Methodology

How the numbers are made

Every statistic we publish — on the site, on social, or supplied to journalists — follows the same written discipline. If a figure can’t carry its query, sample size and caveats, it doesn’t ship.

Sources

  • Listings: collected automatically each day from adverts publicly displayed on Rightmove (we are not affiliated with Rightmove); each record keeps first-seen date, price history, tenure, service charge and EPC where disclosed.
  • Completed sales: HM Land Registry Price Paid Data, 1995 to present (about 4.0 million sales in London's 33 boroughs).
  • Value my home also uses sold-price records collected automatically from sold-price pages publicly displayed on Rightmove: its same-postcode comparable sales come from them, and its price model was trained on them.
  • Context layers: EPC register, police.uk street-level crime, ONS rent benchmarks, planning.data.gov.uk.

How we compute

  • Medians over means, always, unless explicitly labelled otherwise.
  • Sample sizes are stated; aggregates with n < 3 are never published.
  • NULLs are excluded, never treated as zero. Partial-coverage fields (e.g. service charges, disclosed on roughly half of listings) are always qualified as such.
  • Resale price changes use consecutive-pair comparisons within the same property, not first-vs-last.

Known limits

  • “First visible” is when we first saw a listing — it can lag the true listing date.
  • Land Registry registrations lag completions by weeks to months; recent months are undercounted and we window around that.
  • Bedroom counts as listed are unreliable at the tail; we guard aggregates against them.
  • We track London and its commuter fringe, not the whole UK.

Checking a single home yourself? Our guide How to check a London property before you buy lists the official sources for each check.