Fashion Brands Already Hold the Measurements That Could Cut Online Returns

Size and fit rank at or near the top of return reasons wherever online clothing returns are measured, and each return can cost a retailer more than half the product’s retail price. Many shoppers now order two sizes and send one back. Brands already hold the data that could reduce this, in specification sheets that rarely reach their ecommerce teams.

Long Story, Cut Short
  • Standard size charts describe bodies rather than garments, so shoppers cannot compare an item’s real measurements with clothes they own.
  • Return reasons sorted by style and size reveal whether a chart is wrong or bulk production drifted from its specification.
  • Online return rates for individual items at one European retailer ranged from 13% to 96%, against 3% in its stores.
The measurements that decide whether a garment fits are often recorded with great care in production and then lost before the product meets its buyer online.
HIDDEN DATA The measurements that decide whether a garment fits are often recorded with great care in production and then lost before the product meets its buyer online. AI-Generated / magnific

Open the tech pack for almost any kurta and the measurements are exact. Across shoulder, chest, front length, waist and hip are graded size by size, each with a tolerance the buyer signed off before bulk production began. The factory works to those numbers, QC checks against them, and a shipment that misses them can be rejected.

Then the same kurta goes online. The product page shows a five-row chart of body measurements, copied from a template the brand uses for every style it sells, and quite often it is a single image. The garment's actual measurements, the ones the factory spent weeks getting right, never reach the shopper.

That gap between the spec sheet and the product page is where a large share of online fashion returns begin. Closing it needs no new technology, because the data already exists and stops at the wrong desk. Prashant Kumar made the same case for AI imagery, which has to be tested against the real garment before it sells anything, and garment measurements need the same check.

How Much of the Fashion Returns Problem is Fit

The numbers vary by market, but size and fit sit at or near the top almost everywhere returns are measured. Coresight Research estimates the average US online apparel return rate reached 23.4% in 2025 [1].

The European Environment Agency puts the return rate for clothing bought online in Europe at around 20%, with 70% of those returns caused by buyer-perceived poor fit or style [2]. Zalando says half of the items ordered across its markets come back [3], and that size and fit account for up to half of returns in European online fashion [4].

Indian data is thinner and needs reading with care. Unicommerce, drawing on about 900 million transactions, reported customer-initiated returns across all categories falling to 7.8% in 2024, with fashion and accessories making up 30% of order volume [5]. Category rates run far higher. Return Prime data reported by Apparel Resources put returns of clothing and shoes as high as 30% to 35%, with size concerns behind almost 40% of returned merchandise [6]. Rest of World reported in 2023 that Myntra and Flipkart had started charging some customers who return too often [7], as ASOS has done with fees for customers who return lots of items, which is a fair guide to what returns cost the platforms.

The cost is more than a reverse shipment. The British Fashion Council's work on returns estimated that processing each online return costs a retailer 55% to 75% of the product's retail price [8]. The EEA estimates that between 22% and 43% of clothing returned after an online purchase in Europe is destroyed [2], and a US study found that more than 90% of the emissions from fashion returns come from transport alone [19].

Shoppers have their own workaround. They order two sizes and send one back. In a KPMG survey of 3,000 UK consumers, 13% said they buy several items in one order to find the size that fits and return the rest, and a third said they only ever return an item when the size is wrong [9]. Bracketing is how shoppers cope with a size chart they don't trust, and each bracketed order ships with at least one return already planned.

Where the Size Data Gets Lost

The tech pack holds finished garment measurements at each point of measure, graded from a base size by a set increment, with tolerances either side. The fit sample session adds comments (tight across the bust, sleeve two centimetres long) and the grade is corrected. Bulk production is then measured against the approved spec, and the buyer's QC report confirms it. By the time a style ships, the brand knows precisely what an M measures.

The ecommerce team rarely sees any of this. They receive photographs, a colour name, a fabric description and a list of sizes. The size chart comes from the brand's standard template, usually a table of body measurements for S to XXL written once and applied to every top, kurta and dress in the range. A relaxed kurta and a slim one get the same chart. When a style runs small because a fit correction came late, the chart has no way of saying so.

The standards bodies don't close the gap either, because they describe bodies rather than garments. ISO 8559-2 (2025 edition) and EN 13402-3 both build size designation on body dimensions [10][11], and EN 13402-3 states outright that garment dimensions are not contained in it [11]. India's own INDIAsize survey works the same way. NIFT scanned 26,324 people aged 15 and over and produced 14 male and 13 female body size charts, with the full charts on the INDIAsize portal from January 2026. National standards for an "IN" size tag are being developed by the Textiles Committee under the Ministry of Textiles [12].

That is valuable work. A brand that builds its base block on Indian body data should see fewer fit failures before a single return is logged. But a shopper choosing between an M and an L still needs to know what the garment measures, and only the spec sheet holds that.

