MaIA Scan

omnichannel beauty profile

Omnichannel Beauty Profile: The Same Profile on the Site, on WhatsApp and In Store

How to make an omnichannel beauty profile work: a single skin and hair record that your site, WhatsApp and store counter can read and update, without asking for the same selfie three times.

5 min readMaIA team · B4A

A shopper takes a selfie on your site, gets a foundation shade suggestion and buys. Two weeks later she opens the brand's WhatsApp and the conversation starts from zero: "what's your skin type?". The following month, in the physical store, the consultant asks all over again. Same customer, same skin, three service moments that never talk to each other. An omnichannel beauty profile solves exactly that: a single record of visible attributes — skin tone and undertone, skin type and condition, hair characteristics — that every channel reads, uses and updates.

This article is about the least glamorous and most decisive part of the project: where the profile lives, who is allowed to write to it, and how each channel consumes that data without spawning parallel versions of the same person.

The problem isn't capturing the profile, it's keeping it in one place

Capturing is the easy part. With skin and hair analysis from a selfie via API, the customer sends a photo and you get back a JSON with the attributes and a confidence score per field. The hard part comes next: the site stores the result in a cookie, the app stores it on the device, the chat stores it in the session and the store stores nothing at all. Within a few weeks the brand has four incomplete profiles of the same person and not one of them is trustworthy.

The practical rule is simple: the profile belongs to the customer, not to the channel. It lives in the CRM or the customer data platform, tied to the identifier you already use — email, phone number, customer ID. Channels are readers and, in controlled cases, writers. No channel owns it.

Identity before analysis

Before calling the API, decide which key will bind the profile to the person. On a logged-in site it's trivial. On an anonymous visit, you store the result locally and merge it when the customer identifies herself — at login, at checkout or when she asks for the result by email. On WhatsApp, the phone number is already the key. In store, it's the tax ID or the customer record the associate opens. Without that decision made before the first request, the omnichannel profile becomes a pile of orphans.

How to build an omnichannel beauty profile in practice

1. Define the profile contract

Store the fields returned by the analysis, the confidence score of each one, the capture date and the channel of origin. Store the customer's declared preferences too — "I like a matte finish", "I avoid fragrance" — in a separate block. An attribute observed in a photo and a declared preference are different things and they age at different speeds.

2. Decide who can write

A good policy: photo analysis writes the visible attributes; human service and conversation write the preferences; no channel overwrites a high-confidence field with a guess. If the store associate disagrees with the recorded undertone, she creates a new capture — she doesn't edit the field by hand.

3. Treat updates as events, not as a form

Skin changes with the season, the routine and age. Instead of an annual form, trigger a new analysis at natural moments: a change of season, a foundation repurchase, a return after months without buying. Each new capture enters as a version, with a date. You keep the history and you get to say "your last reading was in March" instead of treating the profile as eternal truth.

4. Give each channel the right slice

The site doesn't need the full history: it needs to know which shade to feature. WhatsApp needs enough not to repeat questions. The store needs a summary the associate can read in five seconds. Same profile, different slices — and that is solved in the brand's API layer, not inside each channel.

What each channel gains from a shared profile

E-commerce. Category sorting, SKU highlighting and kit recommendations stop depending on browsing behaviour alone. A customer with a recorded undertone doesn't see a generic storefront.

WhatsApp and chat. The AI beauty advisor integrated with the catalog reads the profile and starts the conversation at the right point, recommending the brand's own products instead of opening with a questionnaire. The most visible gain is the question that no longer has to be asked.

Physical store and direct sales. The associate opens the record and sees the skin and hair summary, the products already bought and the date of the last reading. It's the same kind of context a good consultant builds over years of relationship, available at the very first interaction.

CRM. Campaigns start talking about skin and hair instead of age brackets. And the profile also informs the decision about which campaign a customer shouldn't receive at all.

Makeup try-on and profile: two signals that add up

The profile says what the photo shows. Try-on shows the product applied. When virtual makeup try-on runs with palettes calibrated to the brand's real SKUs, the shades tried become an additional signal for the same record: what the customer wanted to see, not only what the analysis described. Store that signal alongside, with its own label.

Governance: fewer fields, more clarity

Three decisions prevent most of the problems:

  • Explicit, purpose-specific consent. The customer authorises the analysis and knows the result stays in her record, with the option to delete it.
  • Image retention separate from the profile. In most cases the brand doesn't need to keep the photo: it needs the attributes. Decide this before go-live.
  • No clinical interpretation. The profile describes visible attributes of a photo. It is not a medical assessment and does not replace a health professional. Write that in the interface, on every channel.

An omnichannel beauty profile is not a giant data project. It's a well-designed table, an identity key decided early, and the discipline not to let each channel invent its own version.

Want to test the first step? Create your account, buy credits and generate your key in the API documentation — or talk to the B4A team to design the shared profile across your channels.

Frequently asked questions

What is an omnichannel beauty profile?

It's a single record of a customer's visible skin and hair attributes — tone, undertone, type, condition, hair characteristics — stored in the brand's CRM and read by every channel: site, app, WhatsApp, chat and physical store. No channel owns the data; they all consume the same source.

Where should the profile be stored?

In the brand's CRM or customer data platform, tied to the customer identifier you already use. MaIA Scan returns the attributes as JSON via API; the decision on where to persist them and for how long belongs to the brand.

How do you avoid duplicate profiles for the same person?

Define the identity key before the first request. In an anonymous session, store the result locally and merge it when the customer identifies herself at login, at checkout or when she asks for the result by email.

How often should the profile be updated?

Treat the update as an event, not as a periodic form: a change of season, a foundation repurchase, a return after months without buying. Each new analysis enters as a dated version, preserving the history.

Does the beauty profile replace a professional assessment?

No. It describes visible attributes in a photo to guide product choice and customer service. It is not a medical assessment and does not replace consulting a health professional.

beauty profileomnichannelCRMbeauty APIe-commerce

← All articles