Suppose Maya clicks Medium on the company store’s new fleece during a preview in Boston. The price and shipping promise are right for Massachusetts. The size chart, however, says “54” under a column marked inches. She has 180 employee invitations scheduled for the afternoon, and the supplier will decorate the first batch after the order file is approved. She can fix the label, delay the fleece, or send the invitations as planned. Before she decides, she needs to establish whether 54 is a valid garment measurement for the exact Medium in the cart.
The practical comparison is about the connection between a chart and an order. Shopify Markets with Translate & Adapt is a reasonable first test when your Shopify catalog already owns the product and the theme can read the chart from an identified field. WooCommerce with WPML merits a test when your team controls the WordPress stack and can make the language-specific product records and custom fields behave consistently. Adobe Commerce store views merit a test when your catalog already depends on defined attribute scopes and store-view overrides. In all three cases, require the correct garment, chart revision, units, and fallback to survive an actual purchase-path test before releasing the item.
Decide what the chart is supposed to measure
Start with the supplier's approved specification. An employee comparing a favorite sweatshirt laid flat across a desk needs a garment width measured the same way. Someone following a recommendation based on body chest circumference needs a chart explicitly built for that purpose. A supplier’s width measured straight across the garment is neither the wearer’s circumference nor a promise that a garment will fit everyone with a particular body measurement. Show the measuring points and distinguish garment measurements from body guidance before any localization begins.
Set one unit as the source of record for each measurement. Store the raw value, unit, measurement method, and any supplier-approved tolerance together. A label such as “Chest” is ambiguous if one supplier means the width of a laid-flat garment and another means the wearer’s circumference. The same problem applies to sleeve measurements taken from a shoulder seam versus the center of the back. A translation reviewer should know what the number describes, where to start the tape, and whether the listed tolerance applies to the garment specification or to rounding in the display.
Give the chart a durable identifier. For example, a midweight fleece might use chart F24 revision 3, approved for the standard fit in sizes XS through 3XL. Record which supplier model and fit that revision covers, along with the selected color or trim if construction changes between options. A person reading a catalog in a second language should be able to choose a familiar size label while the fulfillment system receives the approved, orderable child variant. When a replacement supplier uses the same customer-facing name, create or update the explicit chart mapping before the replacement can be ordered.
A workable catalog separates three things. The first is what the supplier makes, including the variant identifier and measurements. The second is what each audience reads, including units, instructions, and accessible labels. The third is what operations approves for a particular route and revision. One administrator can perform all three jobs in a small team. The records still need separate names and checks because a copy edit, a factory change, and a market release affect different parts of the order.
| Storefront setup | First chart route to test | Proof before acceptance | When to pause |
|---|---|---|---|
| Shopify Markets and Translate & Adapt | Structured measurements in a known product field with an explicit chart mapping and localized theme rendering | Correct market and language route, exact field read by the theme, matching selected variant in cart | The chart is owned by an app or a template with unverified localization or fallback |
| WooCommerce and WPML | Shared measurements in synchronized fields, translated instructions in explicitly configured fields | Product translation linkage, child variation mapping, field sync and no stale image | Source edits or imports overwrite local instructions or leave translated variation links unresolved |
| Adobe Commerce store views | Shared orderable child identity and store-view instructions or chart presentation | Catalog scope, default-value behavior, storefront render, order SKU and tested restore procedure | A view inherits an unapproved chart or a global edit changes other markets |
Shopify Markets and Translate & Adapt
Start by opening the exact country and language combination you intend to send to employees. Shopify’s market domains and languages guide explains that submarkets can inherit those settings and that language and market-specific content can be configured separately. The presence of an English page does not establish that the United States route is in use. Verify the selection and effective URL, then repeat with a browser session that has no saved market preference. Use the same invitation path the buyer will actually receive.
Shopify’s Translate & Adapt app listing describes manual and automatic translation of supported storefront content and market adaptations. That documentation provides a starting point for product copy. For the chart, inspect the live theme and trace exactly what it reads. A section translated in the theme editor, a file referenced by a product metafield, and a modal populated by an app can appear in the same place on screen while taking three different routes to get there. Shopify describes file-reference metafields available to theme code. Test the chosen resource and the actual template together, including the missing-value behavior and any script that updates the modal when the shopper changes size or fit.
