>> Variant B improved conversion rate (CVR) or Add-to-Cart (ATC) rate? Roll it out.
Basing rollouts on add-to-cart rate is generally not a great idea, particularly for experiments on product pages. There are a variety of reasons why shoppers add products to carts that may only be loosely based on intent to purchase. In my years of ecommerce A/B testing, I've seen plenty of tests that improve conversion rate and/or revenue per visitor but negatively impact add-to-cart rates.
For doubling success I suggest ruining the product slowly and testing how crappy the product can still be while the customer is still paying for the brand (as he's expecting some improvement for the higher price), and voila, we've got enshittification.
