Four databases that give a verdict split on 16 of 20 best sellers, and 6 got opposite answers. Skincarisma's score is built on EWG's own ratings, yet the two rank these products almost unrelatedly (−0.07).
SlipperSkin finding
Read on 21 October 2026, four public ingredient databases (EWG Skin Deep, INCIdecoder, SkinSAFE and Skincarisma) disagreed on whether 16 of 20 best-selling skincare and makeup products were clean or flagged; 6 got opposite verdicts, and EWG's own listings of the same product scored differently for 8.
Mostly not, and for reasons you can check yourself:
On 16 of 20 products. We asked each database the same plain question for every product: does it treat this as clean, or flag it? EWG counts as clean at a hazard score of 1 or 2, its own low band; SkinSAFE at its top rating of 100; Skincarisma at its "Excellent" label; INCIdecoder, which gives no overall score, when it rates none of the ingredients "icky". Only 4 products came back the same on all four.
CeraVe Moisturizing Cream (EWG 3, Skincarisma 90, SkinSAFE 100); CeraVe Hydrating Cleanser (EWG 3, Skincarisma 98, SkinSAFE 100); Neutrogena Hydro Boost (EWG 4, Skincarisma 91, SkinSAFE 100); EltaMD UV Clear SPF 46 (EWG 4, Skincarisma 90, SkinSAFE 100); Maybelline Fit Me Matte (EWG 4, Skincarisma 92, SkinSAFE 100); Glossier Boy Brow (EWG 4, Skincarisma 99, SkinSAFE 100). Each sits in EWG's moderate band while Skincarisma calls it "Excellent" and SkinSAFE gives its top rating. The starkest single case runs the other way round too: Laneige Lip Sleeping Mask is the only product EWG puts in its high-hazard band (7, on "Limited" data, last updated January 2020), and Skincarisma rates it 96; Vaseline Petroleum Jelly gets EWG's best score of 1 and Skincarisma's lowest in this set, 55.
| Product | EWG Skin Deep (1 to 10) | INCIdecoder (icky ingredients) | SkinSAFE (rating) | Skincarisma (/100) | Verdicts |
|---|---|---|---|---|---|
| CeraVe Moisturizing Cream | 3 | none | 100 | 90 Excellent | split |
| CeraVe Hydrating Cleanser | 3 (listings 2 to 3) | none | 100 | 98 Excellent | split |
| LRP Toleriane Double Repair | 2 (listings 2 to 3) | none | 100 | 100 Excellent | agree |
| The Ordinary Niacinamide 10% | 2 | none | 100 | 96 Excellent | agree |
| Neutrogena Hydro Boost | 4 (listings 2 to 5) | Fragrance | 100 | 91 Excellent | split |
| Cetaphil Gentle Cleanser | 3 (listings 1 to 3) | none | 91 | 100 Excellent | split |
| Paula's Choice 2% BHA | 2 | none | 100 | 94 Excellent | agree |
| EltaMD UV Clear SPF 46 | 4 | none | 100 | 90 Excellent | split |
| Supergoop Unseen SPF 40 | not listed | none | 91 | 95 Excellent | split |
| Laneige Lip Sleeping Mask | 7 | Fragrance/Parfum, Alcohol | 91 | 96 Excellent | split |
| COSRX Snail 96 Essence | 1 | none | 100 | 96 Excellent | agree |
| Aquaphor Healing Ointment | 2 | none | 91 | 94 Excellent | split |
| Vaseline Petroleum Jelly | 1 | none | 100 | 55 Good | split |
| Fenty Pro Filt'r Foundation | 3 (listings 3 to 4) | Parfum/Fragrance | 91 | 99 Excellent | split |
| Maybelline Fit Me Matte | 4 (listings 4 to 5) | Alcohol Denat. | 100 | 92 Excellent | split |
| NARS Radiant Concealer | 2 (listings 2 to 4) | none | 91 | 97 Excellent | split |
| Glossier Boy Brow | 4 | none | 100 | 99 Excellent | split |
| Drunk Elephant Protini | 4 | none | 91 | 99 Excellent | split |
| Olay Micro-Sculpting Cream | 6 | Fragrance | 91 | 95 Excellent | split |
| Clinique Moisture Surge 100H | 5 (listings 2 to 5) | none | 91 | 98 Excellent | split |
Because they add the ingredients up differently. Skincarisma publishes its whole formula on its methodology page: it starts at 100 and subtracts for the share of ingredients EWG rates moderate or high. EWG weights each product toward its most concerning ingredient, and its FAQ describes the rating as the relative level of concern from the ingredients compared with other formulations. Yuka and Think Dirty go further and let the worst ingredient cap the product. The same page that publishes Skincarisma's formula calls EWG "widely criticised as alarmist", which is a fair warning about inputs Skincarisma itself uses.
On real products the effect is the chart below: if the two databases agreed, the lines would run roughly parallel. They cross everywhere.
