Virtual Makeup Try-On for a Collection Launch: New Shades Without a Tester

Every collection launch has the same blind spot: the customer has to choose a shade that does not yet exist on any counter, based on a campaign photo shot on different skin, under different light, with a different camera. Virtual makeup try-on for a collection launch exists to close that gap — it puts the new shade on the face of the person making the decision, before the product reaches any shelf. For e-commerce and CRM teams, that changes three distinct moments: the teaser phase, launch day, and the week after, when it is time to decide what to restock.

Why a launch is the hardest scenario in the catalog

In the running catalog, the customer has references: she has already bought one of your lipsticks, she knows the finish is matte, she knows "nude 03" works on her. In a launch, none of that exists. The naming is new, the color family is new and, very often, the whole collection is built around a finish the brand has never sold before.

The result is predictable: carts built on campaign impulse, shade choices made on the strength of a poetic name, and a wave of exchanges that arrives two weeks after the media peak — exactly when the right shade is out of stock and the wrong one is sitting still.

What a preview solves and what it does not

A color preview answers the question "how does this look on me". It does not replace a description of finish, texture, wear time or coverage, and it is not a medical assessment — the render shows visible attributes of a photo or of the camera feed. Let the page copy handle the rest; the try-on handles color.

Virtual makeup try-on for a collection launch in three phases

Teaser phase: a waitlist with shade intent

A traditional waitlist captures an email. A waitlist with a preview captures an email and the shade the person tried. Instead of sending "the collection has arrived" to your whole base, you send "the shade you tried has arrived" — and your CRM gains a field worth more than the signup date.

In practice: a teaser landing page with MaIA Try-On, the collection palette already calibrated in the project, and a short form after the try-on. What you carry into launch day is a segmentation by color, not a single list.

Launch day: a page that holds up under campaign traffic

On launch day, traffic arrives concentrated and impatient, much of it from paid media and link-in-bio. Two decisions matter here.

The first is where to place the preview: above the fold on the launch product page and inside the shade selector, not hidden behind a tab. The second is the try-on mode. The browser SDK runs live camera mode, with the render done locally and no image upload — useful on a phone, at home, with the customer deciding on the spot. The photo-based API fits better in flows where the image already exists or needs to come back as a file, for example a CRM send that returns the rendered preview. If you are still choosing between the two, it is worth reading the comparison between photo-based try-on and live mode before settling the architecture.

After launch: reading demand through the shades that were tried

The following week is about restocking. Try-on sessions show which shades were tried a lot and bought little, which were tried by similar skin profiles, and which almost nobody opened. This does not replace sales data, but it arrives before it and helps decide what to order from the factory and what to feature in the second media wave.

How to prepare the palette before the date

Calibration is project work, not a last-minute task. A few points to put on the schedule:

• Freeze the collection's SKU list well in advance: code, commercial name and color reference for every shade.
• Map the render to the real SKU, not to a hex picked by eye. The preview is only useful if the shade on screen matches the shade coming off the production line.
• Decide the product scope. Through the API, photo-based rendering covers lipstick and foundation. In the browser SDK — in early access since September 2026, granted per account — there are thirteen products, from lipstick and foundation to eyeliner, highlighter and lip liner, plus the combined Full Look, always as a local preview in the browser.
• Test on varied skin. MaIA Scan was trained on Brazilian skin; use that in your favor when validating the palette with real images of your audience, not only with campaign models.
• Combine it with a beauty profile when it makes sense: selfie-based skin and hair analysis returns tone and undertone as JSON, which lets you pre-select two or three shades from the launch before the customer even opens the preview.

What to measure, without inventing numbers

Set the baseline before the launch, using a comparable previous launch. Three readings are enough for a first assessment: preview open rate on the launch page, conversion among those who tried versus those who did not, and exchange distribution by shade over the following four weeks. Results vary with catalog, traffic and channel — so compare your brand against itself, not against third-party benchmarks.

Checklist for the final weeks

• Collection palette calibrated and reviewed with the product team.
• Teaser page live with the preview and shade capture active.
• CRM segmentation by tried shade ready for the launch-day send.
• Photo capture guidance (front light, no filter) inside the photo-based flow.
• Data reading plan agreed with merchandising before restocking.

If the next collection already has a date, you can start with the palette: create your account, generate the key and test the integration in the API documentation, or talk to the B4A team to set up the calibration project for the launch shades.