EPL Diamond: UX of the purchase path

EPL Diamond is a jewelry brand with an online store and retail. A diamond purchase is expensive and comparative: shoppers weigh price, stone, metal, delivery and payment terms.

The client had designers and a design system but no verified structure for the mobile path: which steps, in what order, with what information. I was brought in to close that layer.

At the start (expert audit and CJM, not client analytics): many near-identical offers with unclear price differences; no reviews; an overloaded product card; the promo code sits on the product page, forcing a return after the cart; checkout asks for many fields and registration does not state its benefit.

SOMEONE ELSE'S TEAM. I joined an established team. The output had to be buildable by the client's designers without follow-up questions.
DESIGN SYSTEM. Visual design belonged to the client, so the wireframes specify structure, states and behavior, not looks.
DATA. No analytics or interviews: a competitor review, CJM and expert judgement. Some decisions are hypotheses, and they are labeled as such.
FORMAT. Mobile web: every extra step and field is felt immediately.

I used AI during competitor analysis, to gather information and formalize my conclusions. The flows, CJM and wireframes are my own work.

1. Analysis. Six sites (Sokolov, Sunlight, MIUZ, Adamas, Tiffany, Blue Nile): product, delivery and payment information, visible discounts, registration hints, ways around banner blindness.
2. CJM. "Buy a diamond ring in Moscow": search → catalog → product card → delivery and payment terms → cart → payment. Per step: expectation, friction, emotion, possible improvement.
3. User flow and wireframes. Home, search, chat, menu, diamond and jewelry catalog, selection guides, filters, product card, quick order, cart and checkout, account, favorites, personal discounts, help, error pages.

Decision 1. Help people choose, don't just list

• Problem: look-alike offers, unclear price differences, an overloaded card.
• In the wireframes: a five-step selection guide (single and multi-select) leading into a catalog with matching filters; separate filter templates; product, delivery and payment information separated, details in accordions.
• Trade-off: the guide adds steps, so it is optional and the catalog stays a path of its own.

Decision 2. A shorter path to payment

• Problem: many fields, registration with no stated benefit, returning for the promo code, irritation as steps pile up (CJM).
• In the wireframes: step-by-step checkout; skippable registration; optional phone login; quick order from the card; pickup chosen in a modal; SMS confirmation.
• Edge states: payment and filter errors, 404/500/503, empty favorites, signed-out user, unconfirmed purchase on review.
• Trade-off: simpler checkout versus collecting more client data; a completed order came first.

What the client turned down: product shelves on the diamonds page
For the diamonds page, I proposed making whole-stone purchases easier by adding product shelves. The client decided the page stays purely editorial, without product shelves, and found my idea unnecessary. It was their call, and the page was designed in that editorial role. I could not test the hypothesis that this page could sell whole diamonds, and I do not present it as proven.