DNA-personalized skincare

01Overview and value chain#

Markers EC: EU Cosmetics Regulation 1223/2009 | OECD: Genomics & Bioinformatics | Regulator: FDA (USA), EMA (EU), FTC (USA)

The premise is narrow and testable: a handful of common variants influence how fast skin loses collagen, how strongly it responds to UV and oxidative stress, and how well it holds a barrier — so a product chosen against those variants should beat one chosen off a shelf. The commercial expression is a cheek swab returned by post, a report on a defined marker set, and either a bespoke formulation or a mapped regimen. Panel sizes are modest and disclosed: the SkinDNA laboratory test reads 16 genetic markers across five skin-ageing categories, Mapmygenome’s BeautyMap covers more than 40 skin and hair conditions, and LifeNome’s engine claims to analyse over 20,000 gene-ingredient interactions. Turnaround is measured in weeks, not days — six to eight weeks on some Australian panels — and pricing runs from roughly AU$129 for a report-only test to a full DNA analysis bundled with four personalised products.

The key directions of DNA-personalized skincare are:

1. Dermatogenomic SNP panels (Genetic skin typing): a defined set of variants associated with collagen loss, elasticity, pigmentation, sensitivity and oxidative-stress response, read from a mouth swab. SkinDNA’s laboratory test examines 16 SNPs across five categories including firmness and elasticity, positioned to flag accelerated collagen loss before it is visible.
2. Bespoke formulation from a genetic report (Made-to-genotype products): the report drives an actual product rather than advice. Nomige sells a DNA analysis with a skin report and four personalised products, and runs a reorder path for those products — a subscription shape rather than a one-off test.
3. White-label personalisation engines (B2B beauty intelligence): the analytical layer licensed to other brands rather than sold to the consumer. LifeNome’s precision-beauty engine covers DNA, microbiome and epigenetic methylation data and is explicitly offered white-labelled, which is how most large beauty brands enter this space.
4. Epigenetic ageing readouts (Methylation-based testing): a second measurement axis that changes over time, unlike genotype. Muhdo built its epigenetic supplement range on data from 70,000 individuals and publishes gene-level skin-health content, extending the model from fixed variants to modifiable methylation state.

Sectoral value chain#

[Swab kit] ──> [Sample collection] ──> [Genotyping/methylation] ──> [Report]
                                                 │
                                       (gene-ingredient rules)
                                                 │
                                                 ▼
[Reorder / retest] <─── [Formulation or regimen] <┘

Value chain levels#

Cross-cutting technologies of the sector:

• Buccal-swab sample logistics (At-home DNA collection): mouth-swab kits and postal return, the same consumer-genomics infrastructure that made ancestry testing a mass product, reused for a cosmetic endpoint.
• Gene-ingredient rule engines (Ingredient matching): the mapping layer between a variant call and an active — the actual intellectual property in this category, and where LifeNome’s claimed 20,000-plus interaction set sits.
• Methylation clocks for skin (Epigenetic readout): assays that measure a modifiable biological state rather than a fixed genotype, allowing a retest to show change and turning a one-off purchase into a monitored one.

02US#

The US pattern is business-to-business: the technology is sold to beauty brands as an engine rather than to consumers as a brand, which is what lets a small team reach a mass market.

White-label precision-beauty engines, multi-omic personalisation, small teams behind large brands#

• LifeNome: founded in 2016 and headquartered in New York with a team in the single digits and $5.1 million raised across two rounds, it sells a precision-beauty engine covering DNA, microbiome and epigenetic methylation, analysing gene-ingredient interactions and offering white-labelled reports that partner brands ship under their own name.
• Multi-omic framing: the US vendors have largely moved past genotype alone, folding microbiome and methylation data into the same recommendation, which broadens the claim surface and also the evidentiary burden.
• Claim discipline: because the endpoint is cosmetic rather than medical, marketing sits under FTC substantiation rules rather than a regulatory approval process, which is the constraint that shapes how these products describe themselves.

03CN#

China is a large personalised-cosmetics market, but no domestic DNA-skincare vendor cleared confirmation on this screen, so the region is described rather than tabled.

Personalisation without genotyping, regulatory caution on genetic data, unconfirmed vendor set#

• Different personalisation route: the mainstream Chinese approach to bespoke skincare leans on instrument-based skin analysis and AI imaging in retail settings rather than on a mailed DNA sample, so the buyer journey does not pass through a genotyping laboratory.
• Genetic-data governance: human genetic resources are tightly regulated in China, which raises the compliance cost of a consumer service built on collecting and processing DNA for a cosmetic purpose.
• Screening limit: searches for 皮肤基因检测 and 定制护肤 returned no vendor that could be confirmed as a merchant of a DNA-based skincare product, so no Chinese row is claimed here.

