The research behind what we say.
We make specific claims on this site: smoother texture and more even tone over three months, hydration up by as much as 96% over eight weeks, collagen and elastin raised by a plant extract that works where retinol works, an acid that clears as much acne in four weeks as benzoyl peroxide. Here is what those claims are built on. Every one is a published, peer-reviewed study of the ingredient itself, and every one is linked so you can read it yourself rather than take our word for it.
Glycolic acid
The acid in Resurface Toner, at 5%, a few evenings a week.
Thibault, Dermatologic Surgery, 1998
Thibault PK, Wlodarczyk J, Wenck A. Dermatol Surg 1998;24(5):573-577. 75 volunteers, daily 5% glycolic acid for 3 months, double blind and placebo controlled. Statistically significant improvement in skin texture and discolouration.
Read the study
Salicylic acid
The acid in Purify Toner. It goes down inside the pore rather than over it, and the trial below put it head to head with the strongest thing sold over the counter for acne.
Zheng, Cutaneous and Ocular Toxicology, 2019
Zheng Y, Yin S, Xia Y, et al. Cutan Ocul Toxicol 2019;38(1):48-54. Randomised, split-face, single-centre trial; 31 patients with mild to moderate acne completed it. 2% supramolecular salicylic acid cream on one side of the face against 5% benzoyl peroxide with 0.1% adapalene gel on the other, for 28 days. Total lesions fell 44.1 percent on the salicylic acid side and 45.6 percent on the benzoyl peroxide side, a difference that was not statistically significant. Papules and pustules fell 47.9 percent against 49.8 percent.
Read the study
Hyaluronic acid
Two molecular weights in Quench Serum, twice again in Barrier Cream, and on the list of Balance Gel, Brighten Serum and Hydrate Toner. The lighter weight is the one that does the measurable work.
Pavicic, Journal of Drugs in Dermatology, 2011
Pavicic T, Gauglitz GG, Lersch P, et al. J Drugs Dermatol 2011;10(9):990-1000. 76 women, 0.1% hyaluronic acid twice daily for 60 days. Wrinkle depth down up to 40 percent and skin hydration up to 96 percent with low molecular weight hyaluronic acid.
Read the study
Glycerin and aloe
The two humectants that turn up in nearly every formula in the range. Both trials are vehicle-controlled, which is the part that matters: they compare the ingredient against the same cream without it, not against nothing.
Breternitz, Skin Pharmacology and Physiology, 2008
Breternitz M, Kowatzki D, Langenauer M, Elsner P, Fluhr JW. Skin Pharmacol Physiol 2008;21(1):39-45. 24 patients with atopic dermatitis, randomised and double blind, treated twice daily for 4 weeks with a 20% glycerol emollient against its own glycerol-free vehicle. Stratum corneum hydration was significantly improved and epidermal barrier function restored under the glycerol cream compared with the placebo. Erythema and SCORAD did not separate between the arms.
Read the studyDal'Belo, Skin Research and Technology, 2006
Dal'Belo SE, Gaspar LR, Maia Campos PM. Skin Res Technol 2006;12(4):241-246. 20 female subjects, formulations carrying 0.10%, 0.25% or 0.50% freeze-dried Aloe barbadensis extract applied daily to the volar forearm and compared against the same vehicle without it. After a single application the 0.25% and 0.50% formulations raised the water content of the stratum corneum; after 2 weeks of daily use all three did, measured by Corneometer. Transepidermal water loss did not change, so the mechanism is humectant rather than occlusive.
Read the study
Vitamin C and ferulic acid
Brighten Serum pairs 2% ascorbyl glucoside with 1% ferulic acid, and the three papers below are the reason it is built that way. The first tests ascorbyl glucoside at 1.8%, which is about what we use. The second takes it to 28% and fades sun spots over 24 weeks. The third is the ferulic acid paper the whole category is built on.
Jacques, International Journal of Cosmetic Science, 2021
Jacques C, Genies C, Bacqueville D, et al. Int J Cosmet Sci 2021;43(6):691-702. A formulation carrying 1.8 percent ascorbic acid 2-glucoside compared against one carrying 15 percent pure ascorbic acid, on fresh viable human skin explants and a reconstructed human epidermis model. Skin metabolised the glucoside back to ascorbic acid. On all three markers of oxidative stress - malondialdehyde, superoxide dismutase and catalase - the protection from 1.8% ascorbyl glucoside matched that from 15% ascorbic acid. Unlike ascorbic acid, the glucoside was chemically stable in storage and needed no pH below 3.5.
Read the studyTakada, International Journal of Molecular Sciences, 2024
Takada M, Numano K, Nakano M, Yamamoto A, Imokawa G. Int J Mol Sci 2024;25(24):13453. Double-blind half-face trial in 27 Japanese women with solar lentigos: a lotion carrying 28% ascorbyl glucoside-arginine complex on one side of the face and the same lotion without it on the other, twice daily for 24 weeks. Pigmentation scores fell significantly against the placebo side at 24 weeks. Lightness rose and melanin index fell significantly against the placebo side at both 12 and 24 weeks. Two of the paper's authors are employed by the company selling the lotion, which the paper declares.
