Skip to content

We are still building this site. We are adding more listings and features.

Trust and verification

How we decide what you can rely on

We combine public registers, organisation websites, automated checks and human judgement. The listing tells you which of those actually happened.

From source to public listing

Not every record goes through the same path. A small community group may have only its own website; a registered charity may have several official and public sources.

  1. We find a possible organisation

    Public registers, organisation websites and community suggestions can all start a record.

  2. We match sources to the same organisation

    Names alone are not enough. Websites, phone numbers, charity numbers and locations help us avoid combining different organisations.

  3. We check whether it belongs in the directory

    Most of the catalogue is checked by automated systems first. A narrow set of ordinary cases can go live automatically when two independent AI systems agree at high confidence — anything they disagree on, or that falls into a sensitive category, goes to a person. Our aim is for a person to review every organisation eventually, whether or not automation published it first.

  4. Some descriptions are written by automated tools

    This summary was written by automated tools and checked against the source material it was written from. Every listing that has one says so on its own page.

  5. We publish only the safe public view

    Private contacts and residential addresses stay out of the public listing.

  6. We keep checking

    Visitors can report a change at any time, and the date a listing was last checked stays visible. We’re building automatic re-checks against the official registers next — see our coverage page for exactly what’s live today.

What the trust labels mean

Every listing resolves to exactly one of these, by strength — the strongest one that applies is what you see.

Reviewed by our team

Checked by two independent AI systems

Two independent AI systems had to agree at high confidence before this listing went live. No person has reviewed it yet.

Matched to official register records — not yet reviewed by our team

Community verified

Confirmed by our team from public social media, not an official register

Not yet verified against an official register.

Record refreshed

A freshness signal, separate from the five states above: whether we've recently re-checked a listing's data against an official register. It's not a review claim either way.

When sources disagree

If two of our sources disagree about a location and we can't tell which is right, we hide the disputed detail rather than show something that might be wrong, until it's resolved. There's no separate on-page label for this today — the affected field simply doesn't show.

We show uncertainty instead of filling the gap

Missing opening hours, eligibility or access information does not mean a service lacks those things. It means we do not yet have evidence we can safely publish.

What "not yet known" means

For a small number of fields — contact details, area, and opening hours — we only ever show two things: what we have evidence for, or an honest "not yet known". We never guess, and we never quietly treat a gap as a "no".

Corrections stay connected to their source

Every fact on a listing keeps a record of where it came from, including earlier versions when something changes — so it's possible to see what changed and when. If you report a problem, a reviewer uses that history to check your report against the directory's record.

Tell us about a missing service

Where our data comes from

Some listings began as leads from public open-data sources and are cross-checked against independent sources before we publish them.

Data includes information from OpenStreetMap, © OpenStreetMap contributors, ODbL 1.0.