Why Platforms Now Read Size Data

For years a vague size chart cost a brand returns and little else. That has changed, because the platforms sending traffic now read size data and act on it.

Google requires the size attribute for clothing and shoes in free listings, and for Shopping ads in Brazil, France, Germany, Japan, the UK and the US. Its help pages say submitting a size lets a product appear in results filtered by size and affects how variants are shown [13].

Two more attributes feed the same filters, size_type (regular, petite, plus, tall, maternity and so on) and size_system [14][15]. A product with a missing or inconsistent size is unlikely to appear once a shopper filters for theirs.

Indian exporters hit a catch here. Google's list of accepted size systems runs from AU to US, and IN is not on it [15]. A brand selling abroad has to map its sizes to the buyer's market, usually UK or US, and that mapping is only as honest as the garment measurements behind it.

Marketplaces are further along. Amazon uses machine learning to identify defects in size charts [16], and its Fit Insights tool, which analyses returns data, is open to apparel and footwear brands enrolled in Amazon Brand Registry in India [17]. Myntra's sizing rules, as documented by the integrator Fynd, make across-shoulder and chest measurements mandatory for men's topwear and ask for shoulder, chest, front length, waist and hip on ethnic topwear such as kurtas [18]. Those are garment measurements, the same points of measure sitting in the tech pack.

Zalando builds its size advice from information brands supply, alongside purchase and return behaviour and customer fit feedback, and says the advice prevented 8% of size-related returns in 2025 [4]. The platforms are asking brands for data their factories already produce, and the listings that supply it are the ones that survive a size filter and get recommended in the right size.

The Cost of Returns
  • Coresight Research estimates that the average return rate for US online apparel purchases reached 23.4% in 2025.
  • The European Environment Agency attributes 70% of online clothing returns in Europe to perceived poor fit or style.
  • Processing each online return costs a retailer 55% to 75% of the product’s retail price, by British Fashion Council estimates.
  • Return Prime data put Indian clothing and shoe returns as high as 30% to 35%, with size behind almost 40%.
  • In a KPMG survey of 3,000 UK consumers, 13% bought several sizes in one order to find one that fits.
What Platforms Now Demand
  • Google requires a size attribute for clothing and shoes in free listings and in Shopping ads in six countries.
  • Google’s accepted size systems run from AU to US, but an Indian size system is absent.
  • Myntra asks for five garment measurements, including shoulder, chest and hip, on ethnic topwear such as kurtas.
  • Amazon’s Fit Insights tool, which analyses returns data, is open to brands enrolled in Brand Registry in India.
  • Zalando says its size advice, built partly on brand-supplied information, prevented 8% of size-related returns in 2025.

What Production Should Hand to Ecommerce

The fix is a written handover from production to ecommerce, built from files the brand already holds. The usual alternative is new technology, and it carries its own risk. In a study of more than 8,000 shoppers at an online women's apparel store in China, virtual fitting rooms lowered sales among high-BMI customers even as they raised them among low-BMI shoppers [20]. Five things belong in the handover.

Finished garment measurements for every size. Take them from the approved spec, not the first proto. Choose the three to five points of measure a shopper can check against a garment they already own (chest, front length and hip for a kurta, or waist, inseam and rise for trousers) and state whether chest is measured flat or around. A shopper with a tape and a favourite shirt can then compare like with like.

The block the grading was built on. If the base size was drafted on INDIAsize data, a UK block or the brand's own fit model, say which. That record is what lets the ecommerce team map sizes honestly when a listing goes to a market with a different size system.

Then the fit notes, translated into plain words. "Relaxed through the hip, sleeve finishes at the wrist, size down for a closer fit" comes straight from comments a merchandiser already wrote on the fit sample, and it answers the question every shopper has that a chart of numbers leaves open.

Size labels must also match across every channel. The swing tag, the product page, the marketplace listing and the feed sent to Google should all say the same S, M or 38, with size_type and size_system filled in. A store that labels a variant "Medium" in one place and "M" in another splits its own data. A Shopify Google Shopping feed is built from the product's variant options, so on Shopify the fix starts with how variants are named.

The last item runs the other way, and it is the one most brands skip. Return reasons need to reach merchandising by style and size. When returns for one style cluster in one size marked "too small", check the bulk measurements against spec before rewriting the chart. Sometimes the chart is wrong. Sometimes the factory shipped at the edge of tolerance, and the correction belongs in the next production order.

Return rates also vary by style far more than a single average suggests. Researchers working with one large European apparel retailer found online return rates per item ranging from 13% to 96%, against 3% for the same items in its stores [21]. The first test takes an afternoon. Pull the ten styles with the highest return rate, open their approved specs, and lay the finished measurements beside the size chart on each product page. Start the fixes wherever the two disagree by more than the spec tolerance.

