CRM Segmentation by Beauty Profile: Lists That Talk About Skin, Not Age

Most beauty CRM databases are still segmented by age, declared gender, recency and spend. Those are useful signals for contact frequency, but terrible for deciding which product to show: two 32-year-old customers with the same average order value in the same city may need foundations in opposite undertones. CRM segmentation by beauty profile fills exactly that gap — it uses visible skin and hair attributes, captured from a selfie, to build lists that speak the language of the catalog.

What a beauty profile adds to your CRM

A beauty profile is a set of structured fields: skin tone, undertone, skin type and condition, hair attributes, each one with a confidence score. It describes what appears in the photo — it is not a medical assessment and does not replace a professional.

Inside the CRM, that becomes something simple and powerful: new columns on the customer record. Instead of "woman, 25–34, bought makeup in the last 90 days", you now have "cool undertone, combination skin, wavy hair, profile captured 4 months ago". The first line tells you when to send. The second tells you what to show.

Why this matters at send time

Beauty campaigns usually run in two modes: the one-size-fits-all newsletter and the recommendation based on purchase history. The first ignores the customer; the second only works after she has already bought. The beauty profile covers the stretch between first contact and first purchase — which is exactly where most of the database is sitting.

CRM segmentation by beauty profile: how to build the lists

Start with a few fields

Don't try to segment on everything at once. Three cuts are enough to sustain an entire campaign calendar:

• Undertone (warm, cool, neutral) — foundation, concealer, lipstick, color accessories.
• Skin type and condition — skincare, primer, finish (matte or luminous).
• Hair attributes — treatment lines, stylers, kits.

Each field becomes a dynamic segment in your CRM, updated whenever a new profile is written.

Combine with behavior, don't replace it

The profile doesn't retire the data you already have. It multiplies it. "Cool undertone + never bought foundation" is a discovery campaign. "Cool undertone + bought foundation 5 months ago" is a replenishment campaign, with the exact SKU in the subject line. "Oily skin + abandoned cart with a primer" is a reminder with an argument instead of a discount.

Respect the confidence score

Every field comes with a confidence indicator. Use it as an eligibility filter: fields with low confidence stay out of campaigns that assert something about the customer and go into softer flows, the "want to refine your profile?" kind. Stating an uncertain data point with confidence is the fastest way to lose credibility in the inbox.

How to capture the profile without stalling the database

The profile has to enter the database from somewhere. The entry points that tend to work best:

• Signup onboarding: right after the email, an invitation to send a selfie and get recommendations. The selfie-based skin and hair analysis via API returns the fields as JSON, ready to be written to the customer record.
• Product page: shoppers unsure about shade tend to accept help. The profile is born from a real question.
• Post-purchase: in the confirmation email, with the promise of personalizing the next recommendations.
• WhatsApp and chat: with MaIA Beauty Advisor, the profile is generated inside the conversation and flows back into the CRM alongside recommendations from your own catalog.

In every case, be clear with the customer about what is being analyzed, what for, and how long the data is stored. Explicit, revocable consent isn't only a legal requirement: it's what keeps opt-in rates high on the second campaign.

Flows that get better with a profile

Welcome

Swap the generic three-email sequence for a track that branches by undertone and skin type. Same editorial content, different product showcases.

Reactivation

Inactive customers with an old profile are a good audience for a refresh invitation: "it's been a while since you updated your profile — send a new selfie and see what's changed in our line". It's a reason to make contact that doesn't depend on a discount.

Collection launch

When new shades arrive, the CRM already knows who has a compatible undertone. Pair the send with a try-on link: photo-based virtual makeup try-on shows the shade on the customer herself, and the browser SDK, in early access, enables live camera mode on your site, with processing happening on the device.

What to measure

Compare segment by segment, not the overall average: open and click rates for profile-driven campaigns against a control group without a profile; conversion by undertone segment; the profile's share of repeat purchases; and the capture rate (how many customers in the base agree to send a selfie). Results vary with each brand's catalog, traffic and channel — which is why the control group is mandatory from the very first send.

Operations: the profile has to age well

Store the capture date and model version alongside every profile. Define a validity window — skin and hair change with the season, routine and treatment — and create an automatic recapture trigger when that window expires. An outdated profile in a full-base send gets it wrong at scale.

Want to build your first beauty-profile segments? Create your account, buy credits and generate your key in the API documentation, or talk to the team to design the integration with your CRM.