A size label is not a measurement. “M,” “10,” or “42” only becomes useful when a shopper can translate it into the dimensions of this product, how those dimensions relate to their body, and how the product is intended to feel. Good Shopify size chart UX performs that translation without demanding technical pattern-making knowledge.
This topic is narrower than general product-page psychology and earlier than the returns and exchanges experience. The goal is not to promise that returns disappear. It is to make the pre-purchase size decision legible, honest, and recoverable.
What makes a good Shopify size chart?
A good Shopify size chart is product-specific, clearly distinguishes body measurements from garment measurements, explains how and where to measure, supports relevant units, and sits next to size selection. Fit finders should add transparent recommendations and uncertainty—not replace accurate source measurements.
Build One Source of Sizing Truth
Sizing data often arrives from multiple vendors, spreadsheets, product tags, and app fields. Before redesigning the interface, define which data is authoritative for each product or family. If a supplier chart conflicts with a measured sample, interface polish cannot rescue the experience.
| Information | Shopper question | Required label |
|---|---|---|
| Body measurement | What size fits my body? | Body bust, waist, hip, inseam |
| Garment measurement | How large is this item? | Garment chest, waist, length, rise |
| Fit intent | How should it feel? | Slim, regular, relaxed, oversized |
| Model context | How does it look on someone? | Model measurements, worn size, styling note |
Never mix body and garment values
A shirt can measure wider than the body it is designed to fit because it includes ease. If a table says “chest 42 inches” without identifying whether that is a body range or the garment laid flat, shoppers can interpret the same value in opposite ways. Put the measurement type in the heading, not buried below the table.
Define the measuring method
Explain whether width is measured flat and doubled, where waist placement begins, whether inseam follows the inside seam, and whether fabric is stretched. Use a simple diagram when words are ambiguous. State a reasonable tolerance if manufacturing variation makes exact values unrealistic, but do not use tolerance language to excuse unreliable data.
Treat units as persistent state
Offer inches and centimeters where markets need both. The chosen unit should persist between products and should change labels and values together. Do not convert table fields while leaving model measurements or explanatory copy unchanged.
Put Help at the Decision Point
The primary guide trigger belongs next to the size selector. A guide hidden under a long description, inside a collapsed FAQ, or in the footer requires shoppers to leave the decision to find the evidence.
On mobile, a bottom sheet can work when it has an accessible name, predictable focus, a visible close control, and enough height to read without fighting nested scrolling. A dedicated page can work for technical products if it preserves the product and selected variant on return.
Keep the selector visible after guidance
When the guide closes, return focus and scroll context to size selection. If a size was recommended inside the guide, reflect it in the selector only after clear confirmation. Silent selection changes can lead to accidental orders.
Show unavailable sizes without erasing context
Sold-out sizes should remain legible and marked unavailable because their absence makes the range hard to understand. A waitlist can be offered after the shopper identifies the needed size; it should not distort the recommendation toward available inventory.
Write Fit Language That Carries Information
“True to size” is often too vague. Better notes describe observable behavior rather than relying on an undefined baseline.
- Fitted through shoulder and chest, with a straight waist.
- High rise with limited waistband stretch.
- Relaxed through hip and tapered below knee.
- Designed to layer over a midweight knit.
- Fabric has two-way stretch but the waistband does not.
If reviews contain structured fit feedback, show the sample size and distribution rather than collapsing opinions into one unexplained badge. Self-selected reviews can be useful context, but they are not a controlled sizing standard.
Use model information carefully
Model height and worn size offer visual context but do not replace a chart. Add relevant body measurements and note intentional pinning or styling when it changes silhouette. Include more than one body reference where feasible. Matching model height alone does not determine size.
Decide Whether a Fit Finder Earns Its Complexity
A fit finder can reduce cognitive work across a complex catalog, but only if inputs map to reliable product-level grading and fit rules. Installing a questionnaire over inconsistent data adds confidence theater.
Inputs should be necessary and explainable
Ask only for information that affects the recommendation. Explain why sensitive or unexpected information is requested, how it is used, and whether it is stored. A usual size from another brand is useful only when the reference product and mapping are maintained.