Read the studyLin, Journal of Investigative Dermatology, 2005
Lin FH, Lin JY, Gupta RD, et al. J Invest Dermatol 2005;125(4):826-832. 0.5% ferulic acid added to a topical solution of 15% L-ascorbic acid and 1% alpha-tocopherol, applied to shaved pig skin for 4 days before solar-simulated irradiation. Ferulic acid improved the chemical stability of both vitamins and doubled photoprotection, from 4-fold to approximately 8-fold, measured by erythema and by sunburn cell formation. Thymine dimer formation and caspase-3 and caspase-7 induction were also reduced.
Read the study
Bidens Pilosa
The plant retinol alternative in Renewal Serum, at 2%.
Dieamant, Evidence-Based Complementary and Alternative Medicine, 2015
Dieamant G, Pereda MCV, Nogueira C, et al. Evid Based Complement Alternat Med 2015;2015:280529. Standardised Bidens pilosa CO2 extract raised collagen by 32.5 percent and elastin by 25.2 percent in human fibroblasts, acting through retinoid and rexinoid receptors.
Read the study
Zinc PCA
The sebum regulator in both oil-control toners. The work below is on cultured human skin cells and shows how it gets in the way of the enzyme that takes collagen apart.
Takino, International Journal of Cosmetic Science, 2012
Takino Y, Okura F, Kitazawa M, Iwasaki K, Tagami H. Int J Cosmet Sci 2012;34(1):23-28. Cultured normal human dermal fibroblasts. Zinc PCA suppressed UVA-induced activation of activator protein-1 and reduced production of matrix metalloproteinase-1, the enzyme that degrades collagen in photoaged skin. It also raised expression of the ascorbic acid transporter SVCT2, and production of type I collagen rose with it. Cell culture, not skin, and the work was carried out at Ajinomoto, who manufacture Zinc PCA.
Read the study
Sea-buckthorn and watercress
Two of the harder-working botanicals in the range, and the two papers below are both animal studies: rats for the sea-buckthorn oil, mice for the watercress. We say so on every page that mentions either of them.
Gore, Indian Journal of Pharmacology, 2024
Gore DD, Sharma N, Mishra N, et al. Indian J Pharmacol 2024;56(2):120-128. Nanoemulsion cream and gel of Hippophae rhamnoides (sea buckthorn) fruit oil, applied topically to excision wounds in female Sprague-Dawley rats. By day 8 the cream had contracted 78.96 percent of the wound, the gel 72.59 percent, against 62.29 percent for the positive control, 1% framycetin. Neither formulation caused acute dermal toxicity. Rats, not people.
Read the studyCamponogara, Journal of Ethnopharmacology, 2019
Camponogara C, Silva CR, Brusco I, et al. J Ethnopharmacol 2019;229:190-204. Croton-oil irritant contact dermatitis in mice. A 3% Nasturtium officinale (watercress) leaf extract gel applied topically inhibited acute ear oedema by 81 percent, against 70 percent for a dexamethasone gel in the same model, and cut inflammatory cell infiltration by 60 percent. The effect ran through the glucocorticoid receptor and the NF-kB pathway. Repeated application caused no adverse effects. Mice, not people.
Read the study
Calendula
The largest trial on this page by a distance, and the only phase III one: 254 patients, and calendula beat the cream that was the reference treatment for inflamed skin.
Pommier, Journal of Clinical Oncology, 2004
Pommier P, Gomez F, Sunyach MP, et al. J Clin Oncol 2004;22(8):1447-1453. Phase III randomised trial, 254 patients receiving postoperative radiotherapy for breast cancer, allocated to calendula ointment (126) or to trolamine (128), the reference topical agent, after every session. Acute dermatitis of grade 2 or higher occurred in 41 percent of the calendula arm against 63 percent of the trolamine arm (P<0.001). Calendula patients also had significantly less radiation-induced pain and fewer interruptions to treatment.
Read the study
Peony root
For flushing and blotchiness. The paper below is a mouse model, and the claim we make from it is written as one.
Bai, European Journal of Dermatology, 2020
Bai X, Yu C, Yang L, et al. Eur J Dermatol 2020;30(3):243-250. Imiquimod-induced psoriasis model in mice. Topical paeoniflorin, the main active of Paeonia lactiflora root, significantly reduced epidermal thickness, dermal lymphocyte infiltration, antimicrobial peptides, pro-inflammatory cytokines and keratin 17, the protein that drives keratinocyte overgrowth. In cultured human keratinocytes it suppressed NF-kB activation and cut proliferation dose-dependently. Mice and cell culture, not people.
Read the study
How people buy and abandon skincare
The behaviour our whole model is built to interrupt. All commissioned research with a named survey house and a stated sample.
The Body Shop, survey of 2,000 skincare users, 2022
74 percent do not finish a product before throwing it out. Reasons given: wrong for their skin (27 percent), something newer replaced it (26 percent), it stopped working (18 percent). Two abandoned bottles per person on average.
Read the studyCensuswide for Kenvue, 2,001 adults, 2025
60 percent do not keep a consistent skincare routine. Barriers named: cost (30 percent), motivation (25 percent), busy schedules (23 percent), information overload (20 percent), not knowing which products suit them (15 percent).
Read the studyAmerican Academy of Dermatology
AAD public guidance: use a product for 6 to 8 weeks before judging it, and expect up to 3 months for full results. Switching products every week or two irritates skin and causes breakouts.
Read the study