AI Powered Product Discovery

As AI advances, vision language models (or VLMs, large models capable of interpreting images and text simultaneously) are nearing the point where they can go beyond CLIP and enable highly accurate tagging without fine tuning. While it might seem absurd to run every item in inventory through a prompt, this is now becoming possible even on highly performant models, such as Gemini Flash 2.0, which costs only $0.075 per million input tokens when run in batch mode, or about $1 per 10k images.

Showing the right items to the right users at the right time — also known as Product Discovery — means leveraging your unique data to show users personalized content that aligns with your brand. Through talking with dozens of teams, and working with design partners, we saw that the common denominator of great personalization strategies was a flywheel where great product metadata flowed into great experiences across search, recommender systems, marketing, and SEO.

A/B tests are a powerful statistical tool, commonly used (or abused) for making decisions about everything from button colors to machine learning models. But, we’ve seen A/B tests frequently used incorrectly. This blog post describes a light framework for planning out A/B tests, influenced by best practices at places like Square and Stitch Fix.