What 25 A/B Tests Taught Us About Selling Jewellery on Shopify
D. Louise makes distinctive jewellery, and its original Shopify theme made it hard to show that. Products looked generic, and the path from browsing to buying hadn't been designed around how people actually shop for jewellery.
We built a custom theme to give the brand room to express itself. Then, instead of guessing which changes would help, we tested them: 25 A/B tests across collection pages, product pages and the cart.
+24%
Reported total sales across the project comparison.
+17%
Reported average order value.
+33%
Reported revenue from organic search.
These are the results reported across the whole project, alongside the wider optimisation work. They aren't a measurement of any single test or of the theme on its own. The reported add-to-cart rate also rose by 9%, and returning customers converted 24% more often, both relative to the comparison baseline.
How Every Test Ran
Each of the 25 tests followed the same loop:
- Start with a shopper problem. Each test began with a question about how shoppers browse, choose or buy.
- Split the traffic. Half of visitors see the change, half see the current version.
- Measure what matters. Orders, order value and add-to-cart rate, not just clicks.
- Ship it or drop it. Only changes that proved themselves stayed on the site.
The work focused on the moments that decide a purchase: how shoppers browse a collection, choose a piece and move from cart to checkout.
Why a custom theme came first
Testing only works when you can actually change things. The original theme limited how products could be presented and how far the brand could be expressed, so many good ideas were impossible to try.
A tailored theme removed those limits. It gave the team flexible product listings and cart interactions to experiment with, while keeping the jewellery front and centre.
Testing without breaking the store
Every test is a release, and a broken variant costs real sales. Each change went through the same checks as a launch, from product pages to checkout, as described in our Shopify launch QA checklist.
The takeaway
- Fix the foundations first. A theme that limits you also limits what you can test.
- Test where decisions happen. Collections, product pages and the cart matter most.
- Measure revenue, not clicks. Order value and conversion tell you if a change really helped.
- Keep a steady rhythm. Twenty-five small, measured improvements add up.
Read the full D. Louise case study.