Recommendations need uncertainty
“Recommended: M” is incomplete. Better output says, “Choose M for the intended close fit; choose L if you prefer more room through the waist.” Show the answers driving that result and let the shopper edit them. When evidence is insufficient, say so and route to the chart or support.
Do not optimize to inventory
If the best-fit size is unavailable, say that and offer a size-specific notification. Suggesting a poorer fit because it can ship today damages trust and contaminates performance data.
Handle Category-Specific Edge Cases
Footwear may require foot length, width, last shape, sock assumptions, and market conversion. Bras need band and cup relationships. Compression products require fit intent and safety-conscious guidance. Children’s sizing may use height, weight, and age ranges while making clear that age is only a proxy.
Unisex products should explain the grading basis instead of merely relabeling one run. Adaptive products should describe closures, access openings, seated fit, and relevant dimensions. Bundles must state whether one size applies to all components or each can be selected separately.
Make Tables and Tools Accessible
Use real table markup with caption and associated headers; the W3C tables tutorial explains those relationships. Do not publish core sizing values as an image. Ensure zoom and text resize work, and avoid forcing shoppers to pinch a wide table.
Fit-finder questions need programmatic labels, keyboard-operable choices, visible focus, clear errors, and progress that is not communicated by color alone. Guidance should remain understandable without animation or hover.
Measure Confidence and Consequences
- Was guidance available and seen near size selection?
- Did the shopper inspect measurements, change units, or complete meaningful fit-finder steps?
- Did the selected size change after use?
- Did the session proceed to add to cart and purchase?
- Was a later exchange or return coded to size or fit, and which reason?
Usage is confounded by uncertainty: shoppers with the hardest decisions are more likely to open the guide. A higher return rate among users does not prove the guide caused returns. Compare like products, acquisition contexts, and customer types, and use experiments where traffic supports them.
Operational QA checklist
- Audit source data against physical samples and supplier updates.
- Label body versus garment measurements in every view.
- Verify conversions, translated labels, and rounding.
- Test long tables, small screens, keyboard flow, and screen readers.
- Confirm recommendations do not change with inventory availability.
- Review size-related contact and return reasons for language gaps.
- Version changes so support can recover guidance shown on old orders.
Key takeaways
- Size guidance translates product measurements, body measurements, and fit intent.
- Accurate product data is more important than a sophisticated widget.
- Place guidance beside size selection and return shoppers to the same context.
- Replace vague “true to size” claims with observable fit behavior.
- Fit finders should explain recommendations, uncertainty, and data use.
- Never steer recommendations toward available stock.
If sizing logic is fragmented across vendors, templates, and apps, CROVEX can audit the path from source data through mobile interaction and analytics. Review our product-page optimization work or book a free Shopify audit to turn fit guidance into a maintainable system.
Does your size guide create confidence or more questions?
CROVEX reviews sizing data, fit language, product-page placement, mobile interaction, accessibility, and outcome measurement.
Book Free Shopify AuditFrequently Asked Questions
Include product-specific size labels, clearly identified body or garment measurements, units, measurement instructions, fit notes, and a conversion for the markets you serve.
Place it adjacent to size selection and make it available before an unavailable-size error. Repeat a compact link near fit notes when useful, but avoid hiding the only access in a footer or accordion far away.
No. Body measurements describe the shopper; garment measurements describe the physical item. Mixing them without labels can lead shoppers to choose incorrectly because garments include design ease.
No. A simple, accurate chart may be better for small catalogs or products with straightforward fit. A fit finder earns its complexity when product grading, fit preference, and shopper inputs can support a reliable recommendation.
State the recommended size, the inputs used, relevant fit tradeoffs, and what to do when the shopper sits between sizes. Avoid presenting a recommendation as a guarantee.
Open an accessible sheet or page with a stable header, readable tables, persistent unit controls, and easy return to the selected variant. Do not force horizontal pinching to read core values.
Track guide exposure, meaningful use, size selection changes, add-to-cart progression, exchanges by size reason, and support contacts. Compare cohorts carefully and avoid claiming causation from usage alone.