Personalization for fashion, tell me about it

If you have an angel’s face and devil’s body, then every piece of clothing should be looking okay on you; and this species of human is called “model”. But for the mass of ordinary people, we face problems like short legs, big thigh, flappy arms, flat chest or too busty…which means probably only certain types of clothing would match your figure and personal taste. These often require we have a sense of style to enhance our good body features while disguising unsatisfied parts, which is often easier said than done.

Some might talk about “brand loyalty”; once you find a brand fit your style, then you stick to it for the rest of your life, but this also puts you in the lost opportunity of diversity and discovery.

In the meantime, e-commerce shopping for apparel seems to provide unlimited choices along with unlimited struggles. For example, just type “Little Black Dress” on Alibaba’s Tmall, the generated search results of over 12000 items would even make Coco Chanel (the originator of this “must have” piece in every woman’s wardrobe”) head spin, as basic filter functions do not necessarily narrow down your options.

Whispers of AI and big data have been circulating around to solve these issues. A company like Stitch Fix stood out of crowd and even went IPO last year after only 6 years of launch. It rejuvenates the old concept of subscription e-commerce with a box of clothing sent to your door, but instead of a blind box with random selection, this box now can think and act like your personal stylist, leveraging a geeky algorithm based on each customer survey upon signup and purchase behavior after each shipment. In the end, the company still adds a human layer to finalize the robot’s styling thoughts. That is why the company’s team is made up of 70% of data scientists and 30% of the professional stylists. Its solid success has inspired a string of imitators in China, for instance:

-exact copycat like TheLook, BoxFantasia

-Copycat for men’s wear Champzee 

-Copycat for kids wear Xiaolunsenlin

The problem remains how personalized and efficient these AI box e-commerce services can achieve? As local users would complain that quite often they still do not like the items recommended by the box and have to return them.

Another new trend among Chinese millennials is they no longer buy apparel; they rent. China-based Yi23, a female clothing rental platform, has raised a USD 50 million Series C round of funding.

The apparel sharing business also faces the bottleneck of personalization for better customer experience. An insider source told Yi23 is now collaborating with a local AI fashion analytics startup to understand the pattern&insights of its dataset and hopefully to generate a more personalized recommendation for each user. Both startups are invested by one of the nation’s most reputable VC firms which has just made tons of money out of China’s sharing economy frenzy.

Again AI-powered personalization is the new hit fashion, however, there is still a long way to go.

Author: Cecilia Wu