Give the person responsible for production a list of variant IDs and the chart revision approved for each applicable fit. Toggle the variant on a mobile product page and check the table and chart trigger both before and after the toggle. Add the variant to cart, open checkout far enough to see the order selection in a test environment, and compare the cart selection with the expected variant. If a product can be opened from a collection, from a direct invitation link, and through a saved cart, sample those paths too. A good preview in the translation editor is evidence about that editor. The receipt you need is the storefront state that the buyer saw and the variant selected on the order.
This is a good fit for a team that has one main product catalog, manageable theme customizations, and a way to keep all chart inputs in a documented location. A merchant with a third-party chart app should put the app, theme, and Markets behavior on the same acceptance ticket. Ask what happens when a source measurement changes after a translation was saved. Shopify’s theme configuration documentation explains how market-specific theme settings are inherited. As part of acceptance, check for a missing translation or a chart that still shows an older revision. A support screen or green translation status alone does not prove that the variant-specific chart is current.
WooCommerce with WPML
WPML describes WooCommerce translations as linked product records. That architecture is useful when a team needs different editorial treatment for different audiences. It also gives the team multiple records whose relationships must survive edits and imports. Confirm the parent product translation and the child variation association. Compare what the website displays with the SKU that the operations system receives. There can be separate internal database IDs in different language records while operations retains a single physical variant identity for fulfillment.
Classify every chart input. The measuring instruction might be translated. The supplier width should be shared when it describes the same garment. An approved localized diagram might legitimately differ by language. WPML’s custom field documentation explains field preferences including Translate and Copy. Configure the field behavior intentionally, then change a measurement in a test source product and observe what happens to the translated record. If the chart plugin saves opaque data, custom post types, or a list of attachment IDs, ask the plugin vendor how those values are stored and how its runtime selects the language. Verify the answer with the installed versions.
The highest-value failure test is usually a reorder or import. Take a source product with two fits and the same visible size names. Update the garment width for only one fit, change the order of variation options, and refresh the localized page from a clean session. Record whether the intended variation keeps its chart, whether the second fit remains untouched, and whether the sales or reward order carries the expected identifier. In a test copy, confirm that deleting or removing a localized chart produces the agreed behavior. A fallback that looks sensible while the original garment is still in stock may become misleading after an import swaps models.
WooCommerce with WPML can be a sensible choice for teams already comfortable owning plugin compatibility, staging, and regression tests. The WPML WooCommerce guide documents product and variation translation workflows. Apply the same operational tests to the chart plugin and any ERP or warehouse export. The team that fixes a missing measurement must be able to find which field generated the display, verify the sibling language, and show that the fix did not alter existing purchase records.
Adobe Commerce store views
Adobe explains that websites, stores, and store views operate at different scopes. Its store localization guide separates the interface locale from translations of product content. For an apparel chart, decide which numbers and mapping keys belong to the common catalog and which pieces of presentation need a store-view override. Check the scope of the actual attribute and the behavior of any frontend component that renders it.
Adobe’s product attribute documentation describes Global scope for dropdown attributes used for configurable options. You can therefore test the underlying option and its localized label separately. For example, a size label can be understandable in the local view while the configurable product resolves to the same physical child item. Keep the chart separate from the label logic. A change to “Medium” on screen does not certify a measurement, and a chart override should not force you to rebuild the child variant.
Inspect the inheritance state in the store view before editing anything. For supported localized product fields, Adobe describes clearing the default-value checkbox when entering a translation. Rehearse what happens if an editor restores the default or if a catalog import updates the common value. Check whether the selected view uses the new content, the previous approved version, or an unapproved inherited value. Then test the employee-facing route, including the actual commerce frontend, the exact view and a sample order. On a custom frontend, ask which view identifier and attribute response populate the chart.
Use this approach if the team already has defined catalog roles, access to scoped configuration, and a repeatable deployment or import process. Bring the person who owns catalog data and the person who owns the frontend to the same test. They should be able to say whether an unexpected chart came from a global attribute, a view override, a module, or a cached presentation. Ask for a reversible preview and a restore procedure that preserves orders and newer catalog edits. Availability and scope of preview tools depend on the installed setup, so verify the chosen recovery method in that environment.
Worked decision 1: One garment, two display units
Maya asks the supplier for the signed chart. It lists Medium as 54 centimeters measured flat across the garment from the stated points. The product record confirms both market routes are selling the same fleece and construction. The correct conversion is 54 divided by 2.54, about 21.26 inches, or 21.3 inches at one decimal place. The NIST length units reference gives 2.54 centimeters per inch. Store 54 centimeters as the approved source measurement. Display 21.3 inches on the US route with the same flat-measurement instruction, and show 54 centimeters where that is the chosen display unit. Use the raw number for any subsequent conversions.