About the author. Joshua Uebergang founded Digital Darts in 2015, and it has been a Shopify Partner ever since. Today it is a top-rated Google Ads agency for Shopify brands. His team also builds FeedShine, which creates a Google Shopping feed from a Shopify store in a minute and catches bad barcodes before Google does.

References

  1. FashionUnited partner content sponsored by Alvanon, reporting Coresight Research, "Shifting the Size and Fit Paradigm" (2026). https://fashionunited.com/news/business/sizing-intelligence-is-strategic-priority-as-brands-prepare-for-ai-driven-commerce/2026063073231
  2. European Environment Agency, "The destruction of returned and unsold textiles in Europe's circular economy" (2024). https://www.eea.europa.eu/en/analysis/publications/the-destruction-of-returned-and-unsold-textiles-in-europes-circular-economy
  3. Zalando, corporate site on returns (undated). https://corporate.zalando.com/en/node/10413
  4. Zalando, corporate article on size advice (2026). https://corporate.zalando.com/en/node/11013
  5. Tribune India / ANI, reporting Unicommerce e-commerce trends data (2025). https://www.tribuneindia.com/news/business/fashion-accessories-emerges-as-most-selling-products-on-e-commerce-platform-report
  6. Apparel Resources, reporting Return Prime data (2024). https://apparelresources.com/business-news/retail/third-online-bought-fashion-footwear-products-get-returned-says-report/
  7. Rest of World, on Indian marketplaces charging for returns (2023). https://restofworld.org/2023/end-of-free-returns/
  8. texfash.com, reporting the British Fashion Council's Institute of Positive Fashion, "Solving fashion's product returns" (2023). https://texfash.com/update/state-of-fashion-returns-750000-tonnes-of-co2-emissions-ps7-billion-in-costs-and-counting
  9. KPMG UK, consumer research on delivery and returns (2025). https://kpmg.com/uk/en/media/press-releases/2025/06/research-shows-consumer-demand-for-free-delivery-and-returns.html
  10. BSI, BS EN ISO 8559-2:2025, size designation of clothes. https://knowledge.bsigroup.com/products/size-designation-of-clothes-primary-and-secondary-dimension-indicators-1
  11. EN 13402-3:2017, size designation of clothes. https://shop.standards.ie/en-ie/standards/en-13402-3-2017-324652_saig_cen_cen_748616
  12. NIFT, INDIAsize. https://www.nift.ac.in/indiasize
  13. Google Merchant Center Help, size attribute. https://support.google.com/merchants/answer/6324492
  14. Google Merchant Center Help, size_type attribute. https://support.google.com/merchants/answer/6324497
  15. Google Merchant Center Help, size_system attribute. https://support.google.com/merchants/answer/6324502
  16. About Amazon, on AI in shopping and size recommendations. https://www.aboutamazon.com/news/retail/how-amazon-is-using-ai-to-help-customers-shop
  17. Amazon Seller Central India, "Fit Insights tool now available in Europe and India" (c. 2024). https://sellercentral.amazon.in/seller-forums/discussions/t/5eb9e6fc-4878-400b-a4e9-b3a657813dea
  18. Fynd documentation, Myntra sizing requirements. https://documentation.fynd.com/commerce/docs/openapi/listings-management/myntra/sizing-myntra
  19. texfash.com, "Inside the growing carbon crisis behind fashion's easy return culture". https://texfash.com/special/inside-the-growing-carbon-crisis-behind-fashion-s-easy-return-culture
  20. texfash.com, on virtual fitting rooms and shopper self-esteem. https://texfash.com/update/brands-retailers-have-you-measured-impact-of-your-virtual-fitting-room-on-overall-sales
  21. texfash.com, reporting Dzyabura, El Kihal, Hauser and Ibragimov, "Leveraging the Power of Images in Managing Product Return Rates". https://texfash.com/update/better-predictive-models-could-lead-to-fewer-clothing-returns-finds-study

The tech pack holds finished garment measurements at each point of measure, graded from a base size by a set increment, with tolerances either side. The fit sample session adds comments (tight across the bust, sleeve two centimetres long) and the grade is corrected. Bulk production is then measured against the approved spec, and the buyer's QC report confirms it. By the time a style ships, the brand knows precisely what an M measures.

 
 
 
Dated posted: 13 November 2026 Last modified: 13 November 2026
 
LATEST SPECIALS
Double Jeopardy / Indian Exports / III
Double Jeopardy / Indian Exports / II
Double Jeopardy / Indian Exports / I
Climate Crisis / El Niño Effect / III
Managing Waste / India Project
Climate Crisis / El Niño Effect / I
Double Jeopardy / Indian Exports / II
Climate Crisis / El Niño Effect / III