Fashion and accessories. Shopify for ranges with a lot of variants.

A fashion article is rarely one product. It is a colour times a size run, often across several widths or lengths. One boot quickly becomes dozens of variants, one collection a great many. How you model that decides your maintenance effort, your filters, your stock logic and how fast the product page loads.

Shopify allows three option dimensions per product. Colour, size and width already use them up. Add material or finish and one attribute has to become a product of its own, linked to its siblings in the theme. That decision is best made before the migration, not afterwards in day-to-day operations.

For Gusti Leder we moved more than a thousand products across from Shopware. The work was not in the import. It was in deciding which attribute becomes a variant, which one stays a metafield, and which property still earns a filter or a place on the page at all.

In fashion and footwear the return rate is not a reporting metric; it is the lever on your contribution margin. Most of it is created before the purchase: vague size information, images without a sense of scale, missing material detail. That is product data work, not something to clean up in a returns portal.

On the product page, what helps is specific: size charts per series rather than per store, notes on whether an item runs small, images on a model with the size stated, materials in plain language. If the same size fits differently depending on the last, that belongs on the product, not on a general help page.

The processing itself should be automated. At Groundies it runs through 8returns, connected to Shopify, alongside Zendesk for support and Workato for the warehouse connection. The effect shows up less in the returns process itself than in the number of manual interventions it used to cost every week.

Fashion runs on a seasonal rhythm. Twice a year a large part of the range turns over, and in between there are drops, restocks and clearance. A setup that treats every collection change as a project slows you down twice a year and ties up development time that is needed elsewhere.

In practice that means collections are filled by rules rather than by hand, campaign pages are assembled from theme sections rather than new code, and stock comes from the system that owns it. At Gusti Leder the ERP feeds inventory, orders and returns into Shopify over a REST interface.

If you also sell in stores or through marketplaces, you need a clear answer to which system owns inventory. At MYCR7 that is a proprietary ERP; Shopify is a sales channel, not the source of truth. Settling this early spares you the later argument about which stock figure is correct.

In fashion, going international is more than currency and language. Sizing systems differ by country, tax and duty rules apply differently per market, and return addresses have to be local. Shopify Markets covers pricing, tax and domain logic; the sizing logic belongs in your product data, where it can be maintained.

At Groundies two legacy systems sat side by side: OXID for the United States, Shopware for international. We merged both onto a single shared Shopify Plus codebase and launched in two steps, the United States first, international sixty days later. The project ran six months from kickoff to international rollout.

A staged launch is a method, not a compromise. One market goes live, the team learns from real orders, then the rest follows. MYCR7 worked differently: the EU and Switzerland at the same time, with EUR and CHF and separate tax logic, because both markets were needed at launch.

Engraving, monograms, colour combinations: personalisation raises the value of an article and makes it non-returnable at the same time. Technically there is more to it than an input field. The configuration has to reach the order, the warehouse, the delivery note and the returns rule, or the work turns up manually later.

For Gusti Leder we built a dedicated Shopify app for this: an engraving configurator with a live preview directly on the product page, with no page change. Custom-built rather than a third-party app, because the preview has to match the product image and the data has to run cleanly into ERP and fulfilment.

The Gusti Leder store went live in eight weeks and the conversion rate rose by 45 per cent. We do not credit that to the configurator alone: the migration, the theme and the checkout were all renewed in the same pass.

Fashion sells through images. A product page with many views, zoom, a model video and a colour switcher quickly weighs more than the rest of the page combined. The answer is not fewer images but proper delivery: responsive sizes, modern formats, lazy loading below the fold, fixed aspect ratios against layout shift.

The second lever is the colour switcher. If every colour variant reloads half the product page, you pay for that on every click. Built properly, only the image sources are swapped and the selected variant is held in state. That is work on the theme, not a question of which app you choose.