Methodology
Every ownership verdict on this site is the output of a documented pipeline: sources are checked in a fixed order, weak signals go to human review instead of publishing, and each practice page shows how its verdict was reached. This page describes that pipeline exactly — including its limits.
The evidence we accept
Practice listings come from the RCVS Find a Vet register; ownership verdicts rest on the evidence standards below — stronger evidence always outranks weaker. Every verdict records which evidence produced it, and that audit trail is what practice pages render as “How we know”.
-
Human verification outranks automation
Once a person has verified and locked a practice’s ownership, automated detection cannot change it. Anything that disagrees is escalated for human review instead.
-
The practice’s own statements
CMA rules require practices to disclose corporate ownership on their own websites. A practice’s own published statement — a disclosure naming its owner, or a claim of independence — is first-party evidence, and the practice page links to the site so you can read it yourself.
-
Official registers
Official company and charity registers are the strongest evidence we hold. Where a practice is owned through a company, its ownership is established from public register filings; where it is run by a charity, we record the charity’s registration. Because these records are public, the link between a practice and its owner is one you can check for yourself.
-
Independent signals are triangulated
We collect several independent evidence signals per practice — the register control chain, the registered-office address, shared company officers, the practice’s own disclosure, charity-register matches. Two or more independent signals agreeing is the strongest verdict we publish; any disagreement between signals is escalated to human review rather than resolved silently.
-
Brand names
Practices trading under a known corporate brand (e.g. Vets4Pets, Vets Now) are linked to the brand’s owner. Ambiguous brand resemblances are not published — they are held for human review.
-
Third-party lists are leads, never verdicts
Externally compiled lists of corporate-owned practices are used only as internal verification leads. A match alone never publishes: it queues the practice for first-party verification, and is never cited as a source.
-
No evidence → “no corporate owner found”
A practice with no corporate signal is listed as independent, but its page says exactly what we know: no corporate owner found — not “verified independent”.
Verification tiers
Each practice carries one of three verification tiers, shown as a chip on its page, on search results and in comparisons. The tier states how well the verdict is evidenced — nothing more. We deliberately do not publish an internal confidence score; the tier is derived only from what kind of source backs the record.
Verified
Ownership confirmed against official registers.
Qualifying signals: Confirmed against official registers.
Reported
Ownership stated by the practice itself (its website or trading name) or confirmed by a human check.
Qualifying signals: The practice’s own website disclosure or independence statement, a corporate trading name, a manual check, a confirmed reader report.
Unverified
No positive ownership signal on record — classification not yet verified against a primary source.
Qualifying signals: The no-signal default, a clean-but-inconclusive website scan, records classified before the current evidence standard.
An “unverified” chip is not an accusation — most unverified records are small independent practices with no corporate trail to find. It means we haven’t yet confirmed the classification against a primary source, and the page says so.
Freshness rules
- Continuous re-verification. Records are re-checked on a rolling basis, and a corporate claim that can no longer be confirmed against fresh evidence is not silently republished — it is pulled back for human review. The data-sources page publishes live counts of how recently the dataset was re-verified.
- 365-day public staleness flag. Records not verified within a year carry a visible “this may have changed” warning. Re-verification normally happens well inside this window, so the flag only appears when the pipeline has genuinely failed to refresh a record.
Corrections
Reader reports go through the same review process as our own findings: a correction report becomes a proposed change, a human verifies it against primary sources, and an accepted correction is applied, recorded in the public change log as a correction, and credited to reader feedback in the practice’s “How we know” line. If you leave an email address, we’ll let you know when your correction goes live.
What we don’t do
- We never publish an ownership verdict from a third-party list alone.
- We never let practices edit their own ownership classification directly — corrections go through independent verification.
- We never take payment that could influence an ownership verdict or a ranking.
See also: where the data comes from, with live freshness statistics drawn from the dataset, and the public log of recent changes.