Case Study: Wayfair achieves 99% category win rate and 7-point clickthrough lift with Snorkel AI

How Wayfair accelerated product tagging automation with Snorkel Flow

Wayfair, the home goods and furniture e-commerce retailer, needed a better way to manage product tagging across its 40M+ catalog. Inconsistent supplier data, noisy labels, and slow manual labeling were hurting search performance and making it difficult to capture important design nuances in product images. Wayfair partnered with Snorkel AI to improve its data-centric AI workflow for catalog tagging and visual information extraction.

Snorkel AI co-developed a programmatic labeling and computer vision solution with Wayfair, using weak supervision, iterative data curation, and model-guided error analysis to build better training datasets and custom tag models faster. The initiative helped Wayfair create 46 tag models in days instead of months, achieve a 98.97% win rate over previous baselines, and deliver measurable business gains including a 7-point lift in clickthroughs, a 5-point increase in add-to-cart rates, and 10x faster model development cycles.

Wayfair