Igor Swatowski
ENPL
Get a free quote
Selectedprojects

VeraBotanica. An online store design for a skincare brand

A woman spending €180 on a serum isn't shopping by price. She's shopping by her problem.

Project

VeraBotanica

Industry

Beauty · E-commerce

Year

2025

My role

UX/UI Designer

Tools

Figma · Midjourney · ChatGPT

Services

UX/UI design, Visual identity, AI imagery

Date

2025

The premise

A woman spending €180 on a serum isn't shopping by price. She's shopping by her problem.

[01]

€180 for a serum from a brand nobody has heard of.

The whole project revolves around one question: what makes someone comfortable spending €180 on a product from a brand they don't know?

I designed a full store for a natural skincare brand in the €50–180 range. I was testing how skin-type-led navigation and trust signals placed at the decision points could justify that price.

There was no photography and no budget for it. I generated every product image in Midjourney, which freed up the time for what actually drives a premium purchase.

Every page in the path has one job. The homepage builds credibility. The category page helps her find her product. The product page justifies the price. The cart removes reasons to hesitate.

The VeraBotanica homepage, desktop and mobile. Header, collections, trust bar, botanical innovation section and FAQ.

[02]

Three decisions that hold up the price

Trust doesn't come from a badge stuck on at the end. It's built at the three moments where she decides whether to keep going.

  1. [01]

    Skin-type filters above the product grid

    Categories use standard navigation by product type: creams, serums, cleansers. But once she's inside a category, her next question is “which one is for my skin”, not “sort by price”.

    I put skin-type filters horizontally above the product grid, so narrowing the results is one tap without leaving the page. A small change that makes a catalogue start working like a recommendation.

    The Creams category page. Skin-type filters above the product grid, desktop and mobile.

  2. [02]

    Trust signals where the hesitation happens

    At €50 to €180, she needs reassurance before clicking “Add to cart”. I put four trust signals directly under the button, visible without scrolling, exactly where the doubt appears.

    Deeper product information sits in expandable sections. The page stays clean for those who buy fast, and the more cautious can go deeper without leaving it. “Botanical origin” and “Application ritual” are patterns borrowed from luxury brands, so it's clear this isn't another drugstore cream.

    The product page. Four trust signals under the button, with Description, Details, Botanical origin and Application ritual below.

  3. [03]

    A two-stage cart: a drawer for momentum, a full page for the decision

    Instead of one cart I designed two. The drawer appears the moment something is added and exists to keep momentum: it doesn't interrupt the shopping and it gives a reason to add one more thing.

    The full cart page does something else. It's where a decided buyer checks everything: itemised costs, delivery and VAT split out, a field for a discount code. Two different states of mind, two different interfaces.

    Both carts in one frame. The drawer with a free delivery progress bar next to the full page with an itemised total and a discount field.
Newsletter modal with a first-purchase offer, desktop and mobile.
Order confirmation: order number, delivery details and an itemised summary.
Post-purchase account creation: the prompt from the confirmation page alongside the sign-in and sign-up modals.

[03]

How the imagery was made

I generated every product image in Midjourney. It was its own process, not a shortcut:

  • Three to four reference images to start, to lock the aesthetic: quality of light, colour temperature, and a mood somewhere between botanical and clinical.
  • Prompts with equipment specs: lens type, focal length, lighting setup. A general prompt gives you a general result. A technical prompt gives you something you can control.
  • One good output as the style reference for every generation after it. That's what creates consistency across a product line. Without it, every shot looks like a different photographer took it.
  • Label typography handled separately, in ChatGPT's image generation. Midjourney can't do text, and knowing which tool handles which job is part of the craft.

Next step

Has your business outgrown its website?

Projects

Seeall

More projects

Seeall