Stitch’s Artificial Intelligence Project: Shop Your Looks

Online retailers have long lured customers. They attracted them with the ability to browse a huge selection of merchandise from home, have goods conveniently delivered to their doorstep, and quickly compare offers and prices. Nevertheless, much of the in-person shopping experience has been lost, not least with how colors work with your complexion, trying on clothes to see how they fit, and so on. Companies like Trunk Club, Stitch Fix, Wantable have attempted to address this problem. They have been hiring professionals to choose clothes based on their custom parameters and ship them out for customers. Customers can try things on, keep what they like, and send back what they do not. Fixes is Stitch Fix’s version of that service.

Customers get a personalized Style Card with some inspiration for an outfit. It is driven algorithmically. Moreover, it helps human-style experts to match a garment with a shopper. Each Fix includes a Style Card that shows clothing options to complete outfits based on the various items in a cutromer’s Fix. Last year, the company started testing a way for shoppers to buy those related items directly from Stitch Fix. This was through a program called Shop Your Looks, and they ended up using it due to popular demand.

Artificial intelligence is a natural for for such services. Moreover, Stitch Fax embraced the technology to improve and accelerate Shop Your Looks. That puts the company in direct competition with behemoths like Google, Facebook, and Amazon, on the tech front. All those companies are aggressively building out artificial intelligence-powered clothes shopping experiences as well.

Shop Your Looks

Stitch Fix told VentureBeat that they had a Shop Your Looks beta period. Within that period, more than one-third of clients who bought with Shop Your Looks engaged with the feature multiple times. Moreover, 60% of clients who purchased with the offering bought two items or more. It has been successful enough that the company recently expanded to include an entire shoppable collection using the same underlying technology for personalizing things and outfit recommendations as customer shops. Natalia Gardiol and Hilary Parker are Stitch Fix data scientists. They had an email interview with a VentureBeat. Here, they explained what drove them to develop Shop Your Looks, how the team used artificial intelligence to build it out, and the methods they used, like factorization machines.

In the case study, they had:

• Their problem – How to expand the scope of its service to match outfits to online customers using a mix of human expertise and algorithms.
• The result – Shop Your Looks.
• It was an experiment by a small team of data scientists of Stitch Fix. Then they expanded around other units within the company.
• The biggest challenge – to determine what is a “good” outfit when context matters and the taste is so subjective.
• Stitch Fix used a combination of human-crafted rules to manipulate, store, and sort data. It also used artificial intelligence models called factorization machines.

They had to evolve their algorithm capabilities from matching a client with an individual item in a Fix to now matching an entire outfit based on a client’s past preferences and purchases, to create Shop Your Looks.