Shopping for clothes has never been easier: the offer is extremely wide, from the high street and exclusive conceptual boutiques to online retailers. Yet, too often, such a banal experience turns into a tricky and time-consuming adventure with items that look perfectly fine for shape, patterns or colours in the shops or online only to reveal loathsome on your figure. Enter Stitch Fix.
Founded by Katrina Lake in 2011, the San Francisco-based online styling service was created to rescue consumers from such issues and offer a more personal shopping experience.
The site works in a rather friendly and simple way: you fill out your Style Profile and a personal stylist handpicks pieces that fit your tastes, needs and budget and mails them directly to you. Each box contains five items of either men or women's wear (pants, skirts, shorts, dresses, sweaters, shirts and outerwear), shoes and accessories (scarves, jewelry and bags) by several established and up-and-coming labels and brands.
You can try them at home and keep what you like (average price point is $55 per item), then send the rest back in a prepaid USPS envelope. The company only charges you for what you keep, though there is an initial $20 styling fee to pay that will be applied as credit toward any items you purchase.
So where is the highly innovative aspect behind this transaction? Well, it is driven by algorithms that filter customers' criteria: you're size "Small"? The algorithms will eliminate all the other sizes from your choice. You don't like green, but love electric blue? The former shade will be excluded; the latter will become a priority and so on.
A year after it was founded the company got a machine learning algorithm that becomes smarter when it handles more data. Nowadays the company can rely on hundreds of algorithms, matching stylists with clients, exploring a client's Pinterest pins to see what they like and checking if a customer is happy or read comments and feedback about products and styles.
At the moment the offerings are more and more accurate also thanks to the fact that customers who make purchases tell the company what they liked or didn't like. The algorithms are therefore amassing a huge volume of data that couldn't be dealt with by human beings.
Since last year Stitch Fix also started employing algorithms to design new pieces: the company's Hybrid Design project has so far created several garments (that actually make up less than 1% of the company’s inventory) that have been selling rather well, even though they don't look remarkable at all (they mainly consist of basic tops in a variety of patterns and simple yet flattering tunics).
Yet while the algorithms make most of the job, the human aspect hasn't been forgotten: stylists look at recommendations and edit selections, adding that personalising aspect that technology can't give us. The main idea behind Stitch Fix is indeed marrying technology and data with human experience.
Even in the case of Hybrid Designs the data science team is behind the algorithms analysing 30 trillion combinations (one algorithm may pick pieces of clothing that could be used as templates; another may suggests attributes and characteristics and a last one will add some random elements) predicting what kind of garments customers will want to buy and wear. Human designers eventually look at the algorithms' suggestions and work out the final garments.
So far the company has been rather successful (it now boasts 5,700 employees), something that shouldn't just be attributed to algorithms in general, but to the human staff including its team of data scientists and in particular its Chief Analytics Officer, Eric Colson. He previously worked at Netflix where he helped highlighting the films that would be suggested to viewers based on previous viewing choices.
For the time being Stitch Fix is meeting the needs of those unhappy consumers whose needs may not be met anywhere else, but you can bet that other companies and fashion houses will soon follow and start exploring the power and potential of algorithms, jumping on the data systems bandwagon to see how they can target customers.
The interesting thing about Stitch Fix remains the human aspect behind the technology and therefore that balance between human beings, data science and machines. The problem, though, is that there is no algorithm that will educate consumers to maybe buy less and opt for quality, and machines aren't capable of advising us on how to be more adventurous or unique.
Still maybe data science will help specific companies redefining their positions and their collaborations as well: at the moment the focus remains on celebrities and high profile people including bloggers, editors or models considered as powerful influencers who can sell products. Maybe artificial intelligence (A.I.) will help some companies considering new business models and refocusing on real customers, creating for them more personalized experiences.
Comments