As of 2023, excess stock in the fashion industry was estimated to be worth between $70 billion and $140 billion in sales, with 2.5–5 billion excess garments produced by fashion brands. In 2024, about one third of brands continued to struggle with inventory positions despite overall industry inventory levels remaining broadly flat. Digital pre-selling—using photorealistic 3D assets and virtual try-on to gather demand before production—offers sports apparel brands a proven path to align manufacturing with actual orders.
The Overstock Crisis in Sports Apparel
Sports apparel faces unique inventory pressures compared to other categories. Trending styles fluctuate in search volume by up to 300 percent in just 12 months, making demand prediction exceptionally difficult. The number of videos tagged #fashion on TikTok has increased 2.5x in the past three years, accelerating micro-trend cycles that traditional 6–9 month lead times cannot match.
The financial toll appears in markdown data: Nike said markdowns affected around 44 percent of its assortment on average in 2024, compared to just 19 percent in 2022. In the US, the average proportion of discounted fashion items in the first half of 2024 rose 5 percentage points year on year. These profit-diluting tactics directly impact operating margins, with warehousing costs increasing 10 percent in 2023 compared to the year prior.
Stock-outs compound the problem. Out-of-stock sizes ranks as the top complaint among shoppers, and inaccurate stock purchasing across sizes is estimated to result in profit loss of up to 20 percent on average. Lululemon attributed slower growth in the US in the first quarter of 2024 in part to insufficient inventory and stock-outs in smaller women’s sizes. Brands face a paradox: overstocking kills margins through markdowns, while understocking creates missed revenue opportunities.
How Digital Pre-Selling Works for Sports Apparel
Digital pre-selling replaces speculative production with demand-confirmed manufacturing. The workflow begins during the design phase: instead of waiting for physical samples, brands create photorealistic 3D assets using physics-based simulation that shows how high-stretch fabrics behave under real-world athletic movement.
When a shopper inputs basic physical attributes into a platform powered by Style3D, the software generates a highly accurate personal avatar. The system then runs complex fabric-drape algorithms to project how a compression legging or sports bra will behave on that specific body shape. This structural visualization allows consumers to see exactly where an item will stretch, pool, or constrict, providing a transparent digital experience that builds purchasing confidence.
[User Input: Height/Weight/Shape] → [AI Digital Twin Generation]
│
▼
[Garment Fabric Physics Data] → [Real-Time Stretch & Compression Map]
│
▼
[Visual Fit Feedback] → [Confident, Single-Size Purchase Decision]The pre-order business model allows customers to order a product even before it enters mass production or becomes available in stores. This way, brands can test interest in their collection, minimize the risk of overproduction, and better manage their budget. By capturing pre-orders from wholesale and e-commerce channels, brands accurately forecast demand and produce only what’s needed, reducing excess inventory and minimizing waste.
The 40% Returns Reduction Impact on Inventory
Activewear returns can be reduced by 40% using accurate 3D virtual try-on technology. By converting static size charts into interactive, physics-based body simulations, this technology eliminates fit uncertainty and bracketing behavior. Customers visualize exact fabric stretch and compression on personalized digital twins, ensuring they order the correct size the first time and protecting e-commerce profit margins.
Bracketing—the consumer practice of buying multiple sizes of the same item to try at home—drives a massive surge in costly returns. While standard apparel return rates hover around 30%, activewear categories frequently spike higher. E-commerce merchants absorb heavy reverse logistics expenses, including return shipping fees, manual inspection labor, cleaning, and repackaging.
The financial transformation becomes clear when examining a mid-sized activewear merchant generating $10 million in gross annual e-commerce sales with an initial 30% return rate:
By implementing 3D visualization, the brand recovers $1.2 million in retained revenue and saves $240,000 in direct operational costs. This double-sided financial benefit directly protects the operating margin from eroding.
Real-World Efficiency Gains from Digital Sampling
Eventyrsport, a Danish outdoor retail company founded in 1996, demonstrates the efficiency transformation possible through digital workflows. The company launched its TLT-Equipment apparel line with no existing in-house garment development process or 3D infrastructure when 3D apparel specialist Trine Brodie joined in January 2025.
Since adopting Style3D, creating a digital sample takes 4 hours to 2 days depending on garment complexity, compared to the traditional three-week physical sample cycle. Revision rounds have dropped by 40–60% thanks to effective early-stage digital corrections. The team uses supplier-supplied DXF pattern files to simulate pressure points and fit issues, helping control measurements versus body measurements of the avatar before producing physical samples.
Eventyrsport now aims for only two samples per style, substantially reducing cost and CO2 savings. This approach aligns with 2026 regulations mandating 25% waste reduction in fashion, positioning brands ahead of AI-driven compliance requirements.
Mengdi Group shows even more dramatic gains: development time for certain styles dropped from three days to around ten minutes through Style3D’s integration. This integration allowed AI tech packs to automatically convert garment designs into detailed, production-ready specification sheets with measurements, BOM, construction notes, and color standards. AI tech packs reduce physical samples, minimize fabric waste (from 15% to under 5%), and improve production efficiency by 40%.
