The hidden-risk & due-diligence layer for UK property.

Every lens is a transparent weighted composite of official UK signals. Each input is first ranked as a percentile against all 846,137 covered UK postcodes (so "73rd percentile for crime" means more crime than 73% of the country), then combined with the emphasis shown below. The family, renter and investor lenses use the same signals re-weighted for different priorities — contextual intelligence, not different data. Nothing is modelled or predicted; these are present-day signals, shown openly so you can audit them. We only use signals we genuinely hold — transport frequency, nightlife, walkability and rental yield are not in any score, so a lens never implies data we don't have.

Family-fit — how well-suited an area is for raising a family

Higher is better (0–100). Safety and school access carry the most weight, deprivation and green space sit close behind, and Ofsted quality is a supporting factor (inspection coverage is incomplete). This mirrors the consensus across published family-suitability frameworks — ONS area classifications, academic "child-friendly neighbourhood" indices and property-portal family rankings — which all centre on safety, school access & quality, green space and socioeconomic stability.

Renters & young professionals — everyday-living fit for renters

Higher is better (0–100). Weighted toward fast broadband, safety and everyday cost — what renters tell us matters most among the signals we hold. Transport frequency, nightlife and walkability are not yet modelled, so treat this as a connectivity-and-safety read, not a "vibe" score.

Investor fundamentals — area fundamentals, not yield

Higher is better (0–100). This reads area fundamentals — quality, downside risk, affordability and connectivity. It is deliberately NOT a rental-yield, capital-growth or price forecast: we don't model rents or prices at postcode level (too few transactions to be honest), so this is a fundamentals lens to pair with your own yield numbers, read alongside the direction of travel below.

Direction of travel — which way the area is heading

Two real historical signals, not predictions: deprivation trajectory (IMD decile 2019 → latest — a higher decile means less deprived, so "improving") and council-tax trajectory (Band D, earliest → latest, as a total % and rough per-year rate). Deciles are coarse (1–10) so only larger shifts show; council tax rises almost everywhere, so the pace is the signal. The 24-month crime trend lives on the full postcode report.

Hidden-risk — risks a property listing won't show you

Higher means more under-the-radar risk (0–100): 0–25 Low, 25–50 Moderate, 50–75 Elevated, 75–100 High. Flood is the dominant factor, crime a major one, and air pollution a supporting signal.

Insurer risk — a climate & peril lens for underwriters

Higher means more insurance-risk exposure (same 0–100 bands as Hidden-risk). Unlike the Hidden-risk lens — which takes the worse of the two flood perils — this keeps river/sea and surface-water flood separate, because insurers price them as distinct perils. Crime is a loose theft / malicious-damage / escape-of-water claims proxy; air is an environmental-quality signal. It is a signal for underwriting, not an underwriting decision — no premium, no quote, no regulated advice.

Primary signals — how we rank "what matters most"

Above the lenses, the Primary signals panel ranks the signals we hold by impact for that specific area. Impact blends each signal's inherent stakes (flood, deprivation and crime are high-stakes; broadband is low) with how it actually reads in this postcode (a normally mid-stakes signal that's severe here is elevated). It's a presentation of the same official data — no new inputs, no scoring of anything we don't already show — designed to answer "what matters most here?" before the detail.

Missing data & confidence

If a signal isn't available for a postcode (e.g. an unrated nearest school, or no monitored air reading), we drop that input and re-weight the rest — we never substitute a guess. The confidence label reflects how many inputs were available: High (all/most present), Indicative (a few present), Low (sparse).