The conversion chain, in one paragraph
Only one molecule does anything in skin: retinoic acid. Everything else sold as a retinoid is a precursor that has to be converted into it inside the skin, and each conversion step loses some of the dose. Retinyl esters convert in three steps, retinol in two, retinaldehyde in one, and retinoic acid — prescription tretinoin — in none, which is why it is both the most effective and the most irritating. That is the whole hierarchy. It explains why a stated 0.2% retinal is not the same proposition as a stated 0.2% retinol, and why a “retinol” whose active is actually retinyl palmitate is the weakest thing on any shelf. Our retinyl palmitate comparison covers that last trap in detail.
Where adapalene sits, and why it is the odd one out
Adapalene is not a precursor at all. It is a synthetic third-generation retinoid that binds retinoic acid receptors directly, and in the United States it is regulated as an over-the-counter drugfor acne rather than as a cosmetic. That single fact changes what you can know about it. Its strength is on the label because the monograph requires it. Its indication is on the label. Its warnings are on the label. Compare that with every cosmetic retinol on this site, where the percentage is usually absent and the federal rule on ingredient lists only orders ingredients by predominance above one percent — so a late position proves presence and nothing else.
The trade-off is that adapalene is built for acne, not for photoaging. Its evidence base is comedonal and inflammatory acne; the collagen record that makes people buy retinoids for lines was built on tretinoin and, at a longer distance, on cosmetic retinol. If your problem is breakouts, that is the right trade. If your problem is lines, it is the wrong one, and a retinal is the better buy. Our retinol vs adapalene page works through the decision properly.
What hydroxypinacolone retinoate is, and what it is not
HPR — sold here as The Ordinary’s Granactive Retinoid 2% Emulsion — is an ester of retinoic acid, and the pitch is that it binds receptors without the conversion chain and without the irritation. The citable evidence for it is one open pilot study of a fixed combination containing 0.1% HPR plus 1% encapsulated retinol in mild to moderate acne: a 41% mean reduction in acne grade, 15.3% mild-to-moderate local side effects, and the authors themselves calling for a controlled trial. That is genuinely promising and it is genuinely thin, and anyone telling you HPR is “as good as tretinoin without the irritation” is well ahead of the published record.
Also worth knowing before you read the bottle: the 2% on the front refers to the Granactive Retinoid complex, not to HPR itself. The same pattern — a large number describing the delivery material rather than the active — catches buyers out repeatedly, and our granactive retinoid comparison sets out exactly what the figure measures. We rank it fourth, as the one to try when plain retinol has already failed on tolerance rather than as a first choice.
The criteria, in order
- A published strength beats an unpublished one.Five of the six picks print a figure. That is not a coincidence — we rank disclosure, because without it you cannot step up deliberately, you cannot compare two bottles, and you cannot tell a token dose from a working one.
- The molecule’s own evidence, not the category’s. Adapalene gets credit for the adapalene meta-analysis; it does not get credit for the tretinoin collagen trials. Retinal gets credit for the tolerance study that tested retinaldehyde; it does not inherit tretinoin’s biopsy record either.
- Tolerability at equal evidence. The molecule you can use four nights a week for ten months beats the one you abandon in week three. The published retinoid record plateaus at roughly ten months of continuous use, so quitting is the main failure mode, not under-dosing.
- Cost of a ten-month run, not cost of a bottle.
- A base that suits the skin using it. A buffered serum with niacinamide and ceramides is a different product from an anhydrous oil, even at the same stated percentage.
Which one, by situation
You have never used a retinoid: The Ordinary Retinol 0.2% in Squalane, two nights a week, with a moisturizer over the top. Our beginner picks explain the ramp, and our beginner guide covers what the first six weeks actually feel like.
You are breaking out: Differin. It is a drug with an acne indication and it costs less than most serums. Read retinol for acne for how it fits around the rest of a routine.
You already tolerate a retinol and want more effect: move sideways to a retinal rather than up to a higher retinol. The Ordinary Retinal 0.2% first; Naturium Retinaldehyde 0.10% if you want a 50 ml cream-serum bottle big enough for neck and chest as well. Our retinal roundup compares the whole retinal shelf.
Retinol has already made your skin angry:the Granactive Retinoid emulsion, or the buffered CeraVe serum with niacinamide and ceramides. Both are real answers to the same problem from opposite directions — change the molecule, or change the base.
What we left out, and why
Prescription tretinoin, because it is not over the counter in the United States and this page is about what you can buy without a visit. It is the benchmark the whole category is measured against, and our retinol vs tretinoin comparison is the honest version of that gap.
Retinyl palmitate products, on substance. Three conversion steps from retinoic acid is a long way to travel, and a product marketed as a retinol whose active is an ester is the one category we will not rank.
Bakuchiol and other “natural retinol alternatives”, because they are not retinoids. They belong on our retinol alternatives page, where the evidence can be stated at its real strength rather than borrowed from this category.