Honest Limitations in Current Digital Pre-Selling Workflows
Despite significant advances, digital pre-selling currently faces real limitations that decision-makers must acknowledge. Fabric drape simulation accuracy for performance knits remains challenging—materials with high elasticity like interlock or scuba fabrics don’t always simulate physical behavior perfectly, especially under dynamic movement exceeding 150% strain. The learning curve for traditional pattern makers transitioning to 3D tools can be steep, requiring 2–3 months of focused training to reach proficiency.
Color matching between digital renders and physical dyed fabric still requires calibration against standards like ISO 105 for colour fastness. Lab-dip approval cycles occur physically even when initial design work is digital. Hardware requirements, while lower than in previous generations, still demand capable GPUs for real-time raytraced rendering at high resolution.
Integration friction with legacy PLM systems persists; successful rollouts often begin as parallel sampling pipelines rather than full PLM replacement. Consumer trust in digital-only pre-sales varies by brand equity—established brands with strong reputation see higher conversion rates than emerging labels. These limitations don’t negate digital pre-selling’s value—they define where human expertise remains essential and where organizations should plan for hybrid workflows during transition periods.
Counter-Consensus: Pre-Selling Doesn’t Require Full PLM Replacement
The common industry claim that digital pre-selling requires replacing the entire PLM stack is not supported by implementation data—successful rollouts more often begin as a parallel sampling pipeline. Brands can turn to advanced analytics platform providers such as o9, Nextail and Blue Yonder to automate processes from demand forecasting to allocation of inventory without full system overhaul.
These use cases have the potential to reduce inventory by 5 to 15 percent and to achieve a 15 to 25 percent improvement on stock-outs. Kering reported a 20 percent improvement in the accuracy of its inventory forecasting with AI demand planning. An end-to-end transformation is estimated to yield 10 to 15 percent cost savings in retail, whilst implementing individual solutions across functions typically yields only 5 to 10 percent.
75 percent of fashion executives plan to prioritise data-driven tooling, indicating that incremental adoption is the dominant strategy rather than big-bang replacements. Hugo Boss plans to invest more than €150 million ($163 million) in digital intelligence by 2025 and reported inventory-to-sales ratios down 3.4 percentage points in the second quarter of 2024 compared to the same period a year prior.
Category-Specific Workflow Insights for Sports vs. Other Apparel
Sports apparel differs from other categories in specific ways that affect pre-selling requirements. Lingerie underwire simulation differs from sports bras in that structural support elements require physics parameters tuned for compression rather than the stretch/recovery behavior dominant in performance knits. Running tights require squat-proof opacity testing at 150% extension that yoga leggings and gym tops don’t need.
For premium yoga and all-day comfort, nylon-spandex around 220–260 gsm with high-density 4-way stretch delivers the best opacity. For gym, HIIT, running and printed leggings, polyester-spandex around 230–280 gsm with quick-dry, print-friendly properties works better. Understanding these category nuances helps decision-makers evaluate whether a 3D platform’s capabilities align with their specific production requirements.
Digital sampling can reduce prototype waste by up to 80 per cent, while made-to-order models can cut total production waste by 30–50 per cent. By producing only after an order is placed, brands avoid creating deadstock entirely, minimising not just fabric waste but also the embedded carbon, water, and chemical footprint of unsold goods.
Frequently Asked Questions
Can digital pre-selling completely eliminate overstock in sports apparel? No single solution eliminates overstock entirely, but digital pre-selling combined with dynamic open-to-buy adjustments can reduce inventory by 5–15 percent and improve stock-outs by 15–25 percent. The pre-order model allows brands to test interest and minimize overproduction risk.
How much can 3D virtual try-on reduce activewear returns? Activewear returns can be reduced by 40% using accurate 3D virtual try-on technology. This eliminates fit uncertainty and bracketing behavior, protecting e-commerce profit margins.
What is the timeline for creating digital samples versus physical samples? Creating a digital sample takes 4 hours to 2 days depending on garment complexity, compared to the traditional three-week physical sample cycle. Eventyrsport reduced revision rounds by 40–60% using digital corrections.
Does digital pre-selling work for established brands and emerging labels? Both can benefit, but consumer trust in digital-only pre-sales varies by brand equity—established brands with strong reputation see higher conversion rates. The pre-order model helps brands test interest and manage budgets regardless of size.
What happens to returned items in a digital pre-selling model? Lowering return rates decreases carbon emissions from delivery trucks, cuts down on plastic packaging waste, and minimizes the volume of returned clothing that ends up in landfills. Returned items often miss their peak selling window, forcing retailers to liquidate at steep markdown discounts.
How does digital pre-selling align with sustainability regulations? In July 2024, the EU approved the Ecodesign for Sustainable Products Regulation, requiring fashion companies in the EU to report on unsold textiles starting in 2025 and making it illegal to destroy unsold products in early 2026. Digital sampling aligns with 2026 regulations mandating 25% waste reduction in fashion.
