Editorial standards

How CardLoupe writes, reviews, and signs its work

Three columns, no personality

The library is organised into three editorial columns, one for each starting point we score against: where you are today decides which column is written for you. CardLoupe has no single named personality as the face of the site, so a guide is signed by the column that produced it rather than by an individual writer. That is a deliberate choice, not a way to hide: an invented author with an invented headshot and invented credentials would be a fiction, and the accountability that matters — who checked the numbers, against which issuer disclosure, on what date — belongs to a real named reviewer instead.

• The Starting Line

  For readers whose situation is “My first card”. Byline: The Starting Line — CardLoupe Editorial. Scores in this column disclose the First card weighting — 40% accessibility & fit.

  Read The Starting Line

• The Second Card

  For readers whose situation is “Another card”. Byline: The Second Card — CardLoupe Editorial. Scores in this column disclose the Another card weighting — 30% ongoing value, 20% welcome value.

  Read The Second Card

• The Long Game

  For readers whose situation is “Maximum benefits”. Byline: The Long Game — CardLoupe Editorial. Scores in this column disclose the Maximum benefits weighting — 25% benefits & protections.

  Read The Long Game

AI-assistance disclosure

Where AI tools help draft, structure, or edit an article, that article says so in the byline area: “AI-assisted, human-reviewed.” The flag is stored on the article record itself, not applied by hand, so it cannot be quietly dropped. AI is never the last step: no rate, fee, bonus, or score reaches a published page without a person confirming it against the issuer’s own current disclosure.

The human review process

Every article starts as a draft, moves to pending review, and can only publish once a named human reviewer is recorded on it — the database refuses to publish an article with that field empty, and refuses to publish a hero image without alternative text. Status changes are written to the same append-only audit log used for card data. Card facts inside a guide are read live from the card record, so a guide cannot drift from the comparison table, and every ranking-affecting change stays visible in the Trust Ledger.

Guides are editorial information, not individualised financial advice, and nothing here is a guarantee of approval. See our methodology for how scores are calculated and why affiliate compensation is not one of the inputs.