drives the Crime dimension. Aggregated into H3 res-9 hexes (~0.1 km², ~330 m across) covering the listing's centroid + its 6 neighbours.

Licence: Open Government Licence v3.0

ONS Census 2021 — Usual residents (Nomis NM_2021_1)

What: Population counts for every Lower Layer Super Output Area (LSOA21) in England + Wales — 35,672 records.

How we use it: The denominator for crime rate-per-1,000 calculations. Shipped as a static JSON inside the backend image.

Licence: Open Government Licence v3.0

postcodes.io

What: Postcode → (lat, lng, LSOA21, admin district, region, county). Free, no API key.

How we use it: Resolves the listing's postcode to a centroid + LSOA so we can geocode and look up population.

Licence: Open Government Licence v3.0 (built on ONS Postcode Directory)

OpenStreetMap (Overpass API)

What: Points of interest within 1.5 km of the listing centroid: bus / tram / train / tube stops, supermarkets, cafés, gyms, GPs, schools, motorway / trunk / primary roads, and railway lines.

How we use it: Drives Transport, Convenience and Noise dimensions. One Overpass query per cold cell, then cached in our hex store for 30 days.

Licence: Open Database Licence (ODbL) — © OpenStreetMap contributors

Uber H3 (geo grid)

What: Hierarchical hexagonal grid system used to bucket geo lookups.

How we use it: Resolution 9 (~0.1 km², ~190 m edge) is the natural cell size for street-level analysis without over- or under-resolving.

Licence: Apache-2.0

Rightmove

What: The property listing itself — price, address, postcode, advertised bedrooms, floor-plan images.

How we use it: We fetch the listing page only when the user explicitly pastes its URL. No mass scraping, no aggregation, no resale.

Independent verification

We don't want you to take our scores on trust. These references let you cross-check what we say about an area against the authoritative reports the same data feeds.

What we deliberately do not use

No proprietary house-price or estate-agent feeds — keeps the service free of conflicts of interest.
No paid “area scoring” APIs — every dataset above is free and openly licensed.
No mass scraping of competitor sites. We only fetch the single Rightmove URL a user pastes.