Is that active doing anything?
We sell some of what is named here
Routinae sells some of the products written about here, so read this the way you would read any shop’s opinion of its own shelves. Where we think a cheaper product does the same job we say so, and where the evidence is thin we say that too. Our buying standard is published so you can check the argument against it.
A bottle says Centella. Another says Vitamin. A third has a number in the name. All three are telling you what the marketing department chose to put on the front, and none of them are telling you how much of it is in there.
The information you want is on the back, it is legally required to be there, and it is readable in about twenty seconds once you know the one rule that governs it.
The ingredient list is the only regulated part of the box
EU law requires the ingredients of a cosmetic to be listed in descending order of weight, using standardised international names, so that the same substance is called the same thing on a Korean bottle and a French one. That is why a label says aqua and not water, and it is why the list is useful at all.
There is one enormous qualification, and everything else in this article follows from it: the descending order only has to hold above one per cent. Anything present at less than one per cent may be listed in any order the manufacturer likes.
So the top of the list is genuine information and the bottom of it is essentially unordered. What you are actually reading is: these things, in this order, until roughly one per cent — and then everything else, arranged however.
Where the one per cent line usually falls
You cannot see the line, but you can estimate it. It sits at, or just after, the first preservative or the first thickener — names like phenoxyethanol, ethylhexylglycerin, carbomer, xanthan gum, or the -eths and -ates that stabilise things. Those are almost always used well below one per cent. Anything below them is in the unordered zone.
Which gives you the twenty-second test. Find the hero ingredient from the front of the bottle. Is it above that line or below it? If it is below, its position tells you nothing, because it could be at 0.9 per cent or at 0.0001 per cent and the label would look identical either way.
Three things that count as evidence, and one that does not
A number printed on the ingredient list itself. Some Korean brands print the concentration next to the ingredient in parts per million, which is unusual and worth rewarding. The Purito Seoul centella serum we sell says 34,860 ppm on the list, which is just under three and a half per cent — not on the box, where anyone can print anything, but in the regulated part. That is the strongest signal available to a shopper.
Position above the one per cent line. In the Anua niacinamide serum, niacinamide is third and tranexamic acid fourth, with two more tone ingredients below them. You still do not know the exact percentage, but four related ingredients stacked near the top is a real formula rather than a decorative one — which is also why it is the most likely thing in this shop to sting.
A very short list. The I'm From mugwort essence has three ingredients in total. The mixsoon bean essence has seven, with the fermented extract fourth. When there is almost nothing else in the bottle, there is nowhere for the hero ingredient to hide. Short lists are the most honest packaging in this category and they photograph terribly, which is presumably why there are so few of them.
What does not count: a percentage on the front of the box. A figure in a product name is the brand's claim about its own formula. It is not part of the regulated label and you cannot check it. The COSRX snail essence has 96 in the name; snail secretion filtrate is genuinely first on a list of thirteen, which is consistent with it, but the 96 itself is the brand's figure and not one you can read off the bottle. We sell the Beauty of Joseon glow serum, whose brand publishes a propolis percentage; we have not been able to verify it against the label, so we do not repeat it on the product page.
An example against our own shelf
The Round Lab Birch Juice cream is named after birch sap. Birch sap is on the ingredient list at 10,000 ppm — one per cent — and it sits eighth. So the name is a flavour rather than a formula.
Is that a scandal? No. It is a good, plain, unfragranced cream that a lot of people get on with, and it is priced like one. The name is simply not the reason to buy it. That is the useful conclusion in most of these cases: the hero ingredient is often decorative, and the product is often still fine. What you are protecting yourself against is paying a premium for the name.
The Numbuzin No.5 is the same lesson from the other direction. It is sold as a vitamin serum, but the list opens with lingonberry extract, niacinamide and tranexamic acid, and the ascorbic acid and its derivatives sit a long way down. Read as a vitamin C serum it will disappoint. Read as what it is, it is a reasonable thing to own.
When a low position is exactly right
Low is not automatically a con. Some ingredients are meant to be down there.
- Retinoids. In the Beauty of Joseon eye serum, retinal is twenty-eighth of thirty-six ingredients. That is the point. A high dose around the eyes would be a design fault, not a feature.
- Preservatives. You want them present and you want them low. An unpreserved water-based cosmetic is a worse problem than any of this.
- Fragrance and essential oils. Low is better. Absent is better still if you react to them — which is a separate can of worms.
- Anything with a legal maximum. Several actives have concentration limits in EU law precisely because more is not better.
The honest limits of reading a label
This is a filter, not a verdict. Position is not concentration. Concentration is not delivery — the same percentage in two different bases will not behave the same way. And for several of the ingredients that sell the most bottles, the published evidence is thinner than the reputation: snail is a decade-old favourite that rests on how skin feels afterwards rather than on much that has been demonstrated in a trial, and we would rather say that than borrow the brand's confidence.
Where the evidence is thin, our product pages say so. Where a number is the brand's rather than the label's, we say whose it is. That is the whole of our position: read the back, be sceptical of the front, and buy the base rather than the name.