Often not. Formulas change, and a database keeps whatever list it was given. Skincarisma's CeraVe Moisturizing Cream page still lists methylparaben and propylparaben, which the other four do not, and that is what triggers its "Contains paraben" concern. EWG's Cetaphil Gentle Skin Cleanser listing still shows the older formula with parabens and sodium lauryl sulfate. SkinSAFE lists five parabens in Olay's Micro-Sculpting Cream; SkinSort lists none. Even one database disagrees with itself: EWG keeps several listings of the same product, and for 8 of the 14 products where it does, the listings score differently.
Not always, even on a yes or no question. Leaving out SkinSAFE, whose "Fragrance Free" mark is a certification rather than a claim that a product contains fragrance, the other four disagree on 5 of 20: Laneige Lip Sleeping Mask, Fenty Pro Filt'r Foundation, NARS Radiant Concealer, Drunk Elephant Protini, Olay Micro-Sculpting Cream. With SkinSAFE included it is 10 of 20. Some of that is rules: on 21 October 2026 SkinSort counted linolenic acid and phenylpropanol as fragrance ingredients in Drunk Elephant Protini; re-read on 23 October 2026 it names only phenylpropanol, and still tags the cream not fragrance-free (more in SkinSort vs INKEEDecoder). Some is a page contradicting itself: Skincarisma's Laneige Lip Sleeping Mask page says "Fragrance-free" at a glance while its own ingredient list includes fragrance.
Yes. We repeated the read on 88 products from 48 brands (read 21 and 23 October 2026): the databases that give a verdict split on 80 of 88, across 46 of the 48 brands. Every brand has its own page with the product-by-product grid on the brand files, and the 20-product table is the CSV linked below. A hazard score also is not a legal line: see how EWG scores the ingredients California bans.
The one that answers your question, read against the box in your hand. They are not competing answers to one question. EWG asks how concerning the ingredients are as hazards. SkinSAFE, built with Mayo Clinic dermatology, asks whether a product is free of the most common contact allergens. INCIdecoder, now renamed INKEEDecoder after a trademark dispute, explains what each ingredient does and refuses to score the whole. None of them weighs how much of an ingredient is in the product, which is what the safety assessments by the Cosmetic Ingredient Review panel and the EU's Scientific Committee on Consumer Safety are built around. We report what each database says; we do not issue our own safety verdict on any ingredient.
1. Is this the formula on MY box? Compare the first five ingredients with the app's list and its date. 2. What does the score measure: hazard, allergens, pore-clogging, or "clean"? They are different questions. 3. Does one ingredient decide it? (Yuka and Think Dirty cap by the worst ingredient; Skincarisma uses shares.) 4. Is dose considered? Most apps rate an ingredient, not how much of it is in the product. 5. Does a second app agree? If not, read why on both.
Twenty widely sold skincare and makeup products, chosen before any database was read. Each was looked up on every database that shows per-product results on the public web: EWG Skin Deep (sunscreens through EWG's sunscreen guide), INCIdecoder, SkinSort, SkinSAFE and Skincarisma, all on 21 October 2026. Where a database had several listings, we took the base version and, on EWG, the listing with the newest "data last updated" date, and reported the others as a range. CosDNA refused automated reading that day, and Yuka, Think Dirty and OnSkin publish no product pages on the web, so they are not in the grid. Every page we read is kept, and every figure here links to its source. This compares what the databases publish; it is not a test of any product's safety.
The data: every value above, with its source URL, is free to download as a CSV under CC BY 4.0. Cite it as: SlipperSkin, "Do skincare ingredient apps agree?", 21 October 2026, slipperskin.com.
As an Amazon Associate I earn from qualifying purchases.
Mostly not. Reading 20 best-selling products on 21 October 2026, EWG Skin Deep, INCIdecoder, SkinSAFE and Skincarisma split on whether 16 were clean or flagged, and 6 got opposite verdicts.
There is no single answer key, because they answer different questions: EWG rates ingredient hazard, SkinSAFE checks for common allergens, INCIdecoder explains ingredients without scoring them, and Yuka lets the worst ingredient cap the score. Choose the one that matches your question and check its ingredient list against your box.
EWG's listing read on 21 October 2026 scores it 3, its moderate band, with allergy and immunotoxicity rated moderate and use restrictions rated high. Skincarisma rates the same product 90 "Excellent" and SkinSAFE gives it 100.
Yes. Skincarisma's methodology page publishes its formula: 100 minus the share of moderate-risk ingredients times 0.45, minus the share of high-risk ingredients, using EWG's ratings. Because it counts shares rather than the worst ingredient, its rankings of these products barely relate to EWG's (rank correlation −0.07).
Because databases keep the list they were given, and formulas change. On these 20 products, at least one database held a different formula for 8, for example parabens CeraVe has since removed, or fragrance missing from one list of Olay Micro-Sculpting Cream.
Not on its own. These databases rate ingredients, mostly without regard to how much is in the product. Dose-aware safety assessments come from bodies such as the Cosmetic Ingredient Review panel and the EU Scientific Committee on Consumer Safety. SlipperSkin reports what each database says and does not issue its own safety verdict.
Ingredient apps are useful instruments that disagree because they measure different things, apply different rules and often hold different versions of the formula. Pick the app whose question matches yours, compare its ingredient list with your box, and when two apps disagree, read why before you bin the product.