The real challenge was the UX: building a zero-friction experience for shoppers to maximize conversion. I’ve run a some of A/B tests on the widget design and copy, and learned not to trust my intuition, multiple times, the "best" variant wasn't the one I expected.
It's hard to gather strict numbers without large-scale data yet, but early estimates show that customers who engage with the try-on are ~25% more likely to buy.
Live demo here: https://demo.genlook.app