Two systems, never blurred
Everything you read here is one of two kinds of content, and we don't let them bleed together:
- Programmatic pages are built from the dataset. Every word is derived from data — scores, costs, rankings — and contains no hand-written per-page opinion. Change the data, the page regenerates. These are the "current facts for X" pages.
- Written guides are hand-authored by our editorial team when the value is an argument, a story, a method, or original analysis — like the safety playbook or the solo tax report. Some embed live data widgets; the prose around them is still written by a person and marked as editorial.
We keep the line sharp so you always know what you're reading: a machine-summarised fact, or a human making a case. Confusing the two is how sites end up passing off opinion as data, and we'd rather not.
Sourcing
- Every factual claim is sourced. Survey figures, safety statistics, cost data — where it comes from is part of the claim, not a footnote we hope you skip. A number we can't attribute is one we don't publish.
- Every score is dated. Safety and cost change, and an undated score quietly pretends to be current. Last-updated dates are non-negotiable across the site.
- We distinguish figures from feelings. A cited statistic and an editorial judgement are written differently and never dressed up as each other. When we're aggregating perception rather than counting something, we say so.
Corrections
We get things wrong, and when we do we fix them in the open. Our policy is simple:
- If a claim is wrong, we correct the content and note that it changed.
- We don't silently delete a mistake and pretend it never ran — for anything that materially affected a reader's decision, the correction is visible.
- Reader reports of errors are taken seriously; the community catching our mistakes is a feature, not an embarrassment.
Honest corrections build more trust than a false record of never being wrong. We'd rather be known for fixing things well than for looking infallible.
Commissioning identity content
A place doesn't respond to every woman the same way, and we won't pretend otherwise or generalise across experiences that differ. Our rule: content about a specific community's experience is commissioned from within that community. We don't write the Black solo traveller's guide, or the disabled traveller's guide, or the trans traveller's guide, from the outside and call it representative. Where we can't yet speak to an experience with real, community-grounded knowledge, we say so plainly rather than paper over the gap — a point we make openly in what our data can't tell you.
Duty of care on safety and health
Safety and health content can affect real decisions in real situations, so it carries a higher bar:
- We decompose risk honestly — petty theft, harassment, and violent crime are different risks with different odds, and we never blur them into vague reassurance or vague alarm. False comfort and false fear are both failures of care.
- We point to real help generically — local emergency numbers, embassies and consulates, established support services — rather than inventing specifics we can't stand behind.
- We do not fabricate experts. We will not attach a named doctor, a fake credential, or an invented "reviewed by Dr. X" to lend borrowed authority to health or safety claims. If a piece is reviewed, it's reviewed by our editorial team, and it says exactly that. Anything beyond our competence, we point outward to qualified sources instead of impersonating them.
These standards exist for one reason: the product is trust, and trust is spent the first time a reader catches us cutting a corner. So we try not to cut them.