DTC Skincare for Sensitive & Reactive Skin

Summary

What this report found.

Portfolio

What each brand runs.

Hook, format, offer, awareness stage, angle and persona, counted over distinct (brand, concept) pairs, so a brand's 40 near-identical variants of one idea count once, not 40 times.

Traction

What these advertisers keep paying for.

Ranked within each brand, at most three ads per brand, so a large budget cannot fill the list. Days running is a survival proxy: it says someone keeps funding the ad, not that the ad works.

Landing pages

Where the click stops matching the ad.

Ranked by landing-match score, worst first, at most three ads per brand. Each one is a fix a brand can ship this week without producing a single new creative.

Compliance

Claims worth a second look.

The highest-probability risk question per ad, from the seven-question compliance watch. A disclaimer present elsewhere in the ad does not clear a flagged claim: see the methodology section.

Market norms

Nothing in this niche clears the 90% bar yet.

No single trait is shared by 90% or more of this niche's distinct concepts: the niche is more varied than that.

Mix Engine

persona match × angle × concept archetype × offer type, counted honestly.

Over 6 brands and 35 distinct concepts (70 ads). A cell needs at least 2 brands and 2 concepts to get a number at all; below 3 brands and 5 concepts it carries a low-evidence badge and a wide interval.

Recommended test cells

Under-tested-but-plausible combinations and well-supported combinations that already hold up, never blended into one number.

• #1lifelong reactive + safety or risklow evidence

  6 brands run it: 7 ideas across 18 ads. The ideas have been live 15 to 373 days so far; 7 of 7 are still running. No shared format yet: a feature callout (2 brands), a before-and-after (1 brand), a stat callout (1 brand), a lifestyle scene (1 brand), a product-hero shot (1 brand) and a customer testimonial (1 brand). Low evidence: 6 brands, 7 ideas, 18 ads. How long its ideas stay live was compared with the market alongside 14 other combinations; after correcting for that many comparisons the difference could be chance (score 0.69, a real one needs under 0.10), so it describes what competitors run, not how well it does.

• #2label reader lina + quality or craftlow evidence

  4 brands run it: 7 ideas across 11 ads. The ideas have been live 19 to 161 days so far; 3 of 7 are still running. No shared format yet: a feature callout (2 brands), a talking head (1 brand), a stat callout (1 brand), a native screenshot (notes app, DM or review card) (1 brand) and a product-hero shot (1 brand). Low evidence: 4 brands, 7 ideas, 11 ads. How long its ideas stay live was compared with the market alongside 14 other combinations; after correcting for that many comparisons the difference could be chance (score 0.79, a real one needs under 0.10), so it describes what competitors run, not how well it does.

• #3lifelong reactivelow evidence

  6 brands run it: 9 ideas across 20 ads. The ideas have been live 15 to 373 days so far; 8 of 9 are still running. No shared angle yet: a safety angle (6 brands), an identity angle (1 brand) and a results angle (1 brand). Low evidence: 6 brands, 9 ideas, 20 ads. How long its ideas stay live was compared with the market alongside 14 other combinations; after correcting for that many comparisons the difference could be chance (score 0.79, a real one needs under 0.10), so it describes what competitors run, not how well it does.

• #4midlife maralow evidence

  5 brands run it: 7 ideas across 13 ads. The ideas have been live 8 to 256 days so far; 5 of 7 are still running. No shared angle yet: a results angle (2 brands), a look-better angle (1 brand), an identity angle (1 brand), a gift angle (1 brand) and a health angle (1 brand). Low evidence: 5 brands, 7 ideas, 13 ads. How long its ideas stay live was compared with the market alongside 14 other combinations; after correcting for that many comparisons the difference could be chance (score 0.96, a real one needs under 0.10), so it describes what competitors run, not how well it does.

• #5midlife mara + safety or riskno evidencewhite space 1.01

  None of the 35 ideas here combines a safety angle and for the midlife mara persona. Each part is run here on its own: a safety angle in 10 ideas from 6 brands and for the midlife mara persona in 7 ideas from 5 brands. To complete the brief, the most common choice here that fits it: a feature callout (9 of 35 ideas) and no offer (19 of 35 ideas). A test to run, not a finding: given how often each part is used here, about 2.0 of the 35 ideas labelled on both parts would land here by chance, so an empty combination says nothing yet about how it would do. Its 3 nearest combinations (one part different) are run by 6 brands and 12 ideas; none of them stays live longer than the market once the 15 comparisons are corrected for.

• #6lifelong reactive + quality or craftno evidencewhite space 0.79

  None of the 35 ideas here combines a quality angle and for the lifelong reactive persona. Each part is run here on its own: a quality angle in 7 ideas from 4 brands and for the lifelong reactive persona in 9 ideas from 6 brands. To complete the brief, the most common choice here that fits it: a feature callout (9 of 35 ideas) and no offer (19 of 35 ideas). A test to run, not a finding: given how often each part is used here, about 1.8 of the 35 ideas labelled on both parts would land here by chance, so an empty combination says nothing yet about how it would do. Its 2 nearest combinations (one part different) are run by 6 brands and 14 ideas; none of them stays live longer than the market once the 15 comparisons are corrected for.

