How to Reduce Size-Related Returns on a Shopify Clothing Store
Shopify clothing stores often file look-mismatch as a size error. Split the codes, keep a size app for the label, and use Antla for how the garment looks.
Wrong-size returns on a Shopify clothing store usually mix two failures under one warehouse code. Too-small and too-big tickets need a size chart or size recommendation. Tickets that mean the garment looked wrong on the body need a photo try-on. Antla answers how it will look. It does not pick a size.
Open last week’s return comments. You will see “too tight in the shoulder,” “looked boxy on me,” and “ordered M and L, kept L” sitting next to the same Shopify reason: Too small. The portal is tidy. The file is not.
I’d start there, not in the App Store. Until you split label misses from look misses, you will buy another chart for a problem a chart cannot see.
A clothing return tagged “too small” is often a picture problem. The size selector and the try-on button do different jobs.
The comment is more honest than the code
Shopify lets apparel merchants collect product-specific reasons such as Too big and Too small. Those codes are useful for a weekly export. They are blunt for merchandising.
A shopper who followed your chart can still send the blazer back because the shoulder reads sharp on her frame. Another ordered the labeled size, then wrote “didn’t look like the photo.” A third bought two sizes on purpose. All three can land in the same bucket.
When I read tickets, I sort the free text first:
- The number on the label was wrong.
- The number was right, and the silhouette still felt wrong.
- She never trusted one size, so she shipped herself a fitting room.
The first job is grading and size guidance. The second is how the garment looks on a real body. The third is bracketing, which is a confidence failure that shows up as extra units and a refund.
Treat the code as a prompt. Treat the sentence as the diagnosis.
Clothing returns are expensive enough to justify the extra sorting
NRF and Happy Returns put 2025 U.S. retail returns at $849.9 billion, or 15.8% of annual sales. Online is worse: 19.3% of online sales. Shopify’s 2026 ecommerce returns guide reports those figures and notes that apparel and footwear sit above the online average because fit is hard to judge on a screen.
The same Shopify write-up, citing that NRF survey, says 82% of consumers treat free returns as an important factor when they shop online. Generous policy is how you win the order. It is also how a bad product page turns your warehouse into a fitting room.
Narvar’s 2022 State of Returns still names the mix: size and fit accounted for 45% of returns that year, up from 42% in 2021 and 38% in 2020. Narvar’s own read is that retailers had already spent on size-fit technology and were still missing.
That is not an argument for one more widget. It is an argument for matching the tool to the sentence on the ticket.
Split the file before you buy anything
Export 60 to 90 days of clothing returns by SKU family. Then reread the comments, not only the dropdown.
| What the shopper actually said | What failed | First fix |
|---|---|---|
| ”I’m usually M and this M was tiny” | Label / grading | Garment measurements and an honest chart |
| ”Ordered S and M, sent S back” | Bracketing | Size recommendation plus one-size incentives |
| ”Size was fine, looked huge on me” | Look / silhouette | Photography, fit notes, visual try-on |
| ”Too short,” “too revealing,” “didn’t suit me” | Look / proportion | Try-on, model height and size worn, review prompts |
| ”Hole in the seam,” “wrong item shipped” | Operations | QC and fulfillment. Leave the fit apps out of it |
If defects and damaged parcels dominate, stop. No size app or try-on will inspect a carton. If the expensive rows are dresses, denim, and tailoring with mixed “too small” and “not as expected,” you likely need both layers on the same product page.
When the comments mention S versus M, fix the label layer
A storewide chart for dresses, denim, tailoring, knits, and coats asks the shopper to do your merchandising. Put the variables that change the outcome next to the selector:
- Rise and inseam for denim
- Shoulder and body length for a blazer
- Bust, waist, and finished length for a dress
- Stretch and recovery for knits
- Layering room for outerwear
Write the fit note from return comments on that family. “True to size” is a shrug. “Structured through the rib cage, easier through the hip; take your usual size unless you are between sizes” is a prediction.
State the model’s height, the size worn, and whether the sample is meant to look fitted or relaxed. Ask reviewers for size chosen, usual size, and whether the piece ran short, narrow, or easy. Star ratings without body context do not cut wrong-size tickets.
If Too small and Too big still lead after that work, add a size-recommendation app. Kiwi Size Chart & Recommender publishes a free plan and a $7.99/month paid chart plan; size recommenders start on the $14.99/month Plus plan (App Store, 7 October 2026). Smart Size Chart & Size Guide publishes a free plan with two charts, Starter at $9/month, and “Find my size” on the $19/month plan (App Store, 7 October 2026). Those tools pick a label. They do not show the shopper the dress on her own body.
Why size charts fail on Shopify fashion is the longer version of that limit.
When the size was right and the piece still left, fix the picture
Look-mismatch hides inside size codes because shoppers reach for the nearest dropdown. “Too small” often means “too tight in a way I did not expect.” “Too big” often means “too boxy.” The warehouse hears size. She experienced an expectation gap.
