How PlaceScore is computed

PlaceScore reads its raw data from the HomePortfolio Data API but computes every score with its own deterministic engine. The API also returns a postcode-level walk score, but it is unreliable in dense areas: one London suburb (Isleworth) with over 1,000 nearby amenities and a quarter of households living car-free was scored just 15 out of 100. We show that figure only as a flagged comparison, never as the product score.

Our engine is pure and fully tested. The same address with the same data always produces the same scores and the same breakdown, and every point traces back to a real amenity at a real distance.

The Walk score is a weighted average across eight everyday-life categories: groceries, dining and social, schools, parks and green space, health and pharmacy, fitness, everyday errands and retail, and transit access. Each category combines two signals:

• Density (richness): how many of that amenity are nearby, with diminishing returns, so the difference between 1 and 5 shops matters more than 50 versus 55.
• Proximity (distance decay): how close the nearest one is. Anything within about 400m gets full credit; credit decays smoothly to zero by about 1.6km.

A category where the nearest amenity is genuinely at your doorstep counts for more than one that merely exists somewhere in the postcode. Because we weight proximity heavily, a dense, mixed-use neighbourhood scores higher than a car-dependent town that happens to have a similar raw amenity count. Every category contributes a number of points, and those points add up to exactly the score shown.

The raw amenity counts we read are a wide-area total, so we also apply a walkable-density adjustment: we measure how far you actually walk to reach genuine, everyday amenities (ignoring builders, storage yards, car dealers and other non-pedestrian businesses) and scale the score down where those amenities are sparse or spread out. This stops a new-build estate on the edge of a town from inheriting a high score it hasn't earned, and is shown as a note on the score breakdown.

Transit and Bike are multi-signal composites. Transit blends public-transport commuting share, car-free households, station access (count and nearest distance), population density, a daytime-activity proxy, and amenity diversity. Bike blends cycling-to-work share, the walk score, density, air quality, car-free households, and the number of nearby amenities.

If fewer than three of those signals are available for a location, we do not show a score at all rather than guess. That is why some rural addresses show a Walk score but no Transit or Bike score.

The headline PlaceScore weights Walk at 50%, Transit at 30%, and Bike at 20%. When Transit or Bike cannot be computed, we degrade gracefully using a discounted Walk proxy rather than dropping the headline, and we always show the three component scores so you can see what is driving it.

Each score carries a confidence level: high, medium, low, or approximate. High confidence means we found amenities across most categories with several at known distances. When amenity data is essentially absent (roughly 6% of UK addresses), we return an honest "approximate" state and say so, rather than fabricating a number.

We surface, rather than hide, the realities of the underlying data. The upstream points-of-interest table can lag by several weeks (its refresh pipeline is currently paused), so amenity counts and distances may be slightly out of date, and a specific shop or supermarket can be missing or renamed. Each score page shows a freshness badge with the snapshot date. Air quality and flood risk are shown as context only, not as core score inputs.

The detailed schools list comes from the provider and is not ranked by distance, so it can include schools beyond the immediate area. We therefore derive the genuinely nearest school from the distance-ranked data and show it separately. We never invent an amenity that isn't in the data; where it is incomplete, the score and its confidence level reflect only what we can actually see.

PlaceScore presents derived estimates to help you understand a place. They are not a substitute for visiting, and they are not guarantees. For full property intelligence, including valuation, schools, crime and planning detail, each page links through to HomePortfolio.