Skip to main content
TrueDeed

TrueDeed awards

Honest Agent Awards methodology

This is the complete recipe for the Honest Agent Awards. If an agency is ever scored or ranked, this page is exactly how it was done — computation version 1. Scores come from the transaction record only: no votes, no judging panel, no fees.

Data sources

Completions: HM Land Registry Price Paid Data, matched to TrueDeed listings only where a listing-to-completion match is confirmed — never fuzzy-matched into a score. Listings:the agency's own TrueDeed listings. Every median is a median, never a mean, so one unusual sale cannot move a score.

The three year-1 metrics (lower is better on all)

  • Pricing accuracy:for each confirmed listing-to-completion pair in the award year, the absolute difference between the final asking price and the sold price, as a percentage of the sold price. An agency's score is the median across its pairs.
  • Time to sell vs local median:the agency's median days from listing to legal completion, minus the local median from our published Time-to-Sell dataset (the latest unsuppressed figure for the award year — national level at today's data volume). Negative means faster than the market. If no unsuppressed Time-to-Sell figure exists, no agency is ranked on this metric. Measured on completions only — “under offer” dates are never used.
  • Listing hygiene:of the agency's listings created in the trailing 12 months, the percentage that are withdrawn or stale (still active after 180 days).

The metric we dropped for year 1: fall-through rate

We wanted a fourth metric — the fall-through rate, the share of agreed sales that never reach completion. We are not scoring it in year 1, and here is the honest reason: our sale-progression records do not yet cover enough transactions to score any agency on it fairly (as of July 2026, effectively zero usable progression histories). Publishing a fall-through score from that would be fiction. It returns in year 2 if — and only if — progression coverage supports it, and this page will say so when it does.

The minimum-sample rule (disclosed, exact)

Every metric is computed for every opted-in agency, but an agency is ranked on a metric only when it has at least 5 qualifying observations for it (5 confirmed matched sales for pricing accuracy and time to sell; 5 listings for listing hygiene). Below that the agency is excluded from the ranking and told exactly why on its own dashboard — the score is never estimated, interpolated, or padded. A median over fewer than 5 observations is too easy for one sale to swing, so we refuse to rank on it.

What can never affect a score

  • Votes, nominations volume, or any form of popularity input.
  • Money — there is no entry fee, no sponsorship, and no paid placement anywhere in the awards.
  • Advertising spend or any commercial relationship with TrueDeed.
  • Opting in or out — participation changes visibility, never scores.

Sources

Known caveats

  • HM Land Registry publishes completions with a lag of roughly three months, so the most recent quarter is always under-counted — scores are computed after the lag has settled.
  • Agencies with fewer than 5 qualifying observations on a metric are excluded from that metric, not scored.
  • The fall-through rate metric is dropped for year 1 (sale-progression coverage is too thin to score honestly).
  • Only listings linked to an agency organisation on TrueDeed can be scored; agencies not on TrueDeed are absent, not zero-scored.
  • Computation version 1. Any change to a metric definition increments this version and is documented here.