Shopify’s returns guide tells merchants to govern product information first: dimensions, materials, size guides with the model’s size, images that show how the garment actually fits. Then it recommends 3D and AR “where fit drives returns.” Its worked example is Gunner Kennels placing a crate next to a dog, not a clothing PDP. Photo try-on is the clothing version of that idea: the shopper sees the garment on herself before checkout, without a 3D asset pipeline.
That is the “how will it look” question. Length, silhouette, coverage, color against her skin. It is not size selection. A preview can look right in a size that still pinches at the waist. Keep the size tool next to the selector.
A named public case for running a size tool and a preview together is Snap’s Princess Polly result: shoppers who used Fit Finder and AR Try-On had a 24% lower return rate than shoppers who did not (Snap internal data, 1 July 2020 to 31 October 2022), reported in Snap’s AR Enterprise Services launch. Fit Finder is a size tool, so that figure is not a clean try-on-only result. It still supports running both layers. Caveats live on does virtual try-on reduce returns?
What Antla is for on a clothing product page
Antla virtual try-on is Shopify photo try-on. The shopper uploads one photo in the browser and sees the garment on herself. She does not download a shopper app.
Antla has no size chart or size recommendation of its own. It works alongside any Shopify size app. The Shopify listing line is the right one: size charts tell customers what fits; Antla shows them how it looks. When that listing says Antla works with size charts, unit converters, and sizing templates, it means it pairs with those tools. It does not include them.
Keep the two calls to action distinct. “Find my size” and “Try it on” do different work. One button hides the size tool from shoppers who needed a label and hides the preview from shoppers who needed a picture.
Antla’s privacy policy states: Images are deleted and not accessible within 72 hours.
Antla is Built for Shopify certified. It is not SOC 2 or ISO certified.
Plans on antla.io/pricing, checked 7 October 2026: Trend $19.99/month (100 try-ons, then $0.16), Runway $49.99/month (500, then $0.12), Unlimited Fashion $199.99/month (2,000, then $0.09). Every plan includes a 7-day free trial. There is no free plan. Try the flow on your catalog at https://antla.io/demo before you touch the live theme.
Antla has not published a confirmed first-party return-rate delta. Do not import one from a vendor slide. Your tagged reasons are the evidence.
Score it on the reason codes, not on a launch email
Pick one high-cost category. Dresses, denim, or tailoring are the usual candidates.
Baseline four weeks of comments. Separate too-small / too-big from not-as-expected, quality, and changed mind. Then add only the missing layer: size recommendation if the file is labels, try-on if the file is look, both if the file is mixed.
Hold placement constant. A try-on button three scrolls below add to cart produces a weak cohort.
Compare try-on users with non-users on the same SKUs for conversion, units per order, adjacent-size orders, and return rate within 30 days of delivery. Review at 60 to 90 days so late returns show up. Do not mix denominators. A try-on-user return rate is not your store-wide return rate. People who preview may already be more engaged. Treat the gap as an association until you run a holdout.
Feed the same comments back into merchandising every week. If one dress family keeps coming back as “too short in the torso,” rewrite the length note before you blame the app.
The useful goal is not zero returns. Fashion will always send some pieces home. The useful goal is fewer avoidable wrong orders created by a page that never answered the question she was actually asking.
Frequently asked questions
Why do Shopify clothing stores still get wrong-size returns with a size chart on the page?
Because many tickets tagged Too small or Too big are look problems. The chart can be correct and the shoulder, rise, or length still feels wrong on her body. Read the comments. Fix grading and size recommendation for label misses. Fix photography and visual try-on for look misses.
Which should I add first, a size recommender or virtual try-on?
Whichever matches the comments. If shoppers write about S versus M, start with garment measurements and a size app. If they write that the size was fine and the piece still looked wrong, add photo try-on. Dresses, denim, and tailoring often need both on the same page.
Does Antla include a size chart or tell the shopper which size to buy?
No. Antla has no size chart or size recommendation of its own. It is visual try-on for how a garment looks, and it sits beside any Shopify size app you already run.
How long should a returns team wait before judging try-on?
Use four weeks of tagged reasons as a baseline, then give the pilot 60 to 90 days so delivery and late returns are in the file. A two-week conversion bump is not a returns result.
Which return reasons should I leave off the fit-tool list?
Defects, damaged parcels, wrong-item shipments, and wardrobing. Those are QC, packaging, fulfillment, and policy. A size app and a try-on will not inspect a carton.
What happens to the shopper’s try-on photo?
Shoppers upload the photo in the browser. Antla’s privacy policy states: Images are deleted and not accessible within 72 hours.
Put the picture back on the product page
Some returns are normal. The expensive ones are the ones you could have prevented by answering the right question on the product page. Put try-on on the SKUs where look-mismatch is hiding inside a size code.