04EU#

Europe is where the direct-to-consumer product form is strongest: a founder-led brand selling a genetic analysis bundled with the creams it prescribes.

Founder-led DNA cosmetics brands, bundled analysis and product, epigenetics as the adjacent market#

• Nomige: a Belgian brand founded by Dr Barbara Geusens on dermatogenomics research, selling a DNA analysis with a skin report and four personalised products, plus a day-cream range addressed to specific concerns — stress protection, barrier repair, redness, pigmentation — and a reorder path for the personalised formulations.
• Muhdo: a UK epigenetics company whose supplement range is informed by data from 70,000 individuals and which publishes gene-level skin-health material, representing the methylation-based rather than genotype-based end of the same market.
• Cosmetics framing: under Regulation 1223/2009 these remain cosmetic products, so the genetic report is a selection aid and not a therapeutic claim — a boundary the credible vendors are careful to keep.

05Leading companies and research institutes#

06Tech stack and innovations#

Three layers stack up, and only the middle one is genuinely hard: collection is commodity, laboratory work is contracted, and the interpretation rules are the product.

1. Panel selection and disclosure (SNP panel design):
   • Panels are deliberately small and named — 16 markers in five categories on the SkinDNA test, 40-plus conditions on Mapmygenome’s BeautyMap — because a defined scope is easier to substantiate than an open-ended genomic claim.
   • Categories map to concerns a buyer already recognises: firmness and elasticity, pigmentation, sensitivity, oxidative stress, so the report lands as advice rather than as data.
2. Gene-ingredient rule engines (Interpretation layer):
   • This is where the differentiation lives. LifeNome’s engine is built around gene-ingredient interactions at a claimed scale of more than 20,000, which is the asset a licensing beauty brand is actually buying.
   • The same layer determines whether the output is a bespoke formulation, as at Nomige, or a mapped selection from an existing catalogue — a formulation decision as much as an analytical one.
3. Methylation as a second axis (Epigenetic testing):
   • Genotype does not change, so a genotype-based product can only be sold once; methylation does change, which supports retesting and a monitoring relationship. Muhdo built its range on that distinction with a 70,000-individual dataset behind it.
   • It also shifts the claim from predisposition to current state, which is a stronger consumer proposition and a correspondingly higher evidentiary bar.

07Value chains and production pipelines#

Industrial pipeline for genotype-guided cosmetic formulation (EU Cosmetics Regulation 1223/2009)#

┌───────────────────────────┐      ┌───────────────────────────┐
│ 1. Panel definition       │ ───> │ 2. Swab kit & collection  │
└───────────────────────────┘      └───────────────────────────┘
                                                 │
                                                 ▼
┌───────────────────────────┐      ┌───────────────────────────┐
│ 4. Report generation      │ <─── │ 3. Laboratory genotyping  │
└───────────────────────────┘      └───────────────────────────┘
              │
              ▼
┌───────────────────────────┐      ┌───────────────────────────┐
│ 5. Formulation or regimen │ ───> │ 6. Reorder & retest       │
└───────────────────────────┘      └───────────────────────────┘

Stage 1: Panel definition

A marker set is fixed from published skin-genetics associations and grouped into consumer-legible categories — the SkinDNA panel’s five categories over 16 SNPs is typical of the scale. Fixing the panel also fixes the claim, which is why vendors publish the count.

Stage 2: Swab kit & collection

A buccal swab kit ships with instructions and a prepaid return. The step is commodity consumer-genomics logistics, and its failure mode is a low-yield or contaminated sample rather than anything analytical.

Stage 3: Laboratory genotyping

The sample is typed for the panel or assayed for methylation, usually in a contracted laboratory. Turnaround is the visible cost: six to eight weeks is quoted on some Australian skin panels, an eternity by the standards of consumer beauty.

Stage 4: Report generation

Variant calls become per-category scores against the vendor’s rule set, often combined with a lifestyle questionnaire. This is where a genotype becomes a recommendation, and where two vendors reading the same sample can legitimately differ.

Stage 5: Formulation or regimen

The report either drives a bespoke formulation — Nomige’s model, a DNA analysis bundled with four personalised products — or maps to items in an existing range, which is the lower-cost route most licensing brands take.

Stage 6: Reorder & retest

Genotype-based products monetise through replenishment of the personalised formulation; epigenetic products add periodic retesting, since methylation state moves and a second measurement can show it.