How Do You Simplify a Skincare Routine That Stopped Working?

Take away one leave-on product at a time, starting with whatever you added most recently, and keep everything else unchanged for two weeks. Stopping your entire routine at once feels decisive but tells you almost nothing, because you remove the comparison that makes the answer readable.

Every November the heating goes on across Hongdae and our appointment book changes shape. People arrive holding a phone with eleven products listed on it, certain that one of them has turned on them. Almost nobody can say which one, because all eleven arrived in the same fortnight.

Why remove one product instead of all of them?

Because removing everything destroys the only information you have. A routine can irritate you while no single product in it would — that part is real, and worth taking seriously. In one human study of irritant reactions, serial dilutions of sodium lauryl sulfate, sodium hydroxide and benzalkonium chloride were patch-tested on the backs of volunteers, responses were followed for seven days using transepidermal water loss and laser Doppler flowmetry, and separate protocols tested repeated exposure. Doses too weak to produce a visible reaction on their own caused measurably more irritation when applied twice at one-day intervals. We could only read the published abstract, and it was truncated — the number of volunteers was not stated in the part available to us.

That is the mechanism behind “nothing I use is harsh, but my face is angry.” It is also why a full stop is the wrong opening move: if you take all eleven away and your skin calms down, you have confirmed that eleven was too many — which you already suspected — and learned nothing about what to leave out when you rebuild.

So we go in reverse order of arrival. The product you added last is the one your skin has had the least time to accommodate, and the one you can remove without breaking something the rest of the routine depends on. Cleanser, one plain moisturiser and sunscreen are the floor, not candidates. They stay.

How long should you wait before you decide?

Two weeks to read a removal; six to eight to judge anything you are thinking of keeping. The acute part of barrier repair is fast in principle. In an animal study, sixteen hairless guinea pigs had the stratum corneum removed by microdermabrasion; by twelve to twenty-four hours, dye penetration was largely blocked and electrical resistance was no longer significantly different from untreated skin. The authors say plainly that human repair is likely slower, and they did not measure transepidermal water loss — so take the shape of that curve, not the hours.

If what you suspect is an allergy rather than irritation, there is a standard way to check. The open application test is usually described as a 5 × 5 cm patch on the inner forearm or behind the ear, product applied twice a day for one week, watching for redness, dryness or itch.

Judging an active is a different timescale entirely. In an eight-week split-face trial, fifty women were enrolled — thirty-seven completed — and used 0.025% tretinoin nightly after two weeks of moisturiser preconditioning. Dryness and peeling rose on both cheeks early. On the side using a moisturiser containing niacinamide, panthenol and tocopheryl acetate, dryness returned to baseline by the end of the eight weeks; on the other side it stayed elevated for the entire study. Worth knowing: two of the authors worked for the company that made both creams, and that company funded the trial.

Is the product irritating you, or is your barrier just loud right now?

You can usually tell by whether the reaction follows the product or follows the day. Here is the rule we are comfortable handing over. A reaction that follows the product shows up within roughly twenty minutes of applying it, in the same place each time, and tends to be worst wherever you apply first. A reaction that follows your barrier arrives late in the day, moves around, is worse after a hot shower, a cold walk or a long flight, and turns up even on mornings when you used nothing at all. The first is a product problem. The second is a condition problem, and swapping products will chase it in circles for months.

What we do not publish is the order. When someone arrives with three plausible suspects, a pigmentation goal they do not want to lose, a prescription they would rather not interrupt and nine days left in Seoul, deciding what comes out first and what is reintroduced on which day is the actual craft — and it is different for every face. That part belongs in a consultation, and we cannot write it into a blog post honestly.

Know your own skin before you buy anything

Ingredients should enter your routine in stages, pitched to the state your skin is in this month rather than to how good the ingredient looks in the abstract. Exercise is good for people. It is still not what you prescribe to someone whose body is in the middle of fighting something serious — there, the honest prescription is rest, and pushing harder only borrows against a recovery that has not happened yet. Skin with a broken barrier is in that position, and the useful move is subtraction: fewer leave-on products, one bland moisturiser, sunscreen, and a short list of things you are deliberately not doing this week. Add back only once the baseline has gone boring.

None of this is a diagnosis or a treatment. Redness that will not settle, swelling, oozing, pain, or anything that spreads is a reason to see a doctor, not to run a two-week experiment.

The case against some of this

Three things I said above are weaker than they sound.

  • “Fewer products is safer.” The strongest counter-evidence I found is a 24-week, single-centre trial in 74 subjects using a six-product regimen — cleanser, growth factor serum, brightener, retinol complex, moisturiser and SPF 30+. Mean scores for erythema, dryness, burning, stinging and itching stayed below “mild” at every visit; burning and stinging rose significantly at week four but by under 0.12 on the scale and did not persist, and erythema improved from week four onward. There was no control group, and the manufacturer funded the study with two employees among the authors — so I would not lean on it hard. It still undercuts a simple “six is too many.” Narrowed: the count is not the variable. Speed of introduction and unmanaged overlap are.
  • “Two weeks is enough to read a result.” The tretinoin trial above argues against me: dryness and peeling stayed above baseline for all eight weeks on one side. A two-week verdict there would have condemned a product that was working. Narrowed: two weeks is a window for reading a subtraction, not for deciding whether an active earns its place.
  • “Your routine is stacking subclinical irritation.” That evidence comes from occluded patch tests of surfactants and alkali on backs, not cosmetic serums on faces — and the same study found that rechallenging previously tested sites with subclinical doses did not increase the response, so the stacking is not unconditional. The recovery timings I quoted are from sixteen guinea pigs. Narrowed: the mechanism is plausible enough to act on cautiously; the numbers do not transfer to your face, and I am not claiming they do.

Questions we get asked

Should I stop my whole routine for a week?

Not as a first move. A full stop can calm visible irritation, but it leaves you with no idea which product to avoid, so most people rebuild the same routine and arrive back in the same place. Keep cleanser, a plain moisturiser and sunscreen, and subtract the rest one at a time.

Can I patch test on my face instead of my arm?

You can, but behind the ear or along the jaw is a better compromise than the cheek. Facial skin is thinner and more reactive, and a false alarm there costs you a product you might have tolerated. The open application test as usually described is twice daily for a week on a 5 × 5 cm patch.

Does “barrier repair” mean anything on a Korean label?

Only in one narrow sense. Korea's functional cosmetics system recognises eleven claim categories, one of which is restoring skin barrier function to relieve itching, and a product making a functional claim must clear an MFDS evaluation or reporting route before sale. A cream that went through neither is using “barrier” as marketing language. That does not make it a bad cream — it makes the phrase uninformative.

If you are in Seoul

At the studio in Hongdae we spend the first part of a session on exactly this: what comes out, in what order, and what gets to stay. If your routine has quietly grown past what your skin is asking for, a 1:1 consultation is the short version of a long experiment. Bring the actual bottles, or photographs of the ingredient lists.

Sources

Sources for the case-against section

  • Long-term Multi-product Facial Regimen in Subjects with Moderate-to-severe Photodamage and Hyperpigmentation (JCAD) — 74 subjects, single-centre, 24 weeks, six-product regimen, no control group, manufacturer-funded. Full text read. Used against the claim that fewer products is inherently safer.
  • The Cutis retinization trial above, used a second time — against the claim that two weeks is long enough to judge an active.
  • The irritant study above, used against itself: its own rechallenge arm found no increased response at previously tested sites, and its exposures were occluded surfactants on backs rather than cosmetics on faces.
Next
Next

Ceramide Skincare: Does Any of It Actually Get Into Your Skin?