How to Segment Beauty Customers by Skin
Skin is a better segment than demographics
Most beauty segmentation uses age, gender or spend. Those are proxies. What actually drives what a customer buys in skincare is their skin: their concern, type, tone and stage. Two people the same age with very different skin need very different products. Segmenting by skin gets you closer to what the customer will actually respond to, which makes every message and recommendation sharper.
Build segments around concern and stage
Start with the dimensions that change the recommendation.
- Concern: pigmentation, dryness, congestion, sensitivity and so on.
- Stage: early routine, established routine, or managing a specific issue.
- Type: oily, dry, combination, sensitive.
- Tone: across the full range, so segments stay accurate for every customer.
A skin scan captures these consistently, so the segments are built on real data rather than guesses. See the skin scan.
Keep segments tight enough to act on
A segment is only useful if you can send it a genuinely different message. Too broad, and the message is generic. Too narrow, and you cannot maintain it. Start with your five or six most common concerns and build a distinct journey for each, then split further only when the data shows a difference worth acting on. Tight, meaningful segments feel personal without needing a message per person.
Use segments across the whole journey
Segmentation is not just for email. Use the same skin segments everywhere: matched recommendations on site, tailored bundles, targeted refill timing, and relevant follow up. When every channel reads from the same skin segments, the customer gets a consistent, relevant experience rather than contradictory messages from different systems. A skin aware CRM keeps the segments in one place. See the data view.
Let segments evolve with the customer
Skin changes, so segments should not be fixed. A customer who joined with congestion may later be managing pigmentation or dryness. Re scan customers periodically and move them between segments as their skin changes, so your targeting keeps pace. A segment based on last year's skin is as misleading as no segment at all.
Measure segments by response, not size
Judge a segment by how it responds, not how large it is. Track conversion, reorder rate and revenue for each skin segment against the rest. The segments that convert and reorder best deserve more investment and sharper journeys. If a segment does not respond differently from the whole list, merge it. Segmentation earns its keep only when it changes what customers do. Explore partner results.
Frequently asked
Segment by skin rather than demographics. Build segments around concern, stage, type and tone, since those are what actually drive what a customer buys in skincare. A skin scan captures these consistently, so the segments rest on real data rather than proxies like age or spend.
Because age and spend are only proxies for what a customer needs, while skin concern and type directly determine which products suit them. Two people the same age with different skin need different products, so skin based segments predict response far more accurately.
Tight enough that you can send each one a genuinely different message, but broad enough to maintain. Start with your five or six most common concerns, build a distinct journey for each, and split further only when the data shows a difference worth acting on.
Yes. Skin changes, so re scan customers periodically and move them between segments as their concerns evolve. A segment based on last year's skin can be as misleading as no segment at all, so keep the profiles current.
Measure conversion, reorder rate and revenue for each skin segment against the rest. Segments that respond differently and better deserve more investment, while a segment that behaves like the whole list should be merged. Segmentation earns its keep only when it changes what customers do.
See what skin intelligence does for your business.
Oyster reads skin accurately on every tone and turns it into the right recommendation.