How we review

A FashionScarlet review is an investigation of a brand from the outside. We don't currently buy and test products ourselves, and we don't pretend to. What we can do is read everything a careful buyer would want to read, check it against the brand's own terms, and tell you what it adds up to.

Three kinds of statement

The most important habit in our reviews is keeping three things apart. You'll see them marked in the text of every review written to our current standard:

Checked

Something we verified ourselves against a primary source, usually the brand's own published policy, specification or a reputable report, on the date shown at the top of the review.

Brand says

Something the brand claims about itself or its products that we haven't been able to verify independently: how a fabric performs, where it's sourced, what a formula does.

Buyers report

A pattern in what customers say after buying, drawn from public reviews. It's evidence of experience, weighted by how many people say it, how recently, and where.

Unmarked text is our analysis: what we think those facts mean for you.

What we look at

• Materials and construction: published fabric compositions, leather and metal sourcing, certifications and what they actually certify.
• Fit and sizing: the brand's size guidance, and what buyers report about how it runs.
• Price and value: where the brand sits against comparable labels. We don't print specific prices, because they change too often to keep accurate.
• Shipping and duties: costs, carriers, delivery-duty-paid terms for cross-border orders.
• Returns: window length and when it starts, refund versus store credit, fees, exchanges, final sale and international terms.
• Customer service and complaints: recurring problems in recent reviews, and whether the company responds.
• The company: who owns it, any recent changes, and how transparent it is about manufacturing and sustainability claims.

How we weigh buyer reviews

Review scores are easy to quote and easy to misread. Before we rely on one we ask:

• How many reviews, and how recent? A score from 100,000 reviews and one from 80 are different kinds of evidence.
• Who hosts them? Reviews on a brand's own site are moderated by the brand. Independent platforms are more open, but skew towards people with a grievance.
• Does the company invite reviews? Profiles that actively invite customers tend to score higher and fill up faster.
• Is it a pattern? We report complaints that repeat across many reviewers, not single bad experiences.

Evidence ratings

Instead of a star score, each review carries an evidence rating. It tells you how much our conclusion rests on:

Strong

The brand's terms and specifications are clearly published and checked, and there's a large, recent public record of buyer experience.

Moderate

The key facts are checked, but the buyer record is small, old, one-sided or largely hosted by the brand.

Limited

Important questions can't be answered from public evidence. We'll say which ones.

Why there are no star scores

A score out of ten suggests a precision a research-based review doesn't have, and it invites comparisons between brands that were never measured on the same scale. We'd rather tell you in a paragraph what works, what doesn't, who a brand suits and who should skip it. That's what "Our take" at the top of each review is for.

Keeping reviews current

Each review shows the date its key facts were last checked. We re-check when a brand changes its policies or ownership, or when buyer reports shift. If you spot something out of date, tell us.