Our Process

What distinguishes a reliable source from noise is not format, reputation, or confidence. It is process — a repeatable, visible set of steps that happen before a word reaches a reader.

This page describes exactly how we work. We publish it not as a credential, but as a form of accountability. If you read a piece in Ardent Lens and it does not appear to meet the standards described here, we want to know. The process is only worth publishing if we hold ourselves to it.

Our process has seven stages. Each stage has a gate — a set of conditions that must be met before the piece moves forward. No piece skips a stage because a deadline is close. No piece skips a stage because an editor likes the topic. The process exists precisely to resist those pressures.

Not every important event becomes an Ardent Lens story. We apply three criteria at the commissioning stage: public interest, originality, and what we call the contribution test.

Public interest means the story matters to people who are not directly involved in it — that it has consequences, illuminates a pattern, or changes how something is understood. Originality means the piece adds something that does not already exist: new reporting, a new synthesis, an angle that changes what the existing coverage implies. The contribution test asks a simple question: if we do not publish this, what does a reader lose? If the honest answer is "not much," we do not publish it.

Timeliness alone does not pass these tests. A story that is urgent but derivative — repackaging what wire services have already reported — does not meet our bar. We would rather publish a piece two weeks later that is better sourced and better argued than a quick take that adds nothing to the public record.

Research begins with a source hierarchy. Primary sources — official documents, legislative records, court filings, datasets, peer-reviewed studies, and direct testimony from people with direct knowledge — sit at the top. Secondary sources (reporting from other outlets, academic commentary, published analyses) are valuable context, but they do not substitute for primary material when a material claim depends on it.

Expert sources are consulted for context, interpretation, and identification of what is and is not established in a given field. We identify experts by name wherever possible and include their institutional affiliation and relevant expertise. We do not quote "experts say" without specifying who those experts are. Where a source requests anonymity, we require a credible reason — a genuine professional or personal risk — and we disclose the nature of that reason to readers without identifying the source.

We do not treat any single source as sufficient for a significant factual claim. Where a claim can only be supported by one source, we say so. Where it cannot be independently verified at all, we say that too.

Fact-checking at Ardent Lens is not a single pass at the end of editing. It is woven into the research and drafting process, and it includes a dedicated review stage before any piece is cleared for publication.

What gets checked: every quantitative claim (statistics, dates, figures, rankings); every attributed quote (verified against transcript, recording, or contemporaneous notes); every reference to named individuals (their title, role, and position verified at time of publication); and every causal claim (the logic of the argument reviewed against the evidence cited to support it).

When claims cannot be verified, we have a clear policy: we do not publish claims we cannot verify as if they are established. We either hold the claim until verification is possible, frame it explicitly as unverified, or remove it. Publishing under uncertainty with insufficient disclosure is, in our view, a form of misleading readers — even when the claim turns out to be true.

A draft enters editorial review when the writer considers it complete and the research underlying it is documented. Editorial review is multi-pass and explicitly layered: structural editing comes first, line editing comes second.

Structural editing addresses the argument: Is the central claim clear? Does the evidence actually support it, or is there a gap between what the writer asserts and what the sources establish? Is the structure of the piece the right one for this argument — does the order of sections lead readers to the right conclusion, or could it mislead? This phase often involves the most significant back-and-forth between writer and editor. It is not uncommon for a piece to go through three structural revisions before line editing begins.

Line editing addresses language: precision, clarity, and concision. We edit for accuracy of expression — does the sentence say exactly what the writer means? — and for readability. We do not edit for style alone, or for a house voice that overrides the writer's. A good line edit makes the writer clearer; it does not replace them.

Before a piece is cleared for publication, a final reviewer — who has not been involved in the drafting or editing — reads it with fresh eyes against a standard checklist: Are all factual claims traceable to a source in the file? Does the headline accurately represent the content? Is any language in the piece stronger than the evidence warrants? Are all identifiable individuals given an opportunity to respond to material claims about them?

The right to respond is taken seriously. If we are about to publish a material claim about a named individual or organisation, we contact them before publication. We give a reasonable deadline. We include their response, or note that they declined to comment or did not respond, in the published piece. We do not grant organisations editorial approval over what we write. Response and control are different things.

We publish when a piece is ready, not when a slot opens. We do not hold a finished piece indefinitely, but we do not publish an unfinished piece to meet a schedule. When a story is time-sensitive and a piece is not ready, the honest response is to publish a shorter, clearly scoped piece that we can stand behind — not to rush the full version.

Headlines are held to the same standard as the body of the piece. We do not write misleading headlines, qualify them with body text, or use framing that implies more certainty than the reporting supports. A headline that generates engagement by overstating the story is not a successful headline; it is a small editorial failure multiplied by every person who reads it.

We respect embargos when they serve a legitimate purpose — coordinating scientific announcements, giving governments time to respond to data releases, enabling simultaneous international coverage of shared events. We do not respect embargos used simply to manage media narratives or delay accountability coverage.

Errors are classified at three levels. Typographical errors — spelling, punctuation, formatting — are corrected silently, as the substance of the piece is unchanged. Factual errors that do not materially affect the meaning of the piece — an incorrect date, a wrong job title — are corrected and labelled with a footnote noting what was changed. Material errors — errors that change what a piece argues, what it claims about a person or organisation, or what conclusions a reader would draw — receive a prominently placed correction label above the body text, explaining what was stated, what the correct information is, and when the correction was made.

We do not quietly delete passages, change claims without disclosure, or update articles to reflect new information without noting that the update has been made. When an article is updated with new information after publication — not because it was wrong, but because the story has moved on — we note the update date and summarise what was added.

If you believe we have made an error, or have information that contradicts something we have published, please contact us. We take these communications seriously, and every one of them is reviewed by an editor.