How Our Scores Are Calculated
Every neighbourhood score on PostcodeInfo.uk is derived from official UK government datasets using transparent, reproducible statistical methods. This page explains the exact formulas, weights, and data sources behind each score.
database Primary Data Sources
ONS Census 2021 ↗
Population, ethnicity, occupation, housing tenure, education, industry, marital status, economic activity. Published March 2023. Updated at each Census cycle (~10 years).
Granularity: Output Area (OA21) — approx. 100–600 households per OA.
police.uk Crime Data ↗
Street-level crime incidents reported to 43 territorial police forces in England, Wales, and Northern Ireland. Updated monthly with a 1–2 month publishing lag.
Granularity: Anonymised coordinates snapped to nearest street.
HM Land Registry Price Paid ↗
Residential property sales in England and Wales since 1995. Updated monthly with a 1–2 month publishing lag.
Granularity: Individual transaction level, linked to postcode.
NHS Digital / CQC GP Register
Active GP surgery locations from NHS Digital. Updated quarterly.
DfT NAPTAN Stop Data ↗
National stop and station data for buses and trains. Updated as operators report changes.
leaderboard Score Methodology: 1–10 Decile Ranking
All five census-derived scores (Affluence, Family, Diversity, Homeownership, Employment) are calculated in two stages:
- Raw composite value — a weighted sum of Census indicators for each Output Area (OA), producing a continuous 0–1 score.
- Decile ranking — all ~180,000 OAs in England and Wales are ranked using SQL
NTILE(10), producing a 1–10 score where 10 = top 10% nationally. Scores are also computed relative to the local authority (city) and postcode district (outcode) for contextual comparison.
Plain English: How wealthy and professionally active the neighbourhood is, based on occupations, income levels, and industry mix.
Professional occupations: SOC Major Groups 1 (Managers), 2 (Professionals), 3 (Associate professionals)
Higher NS-SEC: NS-SeC classes 1 (Higher managerial) and 2 (Lower managerial)
Median household income: MSOA-level estimate from ONS income model (2020)
Finance & professional industry: SIC sections J (Information/communication), K (Finance), M (Professional services), N (Admin)
Plain English: How family-oriented the area is — based on multi-person households, house types suited to families, marriage rates, and the proportion of children.
Plain English: How internationally diverse the area is — measures non-UK born residents, passport holdings, recent arrivals, and international migration flows. A high score reflects a cosmopolitan, internationally connected community.
Plain English: How owner-occupied the area is — outright ownership carries the highest weight as it indicates the deepest community roots.
Plain English: How economically active and employed the workforce is. Full-time employment carries an additional weight to reflect economic depth beyond part-time activity.
Plain English: Crime density relative to the catchment population. A score of 10 means the area is in the lowest 10% nationally for crime per 1,000 residents.
Catchment population is derived from the sum of Census 2021 populations across all OAs whose centroids fall within 500m of the OA being scored.
Plain English: Relative property price growth over 1 and 5 years. A high score means prices have grown more than comparable areas nationally.
Where OA-level transaction data is sparse (<2 sales in a year), the score falls back to the outcode-level average. OAs with no sales in the reference period are excluded from ranking.
Active GP surgeries within 2km per 10,000 catchment population. Ranked nationally.
Source: NHS Digital active GP register.
Composite of bus stop density (500m) and proximity to the nearest railway station.
Source: DfT NAPTAN stop register.
smart_toy Automated Narratives & Area Character
In addition to numerical scores, PostcodeInfo.uk uses a deterministic, rules-based engine to generate plain-English summaries of each neighbourhood. We deliberately avoid generative AI language models for these descriptions to guarantee 100% factual consistency and prevent data "hallucinations."
Index Insight
Translates the core 1–10 decile rankings (Affluence, Family, Employment) into a cohesive written summary.
- Evaluates each metric independently.
- Only highlights statistically significant deviations (e.g., top 20% or bottom 30% nationally).
- If an area scores neutrally across all metrics, it is accurately described as a "statistically balanced neighbourhood."
Persona Inference
Identifies the dominant demographic character of the street using a strict hierarchical ruleset.
- Commercial: Triggered by high workplace density.
- Regeneration: Triggered by extreme IMD scores combined with high social housing.
- Renter Hub: Triggered by >40% private renting or high student density.
- Affluent/Family: Triggered by high homeownership, low deprivation, and settled age profiles.
Area Character
Combines the inferred persona with hard street-level data to create a contextual profile.
- Categorises the area into one of 10 distinct socioeconomic tiers.
- Dynamically appends strict factual statements about local crime rates, median property prices, and age demographics.
- Ensures text descriptions strictly match the actual statistical bounds (e.g., crime per 1,000 residents).
help How to Read a Score
| Score | Meaning | Nationally |
|---|---|---|
| 9–10 | Excellent | Top 20% |
| 7–8 | Above average | Top 40% |
| 5–6 | Average | Middle 40% |
| 3–4 | Below average | Bottom 40% |
| 1–2 | Lowest deciles | Bottom 20% |
Scores can be viewed at three levels of comparison using the tab switcher on each profile page:
- UK — ranked against all ~180,000 Output Areas in England & Wales
- City — ranked within the same local authority district
- District — ranked within the same postcode outcode (e.g. B18, SW1A)
For districts with fewer than 10 OAs, the outcode/city score mirrors the UK decile to prevent misleading micro-comparisons.
All underlying data is published under the Open Government Licence v3.0. PostcodeInfo.uk aggregates, normalises, and presents this data free of charge. Have a question about our methodology? Contact us.