How we decide what to publish

Every published piece goes through the same five steps. None of them is optional, and none of them is a formality.

• Commissioning. Ideas come from reader mail, from questions people are obviously asking search engines, and from gaps in what already exists. Search demand tells us what to write about. It never tells us what to say.
• Research. Primary sources first: government and consular pages for anything procedural, national statistics offices for numbers, published research for claims about behaviour, and the organisations that work with victims of romance fraud for anything about scams. Where we give a figure, we say where it came from and what year it describes. Forum posts, Reddit threads and other blogs are anecdote, not evidence, and are labelled that way in the text.
• Drafting. Written to an agreed outline that answers the question in the headline. Word counts are an output, not a target — if the honest answer takes 700 words, the piece is 700 words.
• Editing. An editor reads the draft against the sources, cuts anything unsupported, removes generalisations about nationalities, and checks that the safety and country standards below have been met. Anything that cannot be sourced is deleted rather than softened.
• Publication and review. The piece carries a byline, a publish date and, once its substance changes, an updated date. Country guides and safety pieces are re-read at least once a year. Anything procedural — fees, visa steps, reporting routes — is checked more often, because it is what goes stale fastest.

We do not publish articles supplied by third parties, we do not sell link placements inside existing pieces, and we do not republish other people’s work with a new headline on it.

Country guides are the most useful thing we publish and the easiest thing to get wrong. The line we hold is between a pattern and a stereotype.

A pattern is specific, attributed and bounded: “in most of the country, meeting a partner’s parents happens earlier in a relationship than a British reader would expect, and among urban under-thirties it happens later than that.” A stereotype is a personality handed to fifty million people at once. We publish the first and refuse the second.

The rules that follow from it:

• No national character claims.Sentences of the form “[nationality] women are [trait]” do not appear on this site. Not as praise, not as warning, not as a joke. This is a hard rule. It is how a guide turns into a stereotype, and it makes readers demonstrably worse at meeting the actual person in front of them.
• Internal variation is named, not buried. Capital versus second city versus village, class, generation, religion and observance, education, whether a family has emigrated. If a claim does not survive being qualified in those terms, it does not run.
• No rankings of people.No “easiest country to find a wife,” no “most traditional women,” no beauty tables, no cost-per-date league. The framing itself is the problem; a careful sentence inside that framing does not rescue it.
• Anecdote is labelled as anecdote.An individual story is presented as one person’s experience, never as how a country behaves.
• Language.Terms like “mail-order bride” appear only when we are describing that industry critically, never as a neutral description of a person.
• When in doubt, less. We would rather a guide be thinner than repeat a claim we cannot stand behind.

If you live somewhere we have written about and we have got it wrong, that correction gets priority over almost everything else in the inbox.

Safety pieces have a second audience nobody wants: the people running the scams. Writing them well means being useful to the target without being useful to the fraudster.

• Recognition, not recipes. We describe how a scam is structured, what it sounds like in the early weeks, and the moment it turns to money — enough to recognise it on day three instead of month six. We do not publish scripts, templates, or the operational detail that would make an approach more convincing.
• No stolen faces. We do not republish the photographs used on scam profiles. Those images are almost always taken from a real person who did nothing wrong and is already dealing with the consequences.
• No naming private individuals.We cannot verify a reader’s accusation against a named person, and the cost of getting it wrong lands entirely on a bystander. We describe methods, not people.
• Real reporting routes. Every safety piece points at the actual channels — the platform, the bank or payment provider (speed matters most here), and the national fraud body: IC3 in the United States, Action Fraud in the UK, the equivalent national police unit elsewhere. We are honest that recovery is often not possible.
• No recovery services.We will never recommend, review or accept advertising from a “fund recovery” operation. Those are the second scam that finds the same victim.
• Tone. Nobody gets told they were stupid. Romance fraud works on attentive, careful, intelligent people, and the article that shames the reader is the article they close halfway — which is the paragraph before the one that stops the next transfer.

If you are in immediate danger, contact your local emergency services. We are a website; our inbox is answered in days.

Two Timezones publishes general information for a general audience. It is journalism, not professional advice, and it does not know anything about your situation.

• Not legal or immigration advice. Visa, fiancé, marriage and residency rules change constantly and turn on facts specific to you. Use official government sources and a licensed immigration lawyer or accredited adviser. Nothing here creates a professional relationship, and we cannot review your case.
• Not medical, sexual-health or therapeutic advice. We write about the emotional weight of distance; we are not therapists and this is not treatment.
• Not financial advice. Including anything we write about remittances, transfer services or shared costs.
• Not a crisis line and not an investigator. We cannot verify whether a person is real, trace a payment, contact a platform on your behalf, or intervene in an ongoing situation.
• Not a dating service. We host no profiles, broker no introductions and match no one.

