How we choose

Ten questions, forty-five places, and one product out for every one in.

This is the standard we buy against. It is published so you can check our work, and so we cannot quietly relax it the moment something popular comes along.

45 Products in the catalogue A hard cap. Not a milestone we are working past.
1 in, 1 out The delisting rule Nothing is added until something has been removed.
~300 Products tested to get here Most of them are not here. That is the point of testing them.

The cap, and why it stays a cap

Forty-five is not the number of products we could sell. It is the number we can hold an opinion about. Past that we would be recommending things from a spreadsheet, and you would be able to tell.

The cap does something else that we value more: it makes every addition expensive. To bring in a new cleanser, we have to be willing to lose one we already sell. That argument is much harder to win than “add it and see”, which is how a shop ends up with two hundred products and no view.

One in, one out

When a product joins, another leaves the same week. We write up what left and why, so the catalogue has a history rather than just a state.

Products leave for four reasons: something else does the job better, the price stopped making sense, the formula changed, or supply became unreliable. The fourth is the most common and the least flattering, and we would rather say so than invent a nobler reason.

If something you rely on is delisted, ask us what to use instead. We will tell you even when the answer is a product we do not sell.

The ten questions

A product has to pass all ten. Nine out of ten is a no, and it is usually the tenth one that would have annoyed you six weeks in.

01

It has to fix something we can name

If we cannot write one plain sentence saying what changes and roughly how long it takes, the product is not ready to be sold. “Glow” is not a sentence.

02

The ingredient list has to match the claim

A hero ingredient sitting at the very bottom of the list, below the fragrance, is a marketing decision rather than a formulation one.

03

It has to fit a routine we already sell

Nothing enters if it only works alongside three products we do not stock. That is precisely how catalogues grow to two hundred.

04

The cheaper one gets the benefit of the doubt

If something already in the shop does the same job for less, the expensive product has to justify the gap in words we can say out loud. Most cannot.

05

Someone here has used it for eight weeks

Not a sachet and not a press kit. A full-size product, on one face, through one season, alongside the rest of a normal routine.

06

The claim has to be legal and supported

EU rules require cosmetic claims to be truthful, evidenced and fair. If a claim only survives with an asterisk under it, we will not repeat it.

07

The packaging has to be EU-compliant and obtainable

Responsible Person on the label, INCI list, batch code, and a durability date or period-after-opening symbol. No exception for a product that would sell well.

08

The supplier has to answer questions

Documentation in writing, within a working week. Silence is an answer, and it is a no.

09

The fragrance load has to suit the audience

For anything we would put in front of reactive skin, we read the declared fragrance allergens rather than the front of the box.

10

It has to survive the second order

Repurchase decides who stays. A product people buy once and never again is a good advertisement and a bad product.

What we reject on principle

These are not close calls. No margin, no demand and no supplier relationship buys past this list.

Skin-lightening or whitening framing Including the polite versions of it. We will describe evenness of tone; we will not sell the idea that lighter is better.
Before-and-after photography Lighting, angle and editing do most of the work. It is not evidence, so we do not use it and we do not buy from people who lead with it.
Anything sold on a medical claim Treats, cures, heals, prevents. That describes a medicine. A cosmetics shop has no business implying it, however carefully it is worded.
Miracle, instant, permanent Skincare is slow and reversible. A product that promises otherwise is either overstating itself or is not a cosmetic.
Ten-step sets sold as one decision We will happily sell you the four steps that earn their place. Bundling the other six alongside them is a pricing trick.
Strong actives marketed to children Retinoids and high-strength acids aimed at ten-year-olds. Not from us, at any margin, however well they are selling elsewhere.
Urgency mechanics Countdown timers, stock scares, “three people are viewing this”. Our theme blocks them in code, so nobody can add them later by accident.
Products whose full ingredient list we cannot see first If it is not shareable before the invoice, it will not be shareable after it.

This is what survived it.

Forty-five products. Where one of them is wrong for you, the product page says so before you buy it.

See the catalogue