These two people are the same woman. OK, not really. But they share the same ‘fit profile': similar height, weight, age, and body shape; even the same preference for slightly loose-fitting clothes. We can use Sara's purchase history to recommend a size for Nina.
Nina enters her height, weight and age, and identifies her body shape...
Nina gets a recommendation for the size purchased by shoppers similar to her - shoppers like Sara, who bought this same shirt in size M and did not return it for size.
It's 30 days later. Did Nina keep the shirt, or return it for a different size? Either way, our algorithm learns from the customer feedback and uses Nina's experience to recommend the best possible size the next time around.