Three converging forces make Fioobra possible
The most revolutionary thing in textile recommerce isn't Gen Z buyers, it's agentic commerce. While consumer demand for secondhand is growing (14% US market growth in 2024, 5× faster than traditional retail), the real unlock is on the supply side: technology can now autonomously do what previously required manual human effort for every single garment.
Force 1: The AI capability explosion (technology shift)
Mainstream multimodal AI arrived in late 2023. GPT-4V became publicly available in September 2023 (API access November 2023). The first mainstream commercial model capable of analyzing a garment image and generating a structured description in a single inference call. Before this, automated classification required separate computer vision and NLP pipelines, affordable only for players like ThredUp.
Inference costs collapsed 280-fold between November 2022 and October 2024. Post-January 2024 models driven by DeepSeek and similar efficiency breakthroughs continue to halve costs approximately every 4–5 months. Processing 10,000 garments daily through AI classification would have cost thousands of dollars in 2022; today it's pennies per item.
Agentic AI enables autonomous multi-step workflows. The duration of tasks LLMs can reliably complete has doubled every seven months since 2019 (METR research, March 2025). Post-2023 data shows acceleration to a ~4.3-month doubling period. McKinsey projects AI agents could mediate $3–5 trillion of global consumer commerce by 2030. Salesforce reported that AI agents already influenced 17% ($13.5B) of U.S. holiday orders in 2025.
Computer vision for garments has reached production quality. YOLOv5 models achieve ~98% accuracy on garment type detection. Vision Transformers, CLIP-based zero-shot classification, and diffusion models have all matured between 2022 and 2025.
Force 2: Regulatory forcing function (market shift)
A wave of EU legislation is making recommerce mandatory, not optional:
France pioneered textile EPR in 2007. Now all 27 EU member states must implement EPR. The economic calculus has flipped: not having a resale/recommerce operation becomes a cost center, not just a missed opportunity. The EU generates 12.6 million tonnes of textile waste annually. That’s 12 kg per person. Currently 73% of discarded textiles in the EU are landfilled or incinerated. Less than 1% are recycled into new garments.
Force 3: The secondhand market reached critical mass (demand shift)
The global secondhand apparel market reached $256 billion in 2025 and is projected to hit $367 billion by 2029 at ~10% CAGR, growing ~3× faster globally (up to 5× in the US) than traditional apparel retail (ThredUp/GlobalData Resale Report 2025). Market Data Forecast estimates European secondhand clothing specifically at $32.1B (2025), reaching $75.6B by 2034 at 10% CAGR. Vinted surpassed €10 billion in GMV in 2025 and became the largest clothing retailer by volume in France.
Brand-operated resale programs grew ~244% in 2022 alone, reaching 163 brands with resale shops by 2023. 94% of retail executives say their customers already participate in resale, yet 86% of those same executives don't know how to make resale work for their brand (ThredUp 2025). Zara Pre-Owned expanded to 17+ countries. H&M's Sellpy generated ~€104M revenue in 2023, growing to ~€160M in 2024.
Why now and not three years ago
In 2022, we couldn't build Fioobra. Multimodal AI didn't exist at commercial quality. Inference costs would have made per-item processing prohibitively expensive. Agentic workflows were unreliable. EU textile EPR was still in committee. Today, all three forces have converged: the AI is capable, cheap, autonomous, and improving. The regulations are forcing brands into recommerce. And the market demand has reached critical mass.