Evidence architectureUnited KingdomPublished 18 August 2026

UK Property Evidence Architecture

How official records and observed market evidence can be connected to place, property and time without collapsing different evidence classes.

Explore the interactive Property Atlas

01249,815,084

official source records loaded

Across nine non-exclusive evidence sources

029

official evidence sources

Property, market, planning, energy, place and risk

0311,083,904

dated rental observations

Kept separate from administrative records

041968

earliest series start

UK House Price Index history

Principal findings

What the evidence shows

The evidence base is broader than a rental feed.

Property references, completed sales, planning entities and energy records provide the stable national spine. Repeated rental observations add a changing market layer without being mistaken for the property itself.

Time coverage is source-specific.

The longest monthly series begins in 1968. Sale transactions begin in 1995. Other administrative and observed sources cover different periods, so any analysis must declare its common time window.

Scale cannot replace identity work.

The record total is intentionally described as records, not homes. Linking evidence requires a separate, conservative identity process and ambiguous candidates remain unresolved.

Figure 1

Composition of the loaded official evidence base

Share of loaded rows

Share of the 249,815,084 loaded official-source rows represented by each large evidence family.

Energy performanceCertificates and recommendation rows

58.0%

Property referencesOS Open UPRN

16.7%

Sale transactionsHM Land Registry Price Paid

12.6%

Crime and outcomesGeocoded street-level rows

6.8%

Planning entitiesNational Planning Data catalogue

4.3%

Other official seriesPostcodes, rents, HPI and council tax

1.7%

Reading note. Record shares describe storage composition, not the relative importance of a source and not a count of unique homes.

Method

How this output was produced

Inventory each loaded source at the row type supplied by its publisher.
Retain original geography and coverage dates before creating any cross-source features.
Report overlapping source totals separately and prohibit their use as a unique-property count.
Keep continuously observed market states distinct from official administrative records.

Interpretation boundary

What this release cannot prove

Loaded counts can differ from publisher landing-page totals because releases, exclusions and load dates differ.
National does not always mean the same boundary: several sources cover Great Britain, England and Wales, or England only.
The atlas describes available evidence. It is not a claim that every source is joined to every property.

Source register

Evidence and citation

01UK property data source ledger 02OS Open UPRN — Ordnance Survey 03ONS Postcode Directory — Office for National Statistics 04Price Paid Data — HM Land Registry 05Private rental market statistics — Office for National Statistics 06UK House Price Index — HM Land Registry and ONS 07Council tax levels — Ministry of Housing, Communities and Local Government 08Street-level crime — Police.uk 09Planning Data — Ministry of Housing, Communities and Local Government 10Energy Performance Certificates — Ministry of Housing, Communities and Local Government

Suggested citation

Upgraded Intelligence (2026), UK Property Evidence Architecture, version 1.0, UI-ATL-UK-2026-01.

Questions, corrections or requests for an accessible extract can be sent to hello@theupgradeauthority.co.uk.