Skip to content

Editorial policy and evidence standards

CharityCompare helps people identify organisations, understand dated financial information and compare relevant evidence. A profile is not a guarantee of quality, safety or impact. Our financial ratings are calculated from published fields; they are not donor reviews.

Who is responsible

CharityCompare is published by Elite Digital AI Solutions Ltd, company number 17250017. The publisher is responsible for the interpretation, evidence standards and corrections on this site. Contact the editorial desk at [email protected]. The editorial contact is Kaiser Khan. His background and relevant affiliations are disclosed on that page. Named editorial responsibility does not imply that every automated profile has received an individual human audit.

Four kinds of information, clearly separated

  1. Register facts. Identity, status, classifications and financial fields from the relevant regulator. Activities and classifications are generally supplied by charities to the register; their presence is not independent verification of impact.
  2. Calculated observations. Ratios, annual changes and Clarity Scores derived from the dated data. These are reproducible calculations, not a statement that a person has audited every filing.
  3. Source-checked interpretation. Specific context based on the cited activities, reports or evaluations. Conclusions should stay within what those sources establish. AI tools may assist research, drafting and calculations; they are not credited as human reviewers and their outputs are not a substitute for source checks.
  4. Charity responses. Authorised descriptions or responses are labelled as provided by the charity and displayed separately from register activities and independent interpretation.

Results and impact

A published result must identify its source, reporting period, population or geographic scope, definition and material limitation. We label charity-reported results separately from independent evaluations. Service outputs, funding commitments and programme outcomes measure different things. Missing evidence on a profile means we have not added it; it is not a claim that the charity has no impact. We never invent testimonials, beneficiary figures or user-review ratings.

How ratings work and where they fall short

The Clarity Score V1.1 uses four pillars totalling 100 points. Undisclosed inputs receive no points; missing fundraising costs are excluded from the cost ratio and unknown filing status receives no filing points. V1.1 does not reweight the score for missing data. Consequently, disclosure coverage can affect comparisons between differently sized organisations and different registers. The worked example shows the actual calculation and missing-data behaviour.

A financial score does not measure programme effectiveness or establish misconduct. An inquiry triggers the formula’s zero-score rule, but an inquiry is not a finding of wrongdoing. The formula is applied consistently; changes to it belong in the methodology changelog.

Dates, status and historical records

We distinguish the extract date, the accounts period, the financial snapshot date and the date of a specific source check. An updated register extract does not make an old set of accounts current. Status notices use the regulator’s labels and retain their date. A suppressed return, zero activity or old accounts are not sufficient grounds to label an organisation closed or fraudulent.

Removed registrations retain a dated archive notice. Records with exceptional reporting statuses or old accounts remain available for reference, with historical-rating notices and exclusion from current promotional rankings. Linked funds are represented under their parent registration, with old URLs redirected where the parent is available. We preserve a labelled historical relationship page where the parent’s status prevents ordinary consolidation.

Source review and disagreements

Our editorial review queue identifies service routes for rechecking after 60 days, source-backed basic-profile supplements after 90 days, profile evidence and topic examples after 180 days, and evergreen guides and answers after 365 days. These intervals are prompts to check the original source, not claims that every page is updated on that schedule. A build reports overdue items; it does not silently advance a “last reviewed” date. Register extracts and accounts retain their own separate dates.

When credible sources disagree on a material point, we keep a dated record of each position and its source, state what remains unresolved, and identify what evidence would settle it. A newer statement does not automatically override stronger older evidence. If a source clearly supersedes an earlier fact, we mark the change and preserve the reason in the correction record. We do not publish private correspondence or unverified allegations as a dispute.

Corrections and right of reply

  1. Use the evidence link on the relevant profile or email the editorial desk. Include the registration number, exact field or claim, proposed correction and a public source with its date.
  2. A representative requesting a charity response must identify their role. We check their authority through the profile-claim verification process or established official contact details before publishing their words.
  3. A response is not automatically published. We assess relevance, supporting evidence, privacy and allegations about others. Private correspondence is not published merely because it was submitted.
  4. Accepted responses are clearly attributed. Factual corrections update the relevant information; material changes are dated in the log below. A disagreement does not automatically remove sourced information.
  5. If a decision remains disputed, use the complaints route. Explain the disputed evidence so the publisher can reconsider it.

No purchase is required to request a correction or response. We aim to acknowledge requests within five working days and provide a substantive response within 20 working days. If more time is needed, we explain the delay and give an expected response date. Working days exclude weekends and England and Wales bank holidays. These are our service targets, not a guarantee of publication.

A free approved profile claim can include an attributed description, claimed status and an optional reviewed logo. The submitter must have permission to use the logo; it is not published merely because it was uploaded. Claimed status identifies an approved representative submission, not an endorsement of the charity, its appeals or its impact. Optional paid tiers add enhanced presentation and reporting features; payment never bypasses authority or editorial checks.

Conflicts of interest

Kaiser Khan’s personal biography identifies roles with World Aid Network and World Animal Rescue Network. He must not be the sole decision-maker for disputed content or discretionary promotion involving an affiliated organisation. We retain source-led information, disclose the relationship on relevant editorial work, and require a reviewer without the conflict before publishing discretionary judgments about it. No external reviewer is claimed to have approved our scoring model.

Funding and independence

The site offers optional memberships, business services and reader support, described on our funding page. Commercial options do not purchase changes to ratings, evidence or rankings. Paid placements are labelled and kept outside ranked results. Donate links take visitors to an external charity website; CharityCompare does not process those donations or charge commission.

Relevant editorial links are chosen for usefulness. A requested link does not establish a relationship between two legal entities or earn a rating change. Paid links must be disclosed and marked as sponsored. A charity may not pay to remove a regulator finding or a material correction.

Research and selection

Our research reports state their population, exclusions, definitions and source dates, with downloadable observations and methods. The 50-profile pilot is a provisional editorial sample, not a ranking of search demand or a representative sample of UK charities. We report observed performance only when measurements exist.

Corrections log