Fashion shopping app The Yes is branching out of the smartphone space to the wider web.
At Fairchild’s Tech Forum on Wednesday, chief executive officer and cofounder Julie Bornstein revealed that the company is launching a website driven by the same data science that fuels its popular mobile app.
The start-up launched in May 2020 with an iPhone app that put different looks in front of users and asked yes/no questions to understand their preferences. The answers fueled data models that homed in on their tastes and style, and with 100,000 downloads since then, the flow of information has been gushing in. According to the company, The Yes has clocked more than eight million yes/no clicks within the app so far, and it’s growing 40 percent month over month.
Yet the company still realizes that sticking with an iPhone app alone is a somewhat limited proposition.
“I think the first year was really focused around putting a product out there. And we started in mobile, because not only is it the fastest-growing part, but it’s also a more restrictive format,” she said during Wednesday’s “The View From Next-Gen Fashion and Retail” panel. “And so it required us to have to make tough choices. But ultimately, if you want to reach everyone, you need to be on the web.”
First Look at Dior and Sacai’s Capsule Collection
Today, anyone can launch a webpage or e-commerce site with static listings. But that was never going to do for The Yes. Instead, it touts this launch as the first e-commerce site powered by a neural network — a data-driven approach that trains machines to perform tasks based on data models. Put another way, The Yes created a shopping site that remembers customers and all of their individual preferences every time they come back, so it can actively adapt to each users’ tastes.
Like the app, shoppers can answer the simple yes-no questions on individual products — which, it turns out, users find fun and even addicting, according to App Store reviews. There’s also a quiz about style, price and size preferences, for a broader look at their overall tastes, both of which inform the new site’s smart search feature.
A look at The Yes’ new neural network-powered e-commerce site for fashion. Courtesy image
Bornstein elaborated on the data science in exclusive comments to WWD: “We have rebuilt the architecture of e-commerce to integrate an AI layer into the core technology. This allows us to create a one-to-one neural network with each shopper,” she explained.
The more a mobile user or website visitor interacts and provides information, whether through the yes/no interactions or the style quiz, the better the company learns her preferences. “This is the first e-commerce system to adapt based on each shopper’s explicit signals,” Bornstein continued. “The result is that each shopper’s ‘store’ adapts to her. Whether she is on the homepage, browsing or in search, she will see products ranked for her style, brand, price and color preferences, preselected in her size.
“The effect of this is clearing away all the irrelevant product and helping her find what she loves across hundreds of brands in a whole new way,” she added.
In other words, no two shoppers see the same feed of products.
According to The Yes, the company’s data-driven approach to machine learning means that no two feeds are like. Courtesy image
The Yes calls it “adaptive shopping e-commerce.” But digital natives will find the formula familiar, as it’s similar to the way streaming services like Pandora or Netflix queue up songs or films that they think the user might like next.
The concept isn’t entirely new — eBay launched its “Store of You” in 2018 with a similar premise — making the effort look particularly useful for massive marketplaces that can overwhelm shoppers with so many choices. The e-commerce titan wrangles a high volume of listings that’s quickly approaching two billion, for instance.
The Yes features nearly 100,000 styles, which may seem small by comparison. But it’s only one year in, and the selection is focused specifically on fashion, with a broad spectrum of more than 250 brands, from Mango and Everlane to Miu Miu and Balenciaga, in sizes from 00 to 40.
Its fundamental premise is to help fashion shoppers navigate all the choices, and perhaps home in on just what their individual style is, even if they don’t quite know themselves.
In both cases, the data science hinges on the work of building taxonomies from the ground up. For eBay, a platform that traverses so many categories, it had to develop a so-called “Interest Taxonomy” that helps distinguish why shoppers were flocking to certain types of products, based on their passions. For The Yes, its “Fashion Taxonomy” covers the nuances of style across apparel, footwear and related products.
“We create a lot of our underlying artificial intelligence models through intelligent labeling of the data,” co-founder and chief technology officer Amit Aggarwal previously explained to WWD. Today, every product in the catalogue is automatically tagged via a machine learning model with more than 500 detailed product attributes.
The result is a system that can distinguish between two skirts, dresses or other garments with very similar details, but different styles.
As one might expect, engineers comprise more than half of the 40-person crew at The Yes. They’ve been busy, building a proprietary fit map that normalizes size across all brands — which offers massive efficiency for making size recommendations or product recommendations based on size profiles. To promote brand discovery, they also built a custom matrix that leans on a customer’s shopping preferences and behaviors, as well as the expertise of a particular expert: Fashion and creative director Taylor Tomasi Hill, who previously hailed from Moda Operandi.
The combination of fashion and tech expertise has attracted $31 million in funding to date, along with a few other impressive numbers: According to its data, the platform clocks conversion rates of over 6 percent, with repeat customer rates topping 50 percent. Sixty percent of its shoppers are under the age of 30, and its best customers use the app more than seven times per month and have bought more than 20 items.
So far, the customer base looks like an even split of young shoppers who love to share the app and luxury consumers who appreciate the ability to buy high and low in one cart, at 50 percent each. Additionally, The Yes claims that 90 percent call the app one of their top shopping destinations.
There’s no doubt that mobile has been good to the company, so naturally, it still considers itself mobile-first and recommends using the iPhone app, if that’s an option. The difference now is that it’s not a limiting factor. The data-driven startup can now embrace any online shopper, no matter what tech they use or what style they want.
Source: Read Full Article