• #7minimalist maya + safety or riskno evidencewhite space 0.76

  None of the 35 ideas here combines a safety angle and for the minimalist maya persona. Each part is run here on its own: a safety angle in 10 ideas from 6 brands and for the minimalist maya persona in 6 ideas from 4 brands. To complete the brief, the most common choice here that fits it: a feature callout (9 of 35 ideas) and no offer (19 of 35 ideas). A test to run, not a finding: given how often each part is used here, about 1.7 of the 35 ideas labelled on both parts would land here by chance, so an empty combination says nothing yet about how it would do. Its 4 nearest combinations (one part different) are run by 6 brands and 15 ideas; none of them stays live longer than the market once the 15 comparisons are corrected for.

• #8lifelong reactive + save time or effortno evidencewhite space 0.74

  None of the 35 ideas here combines a save-time angle and for the lifelong reactive persona. Each part is run here on its own: a save-time angle in 5 ideas from 3 brands and for the lifelong reactive persona in 9 ideas from 6 brands. To complete the brief, the most common choice here that fits it: a feature callout (9 of 35 ideas) and no offer (19 of 35 ideas). A test to run, not a finding: given how often each part is used here, about 1.3 of the 35 ideas labelled on both parts would land here by chance, so an empty combination says nothing yet about how it would do. Its 2 nearest combinations (one part different) are run by 6 brands and 11 ideas; none of them stays live longer than the market once the 15 comparisons are corrected for.

• #9midlife mara + quality or craftno evidencewhite space 0.61

  None of the 35 ideas here combines a quality angle and for the midlife mara persona. Each part is run here on its own: a quality angle in 7 ideas from 4 brands and for the midlife mara persona in 7 ideas from 5 brands. To complete the brief, the most common choice here that fits it: a feature callout (9 of 35 ideas) and no offer (19 of 35 ideas). A test to run, not a finding: given how often each part is used here, about 1.4 of the 35 ideas labelled on both parts would land here by chance, so an empty combination says nothing yet about how it would do. Its 2 nearest combinations (one part different) are run by 5 brands and 10 ideas; none of them stays live longer than the market once the 15 comparisons are corrected for.

• #10midlife mara + save time or effortno evidencewhite space 0.58

  None of the 35 ideas here combines a save-time angle and for the midlife mara persona. Each part is run here on its own: a save-time angle in 5 ideas from 3 brands and for the midlife mara persona in 7 ideas from 5 brands. To complete the brief, the most common choice here that fits it: a feature callout (9 of 35 ideas) and no offer (19 of 35 ideas). A test to run, not a finding: given how often each part is used here, about 1.0 of the 35 ideas labelled on both parts would land here by chance, so an empty combination says nothing yet about how it would do. Its 2 nearest combinations (one part different) are run by 4 brands and 7 ideas; none of them stays live longer than the market once the 15 comparisons are corrected for.

• #11label reader lina + quality or craft + feature or reasons calloutno evidence

  2 brands run it: 2 ideas across 3 ads. The ideas have been live 19 to 140 days so far; 1 of 2 is still running. No shared offer yet: a free-shipping offer (1 brand) and no offer (1 brand). Thin evidence: 2 brands, 2 ideas, 3 ads (even a low-evidence badge needs 3 brands and 5 ideas). How long its ideas stay live was compared with the market alongside 14 other combinations; after correcting for that many comparisons the difference could be chance (score 0.79, a real one needs under 0.10), so it describes what competitors run, not how well it does.

• #12lifelong reactive + safety or risk + feature or reasons calloutno evidence

  2 brands run it: 2 ideas across 4 ads. The ideas have been live 15 to 170 days so far; 2 of 2 are still running. No shared offer yet: no offer (1 brand) and a subscribe-and-save offer (1 brand). Thin evidence: 2 brands, 2 ideas, 4 ads (even a low-evidence badge needs 3 brands and 5 ideas). How long its ideas stay live was compared with the market alongside 14 other combinations; after correcting for that many comparisons the difference could be chance (score 0.79, a real one needs under 0.10), so it describes what competitors run, not how well it does.

Audience

What they set, who it reached, who the creative speaks to.

Set and reached are observed facts from the official Ad Library API, EU/UK-delivered ads only. Implied is Jev's read of the creative alone, shown for every brand.

Implied only (no EU/UK delivery observed)

• FernbrightLabel-Reader Lina 25% · Minimalist Maya 25% · Lifelong Reactive-Skin Carla 17% (n=12)
• RenewlyLabel-Reader Lina 25% · Minimalist Maya 25% · Lifelong Reactive-Skin Carla 17% (n=12)
• Alder & WrenLabel-Reader Lina 27% · Lifelong Reactive-Skin Carla 27% · Midlife Mara 27% (n=11)
• PalewickLifelong Reactive-Skin Carla 22% · Midlife Mara 22% · Minimalist Maya 22% (n=9)