Beauty brands should model catalogue complexity, customer choice, subscription management, and market structure before Shopify design begins. A strong beauty storefront turns product information into structured decisions, reducing shade mismatch, support demand, migration risk, and expensive post-launch rework.
Most costly beauty-commerce problems are not theme problems. They are catalogue, customer-data, and operational decisions that were deferred until the design looked finished.
A beauty store is often scoped like any other Shopify project, and that is where most of the later problems come from. The catalogue gets modelled on the SKUs that exist at the time of the brief. The product template is designed around brand photography. Subscription is added near the end, once the theme is finished. All three are expensive to reverse after launch.
I run a Shopify agency that works only with beauty, cosmetics and DTC brands. What follows comes from builds for self-tan, haircare, skincare and colour cosmetics on Shopify and Shopify Plus, and it is the set of questions I would want a founder to have answers to before signing a scope.
Shopify allows three options per product. Apparel rarely needs more. Beauty does: a foundation has shade, coverage and size, and then a mini format and a refill pouch get added to the range. A skincare line has size, strength and scent. Once a fourth axis appears, the choice is between a fourth dropdown that sits awkwardly on the page, splitting the product into several products, or linking separate products with a combined listing so the storefront presents them as one.
The decision affects more than the product page. It sets what a shade swatch grid can do, whether reviews aggregate on one page or spread across a dozen, how collection filtering behaves, and how inventory maps to the 3PL. Reversing it after launch means new URLs, a redirect map, remapped review data and usually a rewritten product template.
Decide at the same time whether reviews and search authority should sit on one page or across several. For a hero product sold in twenty shades, one page is normally better. For separate lines that customers search for by name, separate products are.
An apparel shopper usually knows what they want and is checking fit and price. A skincare shopper often does not. They have a concern, an approximate idea of their skin type, and several products in front of them making similar promises.
In practice that means the concern and skin type appear in the customer’s own wording near the top of the page, ingredients are listed with a short reason for each rather than as an INCI block, the routine step is stated, and there is a line on expected timeframe. For colour cosmetics the equivalent is undertone guidance, swatches photographed on more than one skin tone, and a way to compare two shades without leaving the page.
Return rate shows whether any of this is working. In colour cosmetics a large share of returns comes from shade mismatch, so work here shows up in both conversion and returns.
A product recommendation quiz has the largest effect on conversion of any single feature in this category, and it is also the one most often built as a funnel that recommends the bestseller whatever the answers are. Customers work that out quickly.
A quiz that holds up has a data model behind it. Product attributes such as skin type, concern, texture and strength live in metafields, and the quiz maps answers to those attributes rather than to product IDs, so recommendations survive a catalogue change. The result page states the reason for the recommendation. Email capture comes after the result rather than before it. Answers are written back to the customer record so email flows can use them later, which is where a large part of the value sits.
We built a product recommendation quiz for Suntouched, a Canadian haircare brand, as part of a custom Shopify 2.0 build. Barrel built a skin-type quiz into the first US ecommerce site for Dr. Jart+, and PIVOT built a Shade Finder for Smith & Cult. In each case the quiz reads from the catalogue rather than duplicating it.
Beauty consumables have a refill cycle, which makes subscribe-and-save more useful here than in most categories. The app choice matters less than it looks. Recharge, Loop and Bold all do the job.
What determines whether a programme survives past month four is what the customer can do without contacting support: skip a shipment, swap a product, change the interval, add a one-off item to the next order. If those actions require a ticket, churn becomes a function of support capacity.
Cadence is the other half. A cleanser and a retinol do not run out at the same rate, so a single default interval across the catalogue leaves part of the subscriber base overstocked and part of it without product. Set the interval per product from observed usage.
On Klova, a sleep and wellness brand on Shopify Plus, we built a custom Recharge setup with a customer-facing management dashboard alongside a migration from Shopify 1.0 to 2.0. The dashboard took most of the development time on that project.
Regulation arrives earlier in beauty than in most categories. Ingredient restrictions, permitted claims and marketing wording differ between the EU, the UK and the US, so one product description is not usable everywhere. Payment mix and price points differ as well.
The structural decision is between Shopify Markets on a single storefront and separate storefronts. Markets is cheaper to operate and covers most brands. Separate storefronts make sense when content diverges by market, when local teams publish independently, or when catalogue availability differs enough that market-level exclusions become hard to manage.
Costs rise when storefronts are created country by country as the brand expands, because the same app stack is then licensed several times over. We consolidated Meroda Cosmetics, a Dutch brand, into five storefronts covering 27 active markets on Shopify Plus, which removed around $8,000 a month in platform and app cost.
Beauty brands usually arrive on Shopify from WooCommerce, Magento, a site builder or an older Shopify theme. Four areas account for most of the failures.
Subscriber records. Moving active subscriptions between platforms carries the highest risk in the project, and errors here charge the wrong card or cancel a cohort without notice. Schedule it separately from the site launch.
Reviews. Conversion in beauty depends on review volume and on review metadata at shade or skin-type level. Reviews that arrive as star ratings without their attributes have only partly migrated.
URLs. Product URL structure normally changes, and any shade split or merge enlarges the redirect map. It should be built before launch.
Theme content. Ingredient tables, routine guides and claims text hardcoded into old templates need to become structured data during the move, otherwise the new store carries the same maintenance problem.
Suntouched moved off a basic site builder onto a custom Shopify 2.0 theme. The project began with an SEO audit rather than with design, which is the sequence that protects existing traffic. The owner reported a 45% conversion increase after launch.
Beauty product pages carry more media than most: shade swatches, texture shots, before-and-after images, ingredient graphics and video. Most of it is served to a phone on mobile data.
Constraints worth writing into the brief: defer variant media so the page does not load every shade gallery at once, limit hero video, serve responsive image sizes instead of one large asset, and test the product page on a throttled connection before sign-off. Test the quiz the same way, since it is often the heaviest component on the site.
Sitewide conversion rate moves for too many reasons to show whether these decisions worked. Four measurements are more useful: product page to add-to-cart on the hero product, quiz completion rate together with the conversion rate of customers who finished it, subscription attach rate on first order, and return rate broken down by shade or product.
Our beauty and cosmetics Shopify projects are collected on one page, with the full case studies behind each brand.
Not at launch. Plus becomes worth its cost with multiple storefronts, checkout customisation or higher API limits. A brand running five storefronts across 27 markets is a clear case. A single-market brand at six figures usually is not, and the budget goes further in the build.
Usually after. A quiz built against a catalogue that is still changing has to be rebuilt. Launch with the product attributes already structured in metafields, then build the quiz once the range settles.
The constraint beauty catalogues hit first is the three-option limit per product rather than the variant count. Ranges that need more than three axes are handled by splitting them into separate products and linking those with a combined listing, so the storefront still presents one product.
Modelling the catalogue on the current SKU list. When the range grows a fourth axis, the product template, the redirect map and the review data have to be reworked at the same time. Working through catalogue architecture before design avoids it.
Ivan Tryhub is CEO and co-founder of SPLIT Development, a Shopify agency working with beauty, cosmetics and DTC brands from Tallinn, Estonia and Sacramento, California.
