Buying Guides
Size Charts That Cut Returns: Measurement Specs and Tolerances
A factory guide to sizing returns: the real causes, tolerance windows factories can hold, and the size-chart discipline that protects your review score.

Sizing is the largest single return driver in apparel, and most of it is preventable with two unglamorous tools: measured specs and honest tolerance windows. This page explains where sizing returns actually come from, what a factory can hold, and how to publish size charts that survive contact with real bodies and real lots.
Where sizing returns actually come from
Three buckets dominate. Chart mismatch: the garment does not match the published measurements. Expectation mismatch: the garment matches the chart but not the buyer's mental model of the fit. Lot drift: the re-order does not match the first order. Each bucket has a different owner, and only the second one belongs to the customer.
Measured specs versus graded specs
A size chart is a promise about the garment. It should be written from the sealed sample with a tape, not copied from a grade table. Grade tables describe intent; sealed samples describe reality. The gap between them is where returns live. Our practice: measure the sealed sample at every listed point, publish those numbers, and re-measure at bulk sign-off.
Tolerance windows factories can hold
Every measurement carries manufacturing variance. Holding a factory to zero is fantasy; leaving tolerance undefined is an invitation. Working bands for activewear knits run small: a few millimetres on chest and hip, slightly more on length. The number matters less than the agreement: name the tolerance per measurement point in the PO, so both sides know what pass and fail means before production, not after a return wave.
How to sample-measure and publish honest charts
Measure garments, not patterns. Lay the garment flat, use consistent tension, and measure the points buyers actually check: chest or bust, waist, hip, inseam or length, and rise where fit depends on it. Publish flat measurements with a note on stretch range, because a compression garment measures one way flat and another way worn. A stretch note, this garment stretches to fit the following ranges, prevents a whole class of returns.
Body-data sources for your audience
Charts land better when they map to your market's bodies. Use return data as a body-data source: which sizes return, for which reasons, is a live survey of where your chart meets reality. Supplement with standard size surveys for your region if you sell across markets, because a medium is not the same medium everywhere.
Fit signals hidden in return reasons
Read return reasons as engineering signals. Runs small comments cluster where the grade skips. Baggy-out comments track fabric recovery failures. Size-up advice in reviews marks a chart that under-promises. Each pattern has a fix: re-grade, re-specify recovery, or re-publish honest numbers.
Updating charts across fabric changes
A fabric swap changes measurements: a different elastane share shifts stretch range; a heavier knit shifts fit perception. Any fabric change should trigger a chart review with a re-measured sample. The re-order complaint, my second order fits differently, is almost always a chart that stayed home while the fabric travelled.
Communication: fit notes that prevent returns
Two sentences prevent returns better than a bigger chart: this style runs true to size with compression fit, and if you are between sizes, size up for comfort. Fit notes transfer the mental model the chart cannot. Write them from real fit sessions, not from marketing.
What not to promise
Do not promise perfect fit across all bodies; do promise accurate measurements. Do not promise zero variance; do publish tolerances internally and hold the factory to them. Precision in language earns the trust that vague confidence loses.
Common mistakes
Copying charts between styles with different fits. Publishing grade-table numbers instead of measured ones. Skipping the bulk re-measure. Changing fabric without re-checking the chart. Ignoring return-reason text, which is free fit research.
Questions buyers ask
What tolerance is normal in clothing sizes?
Knits commonly hold a few millimetres on girth measurements and a little more on lengths. The specific numbers belong in your PO as a tolerance table per measurement point.
Why do sizes differ between brands?
Different blocks, markets, and fit intentions. Charts are brand-specific promises, which is exactly why measured numbers beat letter sizes for online trust.
How do you measure leggings for a size chart?
Flat, consistent tension, at named points: waist, hip, rise, inseam, thigh, ankle opening, and publish the stretch range alongside. Measure the sealed sample and re-measure at bulk.
Can size charts reduce returns?
Yes: accurate charts, stretch notes, fit notes, and lot-change re-checks attack the two biggest return buckets directly. The chart is cheap; the returns are not.
A monthly chart review routine
Size charts drift because everything around them does: fabric lots, factories, even grade reviews. A monthly routine holds the line. Pull the month's returns and tag every sizing-related one by style, size, and reason phrase. Compare against the chart in force at the time of sale, not today's chart.
Where a pattern shows, three comments or one percentage threshold, order a re-measure of a live sample from stock, not from the pattern room's memory. Confirm or correct the published numbers, and date the change.
Close the loop with a note in the chart's history: what changed, why, and which returns triggered it. Twenty minutes a month turns the chart from a publishing artifact into a control system, and the return curve tells you within a quarter whether the system is working.
A tagging taxonomy for return reasons
Return-reason text is only as useful as your tagging. Build a small taxonomy and enforce it at entry: too-small, too-large, fit-shape, changed-mind, quality, other, with fit-shape carrying a free-text note. Train customer service to tag at refund time, when the context is fresh.
Review the tags monthly beside the size data. Too-small clusters mark grade skips or optimistic charts. Fit-shape notes are pattern gold: they tell you which body feature the block misses, which is information no survey buys as accurately.
Six tags, one dropdown, and your returns become a research panel that runs itself. The chart fixes that fall out of it are the cheapest product development your brand will ever conduct.
A last operational note: share the tolerance table with customer service. When a return question arrives, the agent can answer from the same numbers the factory signed, and even explain the acceptable range honestly. Alignment across departments reads as confidence to customers, and confident answers reduce the reflex return more than any discount code. The chart is not just a promise to buyers; it is a script for everyone who speaks for the brand.
Bottom line
Sizing returns are a measurement discipline problem wearing a customer-service costume. Measure sealed samples, name tolerances in the PO, review charts at every fabric change, and mine return reasons for the fix list. Your return rate becomes a managed number instead of a monthly surprise.
Send your return-reason export to info@breezsportswear.com or WhatsApp +86 135 1661 0926. We will map the signals to spec fixes and return a measurement and tolerance sheet for your next order.