If it is wrong, we fix it and we say so. The scale of the fix decides what the reader sees.

• Typos, formatting, dead links. Fixed without a note.
• A factual error that changes the meaning. Corrected in the text, with a dated correction note at the foot of the article stating what it said before and what changed.
• Guidance that has moved on.The article is revised and carries an “Updated” date with a line explaining what is different.
• Removal. Rare, and a last resort. We correct in place rather than quietly delete, because a deleted page is a record of nothing.

To request a correction, email hello@twotimezones.comwith “Correction” in the subject line. Include the URL, the sentence you are disputing, and a source if you have one — though a first-hand “I live here and this is not how it works” is a source too, and we treat it as one. We reply to every correction request, usually within two business days, and we tell you what we changed even when the outcome is that we are leaving the piece as it stands.

Organisations described critically get a right of reply, and we will publish a fair summary of their response alongside the piece.

The site is funded by advertising and affiliate commission. Both are visible to you by design.

• Advertising.Every ad unit carries the word “Advertisement” above it, in every position, with no exceptions and no “native” units dressed as articles. Ads are served by third-party networks and are not endorsements; we do not see most creatives before they run. We remove any advertiser reported to us as a dating scam, an introduction agency selling contact with women, or a predatory lender — the full list is on the advertising page. Report one to hello@twotimezones.com and we will pull it.
• Affiliate links.Some outbound links to apps and services pay us a commission if you sign up. It costs you nothing extra. Pages containing them carry a disclosure near the top. Commission rates play no part in what we recommend or in the order of a list, and “none of these are worth it” is a conclusion we have published and will publish again.
• Sponsored content.We do not currently sell it. If that ever changes, a sponsored piece will be labelled “Paid” in the headline area and in the byline, its outbound links will be marked rel="sponsored", it will be excluded from rankings, recommendations and the newsletter, and it will never appear in a safety article. This paragraph will be rewritten before the first one runs, not after.
• Firewall. No advertiser, network or affiliate partner receives a draft before publication, has a veto over a piece, or can have a published piece amended or removed. Advertising is sold against the site, never against an individual article.

We use AI tooling, and we would rather tell you precisely how than let you guess. The tools are general-purpose large language models — the class of assistant that includes Claude and ChatGPT — used the way a newsroom uses a research assistant and a copy desk.

Where they are used: summarising long public documents so an editor knows which pages to read in full; building outlines; drafting passages of a first draft; rewriting clumsy paragraphs; generating headline, subheading and standfirst options; writing image alt text and meta descriptions; drafting quiz questions; and scanning our own copy for the generalisations this policy bans, which a model is genuinely good at catching.

Where they are not: no model output is published unedited. A model is never the source of a fact, a number, a statistic, a quotation, a law or a citation — every one of those is verified by a human against the primary source, and if the source cannot be found, the sentence is deleted rather than published with a hedge. We do not generate first-person experience, reader stories, testimonials or reviews. We do not generate photographs of people; if we ever publish an AI-generated illustration, the caption will say so. Nothing publishes automatically — a person presses the button.

Accountability.The name on the byline is accountable for every sentence in the piece, including the ones a tool helped write. If something is wrong, it is our error and we correct it under the process above. “The AI wrote it” is not an explanation we will ever offer you.

We disclose this because you cannot audit our process from the outside, and a disclosure you can read is worth more than a promise you cannot check.

• We do not accept payment for coverage, for links inside existing articles, or for guest posts. Unsolicited “I’d love to contribute” and “I noticed a broken link” outreach is declined without exception.
• We are not owned by, funded by or affiliated with any dating app, introduction agency, immigration firm or visa consultancy.
• Writers and editors do not accept paid trips, gifts or hospitality from services they cover. Where a company provides free access to a paid product for review, the piece says so in the first hundred words.
• Where a writer has a personal connection to a country or a service they are writing about, it is disclosed in the piece — it is usually a strength, and it is never a secret.

Corrections, complaints about a piece, right-of-reply requests and questions about this policy all go to hello@twotimezones.com. Put “Correction,” “Complaint” or “Editorial” at the front of the subject line and it reaches the right person faster. Response times and the other routes into the inbox are on the contact page.

This policy is reviewed whenever our practice changes — most likely when we adopt a new tool or take on a new commercial arrangement. The date at the top of the page is the date of the last substantive edit.