She should also test an adjacent size. If Large is 57 centimeters, its one-decimal display is 22.4 inches. That creates a visible difference of 1.1 inches, while the source increment is 3 centimeters, approximately 1.18 inches. A buyer should not infer exact grading increments from differences between rounded labels. The table can name its rounding convention. If the vendor confirms a manufacturing tolerance, use the same source and convert the tolerance separately. Keep the tolerance distinct from the decision rule for choosing a size between rows.
The team tries Medium and Large on the intended route, opens the chart from each selection, verifies the supplier reference, and records the orderable child SKU for each. An independent teammate compares the displayed widths and order lines with the source. Once both checks pass, Maya releases the fleece with the approved chart revision. She keeps the page screenshot, destination route, timestamp, mobile view and test-order line in the same approval record. The captured order proves which garment was selected, and the screenshot proves the guidance that was visible at the decision point.
Worked decision 2: A good-looking fallback belongs to another model
Suppose a new supplier makes a replacement fleece under a familiar collection name. The old Medium is 54 centimeters wide flat, and the new Medium is 51. The difference is 3 centimeters, about 1.18 inches before display rounding. Translating the older diagram would leave the new product with an incorrect measurement. So would carrying over the former supplier's chart because its image is still attached to the product template.
The operator tests the storefront with the localized chart removed. The page renders the older image even though the cart selects the replacement. That is a failing purchase path. If the site supports a verified product or market restriction that isolates the affected fleece, hold just that selection. A page-level banner alone is insufficient if a shopper can still add the item from a collection card or reopen it from a saved cart. Where no narrow restriction works, suspend the affected item at the smallest scope that actually prevents orders and verify an unrelated item remains orderable.
An alternate can still serve employees while the fleece is held. For example, offer an approved one-size item through the same invitation with a clear choice and the same eligibility checks, or supply an accessible, revision-controlled chart on the product page if the exact variant and purchase path can be verified. Restore ordering only after the new measurements and revision appear for the replacement in a clean session, the obsolete fallback is absent, and the cart records the replacement SKU. If any orders arrived during the incident, preserve the original lines and route affected selections for human resolution before decoration. Do not replace a selected size automatically.
What to put in the acceptance ticket
Use one representative product, two adjacent sizes, a second fit if offered, and one deliberate missing-chart test. Write the expected outcomes before the demonstration. Include the destination route and locale, supplier model, chart revision, current template or component, measurement method, displayed units, selected child variant, and a test order reference. Pair screenshots with structured order evidence. Use test data and store the smallest amount needed to reproduce the decision.
Give the team a simple restore drill. First capture the before state and identify the exact field or configuration. Change only the intended chart revision. Verify the chosen market and a sibling, then restore the previous approved value in staging and repeat the test. If the prior chart describes a discontinued garment, keep the product unavailable while you restore the ability to render a valid chart. In production, rollback means recovering a known-good purchasing state. Reattaching a dated image because it is available does not accomplish that.
Check practical accessibility during the same drill. A native text table with headings lets someone understand how rows and columns relate, enlarge the text, and compare sizes without decoding a bitmap. If you include a diagram, provide a separate text explanation of the measurements and keep its language and revision aligned with the table. Try opening and closing the chart with a keyboard, reading it on a narrow screen, and returning focus to the selection. Record the observed behavior and assign a fix owner when the theme or app fails.
Run this small acceptance suite again whenever the garment specification, size mapping, field configuration, translation workflow, template, or chart application changes. For each market, name one person who can hold the listing and one person who can approve reopening. A release decision should follow the visible chart and test order together. Where a vendor cannot demonstrate them on the intended stack, mark the capability unverified and retain the evidence that would close the gap.
When a managed storefront belongs in the comparison
If the decision is how to operate a company apparel program, the team can compare a managed storefront with self-operated catalog software. Giftpack’s branded storefronts describe merchandise and store preview alongside production, inventory planning, and fulfillment. Ask for the same demonstration on a proposed Giftpack program: one real garment specification, two units where applicable, one intentionally missing local chart, and the exact order line. Confirm who owns changes and whether the affected route can be held while buyers retain an approved choice. The best evidence for Maya is the same whichever operating model she selects. The right garment was offered, the sizing guidance matched that garment, and the order recorded what the employee chose.