Corrections

A review is only as trustworthy as the publication's willingness to mark its own errors. Premium Travel Review treats the corrections log as part of the editorial product, not an afterthought to it. When a published verdict is found to rest on a fact that was wrong, we record the change here in plain language, with the date it was made and a description of exactly what changed. The entry stays on this page permanently. We do not quietly revise a score and move on, and we do not delete corrections once a reasonable interval has passed. The record is the point.

On launch this log is empty, and that is by design. The publication began under a clean editorial standard, and there is nothing yet to disclose. An empty corrections page on a new title is not a sign that nothing can go wrong; it is the honest starting state of a desk that has committed to logging the first material error the moment one is identified. As the corpus grows, this page will too, and a reader will be able to judge the publication not only by what it gets right but by how it behaves when it gets something wrong.

We draw a firm line between a silent fix and a material correction. A silent fix is a change with no bearing on the meaning of a verdict: a misspelled name corrected in passing, a broken link repaired, a typographic slip, a clumsy sentence rewritten for clarity. These are made without a log entry, because they alter nothing a reader relied on when making a decision.

A material correction is different. If a change affects a published score, a price or rate paid, a date, a named entity — a property, a distillery, a chef, a vintage — or any factual claim that a reader might have weighed in a purchase or booking decision, it is logged here permanently. The test is simple: if the corrected fact could have changed what a reader did, it belongs in this record. When a correction alters the headline verdict in a comparison, both the affected entry and the grid that contained it are noted.

Before the current corpus went live, the desk re-verified every drafted review against external sources. Four drafts were found to rest on premise-level errors and were withdrawn rather than corrected — a review built on a fact that never happened cannot be salvaged by amending a sentence. In the interest of the same transparency the post-publication log promises, those withdrawals are recorded here.

• Atomix, Manhattan — withdrawn. The draft was built around a promotion to a third Michelin star. Atomix holds two Michelin stars (awarded 2021 and retained since); the three-star rating it does hold is from a separate newspaper system. The premise was a non-event, so the piece was withdrawn rather than published.
• Sézanne, Tokyo — withdrawn. The draft described Sézanne as a two-star kitchen and presented Daniel Calvert as its current chef. Sézanne in fact holds three Michelin stars (awarded 2024), and Calvert had departed on 31 March 2026, before the draft's own date. Two load-bearing facts were wrong; the piece was withdrawn.
• Rosewood São Paulo — withdrawn. The draft reviewed the property as a fresh opening. Rosewood São Paulo opened in early 2022, roughly four years earlier, and the draft also misstated the room count and the entire food-and-beverage line-up. Withdrawn.
• The Twenty Two, London — withdrawn. The draft dated the club's opening to 2025 (it opened in 2022), misattributed its founders, overstated the membership fees by roughly three times, and named the wrong restaurant chef. Withdrawn.

These were caught before publication and never carried a public verdict; the same standard will apply to anything that slips through after it.

No post-publication corrections have been logged since the current editorial standard took effect (Version 2026.1).

If you believe a published claim is wrong, tell us. Email corrections@premiumtravelreview.com with the URL of the page in question and a clear statement of the disputed claim — ideally with a source we can check. We read every message that arrives at that address. A reader who flags a genuine error is doing the publication a service, and we treat such notes as a priority rather than a nuisance.

When a correction is warranted, the desk verifies the underlying fact independently before changing anything, updates the affected page, and records the change in this log with the date and a description of what was altered. Where a correction is significant enough to move a verdict, we say so explicitly in the entry rather than burying the change inside revised copy. For the full procedure — how a flag is triaged, who reviews it, and how the affected page is annotated — see the corrections workflow.

This is a slower, more public way to handle mistakes than most outlets choose. We think it is the only way consistent with asking readers to trust a number out of twenty. A publication that scores premium products against a fixed standard inherits an obligation to hold itself to one as well, and this page is where that obligation is met in public.