Multi-active serums live or die on a single question: which of those impressive-sounding ingredients is actually present at a dose that matters? Marketing loves a long list — it implies value, comprehensiveness, science — but a label can name a dozen actives while only one or two are doing real work. This Cos De BAHA serum is a perfect case study, because its very name foregrounds retinol, and the retinol turns out to be the least important active in the bottle. Reading it honestly means reranking the ingredients by what they actually contribute.
Start with the headline, because it needs deflating. The serum advertises retinol at 0.02%, and 0.02% is, frankly, a token amount — far below the concentrations shown to drive the meaningful cell turnover that gives retinol its anti-aging reputation. It is enough to put the word on the label and add a whisper of renewal, but if you are buying this expecting retinol results, you will be disappointed. That is the central myth to drop: a 0.02% retinol serum is not a retinol treatment. Fortunately, the serum does not actually depend on it.
What does the work is the brightening pair and the bakuchiol. Tranexamic acid sits high on the ingredient list, and it is one of the genuinely effective, well-tolerated actives for interrupting the signals that create uneven pigment. Alongside it, arbutin — a gentle, naturally-derived brightener — inhibits melanin formation from a second angle. Between them, brightening and radiance are the serum's clearest, most reliable benefits, which makes the "radiance boost" name far more accurate than the "retinol" one. Then there is bakuchiol at a meaningful 2%, the plant-derived retinol alternative that supplies most of the actual gentle anti-aging and smoothing here, effectively standing in for the underdosed retinol while being kinder to the skin. So the real story of this serum is: brightening plus bakuchiol, with retinol as a garnish.
Where it gets genuinely likeable is the barrier-support layer, which a lot of brightening and anti-aging serums neglect. Ceramide NP, a skin-identical lipid, sits in the formula alongside squalane, a full centella complex (madecassoside, asiaticoside, asiatic and madecassic acids), panthenol and allantoin. That is a properly conditioning, soothing cast, and it means this serum brightens and lightly renews while supporting the barrier rather than taxing it — the opposite of the stripped, irritated feeling that aggressive actives can cause. Combined with the low retinol, it is why the serum reads as gentle and broadly wearable despite its long active list.
In use, it behaves like the well-rounded emulsion it is: a lightweight, slightly creamy serum that absorbs to a smooth, conditioned finish, with a faint floral scent and essentially no sting. There is little to no purging at this retinol dose. You apply it at night, your skin feels comfortable, and over a few weeks it looks more radiant and even, with gradual improvement in tone and texture building over a couple of months. It asks little and gives a gentle, cumulative payoff.
The trade-offs are mostly about expectations and a couple of formula notes. Beyond the token retinol already discussed, the serum is fragranced by rose and lavender botanical extracts, which give it a light scent and make it not fragrance-free — pleasant for most, but a reason for genuinely fragrance-reactive skin to patch test or skip. It contains squalane and caprylic/capric triglyceride, so it is not fungal-acne-safe. Because it contains retinol, even a trace, it is not pregnancy-safe. Most active concentrations beyond the retinol and bakuchiol, and the pH, are unstated. And it should not be stacked with additional retinoids or strong acids — with this many actives already inside, it is meant to be your single treatment step.
Value is, predictably, excellent. A serum combining tranexamic acid, arbutin, bakuchiol, ceramide, a centella complex and squalane for around fourteen dollars is a remarkable amount of well-chosen formulation for the money, even with the retinol as window dressing, and a bottle lasts two to three months. A multi-year track record and a wide review base make the low cost read as lean, direct-to-customer pricing rather than a cheapened formula. You are paying for a genuinely loaded, gentle radiance serum, and getting more than you pay for.
Who should buy it? Normal, combination and dry skin — and many sensitive types — wanting a gentle, affordable all-rounder for dullness, uneven tone and overall radiance with barrier support built in. It is ideal for people who want a low-effort single serum that does several things adequately rather than one thing dramatically. Who should skip it? Anyone shopping specifically for real retinol results, who needs a proper retinoid; the strictly fungal-acne-prone; the fragrance-reactive; and anyone pregnant or breastfeeding. Judged as the gentle, brightening, barrier-supporting radiance serum it actually is — rather than the retinol product its name implies — it is a quiet overachiever.
