Why algorithms are working hard to disrupt the fashion world


Style update

For the first time I wore a suit and high heels to attend a meeting of an American technology company, and I felt as if I accidentally put on makeup. Comfortable wool and comfortable sneakers are the default uniforms in Northern California. Next time I wear jeans.

But just because San Francisco people don’t dress up doesn’t mean that the tech industry is not interested in fashion. Mark Zuckerberg’s grey Brunello Cucinelli T-shirt and the cashmere sweater worn by venture capitalists are unique if it is a standard fashion choice. San Francisco even produced its own global fashion brand in Allbirds wool sneakers.

However, what the industry has not done is trying to subvert the fashion industry itself.E-commerce has put a rocket booster Under fast fashion Sales and financial technology have enabled fast online payments, and social media has provided targeted advertising. But most of the business of choosing and designing clothing escaped the control of the technology industry.

This is not because you don’t want to try.Virtual reality glasses manufacturer Magic Leap promises to completely change the way we try on clothes before collapse Under the weight of its own hype. Google and e-commerce company Zalando collaborated on artificial intelligence-driven design through Project Muze, but the results were not encouraging. According to the tech blog, its proposed design is “same” small tools. Personalized, artificial intelligence-driven fashion sales are still largely a pipe dream.

The company that applies data-driven decision-making to the difficulty of subjective human taste is Stitch Fix, an algorithm-based mail-order clothing company. Established in the shopping district of San Francisco ten years ago, it was created when delivery boxes were all the rage. Vegetables, juices, socks and razors are all sold through subscription e-commerce.

Stitch Fix stands out for its emphasis on data. It appointed Eric Colson, a former Netflix data scientist, as its chief algorithm officer in 2012, making him the first person to hold such a title. The company is equipped with buzzwords such as network effects and proprietary algorithms. It promises to bring science to the art of fashion.

Fears that Amazon may suppress this business model means that stock prices are off to an unstable start. But for Stitch Fix, Amazon has been working hard to sell fashion products. Think of its artificial intelligence camera Echo Look. It was supposed to provide style suggestions, but it was criticized for repeatedly making mistakes in clothes.Even newer StyleSnap search tool, Which allows users to upload photos and recommend similar clothes, was accused of omitting the tag Science and Technology PressThe number of choices can make the site feel like a messy sale.

Amazon’s failure coupled with the online shopping boom in the pandemic era very good Used for suture repair. Reluctant to go to the store and keen to wear comfortable clothes when locked, more and more users turn to the site. In the three months ending in May, sales increased 44% from the previous year. It is experimenting to allow users to purchase goods directly.

As it develops, Stitch Fix seems to be keen on emphasizing the human factor in the business, just like the magic of algorithms. In August, it appointed a new CEO at Elizabeth Spaulding, who Announce The stylist “played a very active role in training our machine learning models with our data science team”.Please note that in the 2017 listing documents, the word algorithm is highlighted 76 timesIt was mentioned only once in a conference call with investors this summer. The number of human stylists has kept pace with user growth-it has doubled since 2017.

The trouble with collecting multiple data points is knowing how to deal with them. In the deepest blockade, I ordered a Stitch Fix box and filled out a lengthy online style quiz about my favorite store and style. Paying a $20 styling fee, keeping the clothes I like and sending the rest back seems to be an effective way to shop. But what was sent was a bunch of disappointing strange clothes, mostly from brands I had never heard of. I want Kate Moss in the office. I got a little royal at the garden party.

Taste is abstract and difficult to determine-no matter how much data you have. The image of the flowered shirt in my eyes is different from the shirt you imagine. It will also not match the first result displayed by Amazon, Google or Stitch Fix.

Stitch Fix says that the more you order, the better the stylists and algorithms will know about you. Who knows, maybe the second or third box would be more suitable. Again, there may be things that algorithms cannot do.

[email protected]


Source link