Many high-end celebrities and fashion models use personal shopping assistants or stylists to help them find the best clothes for them.
For many women, however, that’s not really an option. We often rely on friends or family to give a thumbs up or thumbs down on potential clothing purchases, and that’s about the extent of it.
That may be about to change.
Lily is a virtual, AI shopping assistant that learns about your style and how you feel in your clothes.
The app just won the 2017 SXSW Accelerator Pitch Event in the Social and Culture category. Here’s why it’s so much better than every other shopping app out there.
What Is Lily?
Put simply, Lily is a personalized mobile shopping app that helps shoppers discover clothes that are right for them. For now, the app is only available on iOS (for free), but its creators Purva Gupta and Sowmiya Chocka Narayanan promise an Android version is on the way.
Lily’s stated goal is to help women discover and buy clothes that make them look and feel the best. Created by an all-women team of six industry professionals with backgrounds ranging from Facebook to Macy’s, Lily’s development comes from the very audience it also seeks to help.
“Lily is built by women, for empowering women to be the best version of themselves,” Gupta told me via email. Gupta says she and the rest of the Lily team have learned that there are 120 million U.S. women who are not happy with their appearance, and that women experience about 13 negative thoughts about their bodies every day. “As a team, we are on a mission to change these numbers and solve this problem in the world,” she says.
There are plenty of shopping apps on the market, but what makes Lily stand out is its technology. It doesn’t simply use your purchase history to give you bland or generic recommendations. It goes much deeper, creating an experience meant to feel more like the app is truly getting to know you.
But how does the app “know” what each woman wants in her clothing? That’s where some unique technology comes into play.
How Lily works
The Lily app is able to make suggestions and recommendations based on users’ perceptions and emotions about their own body.
It accomplishes this feat in a fun way. The app asks users a series of questions about body type and style preferences, almost as if you were chatting or texting a friend. One example of a question Lily asks: “How would you describe your décolleté (that’s French for chest — I’m very worldly!)?”
While on the surface it may feel like you’re chatting with some sort of automated quipster, but Lily is hard at work using your answers to learn more about what kinds of clothes make you feel best. The app asks how users feel about their body — what parts you like to accentuate and which ones you’d prefer to minimize — and uses a complex matching algorithm to make recommendations.
Lily’s creators call this algorithm the “Perception and Empathy engine,” the first such engine of its kind.
“I have personally spent more than 10,000 hours in the last three years asking women about what they feel when they are buying clothes online or in stores and why they buy the clothes they buy,” Gupta said when asked about from where the idea for the engine came. “We quickly understood that most clothing recommendation engines are focusing on what users like and buy.”
Gupta says this information is obtained through “millions of tangible actions” like browsing and past purchase history.
Through their thousands of conversations with women over recent years, the team at Lily learned that a true personalization engine needed to get beyond the what to the root of why customers buy the clothes they do.
Instead of focusing on tangible customer actions and rational behaviors, Gupta says Lily’s Perception and Empathy engine considers the intangible perceptions and irrational behaviors that drive consumers’ purchasing decisions.
“We also found in our research that women are buying more clothes every now and then to feel their best in the situation they are in or preparing for. It’s all about satisfying the feelings,” Gupta says.
The app then connects the patterns of responses and generates suggestions that aim to please users’ emotional needs and style desires. It then gives you the option to buy clothing online, reserve items in physical stores or take it along with you on a real-world shopping trip.
As it stands, Lily has a number of well-known retailers on board already, including H&M, Express, Lulus, Macy’s, Bloomingdales, Nordstrom and Banana Republic. Gupta reports that one in three Lily users have already made purchases from their favorite brands through the Lily app.
Why Lily works
One of the reasons Lily is poised for success is because it’s a win-win proposition. Put another way, the app is valuable both for its users as well as for the retailers that are on board.
For users, the app gives them unique suggestions and recommendations based not just on purchase history, but also on emotional questions about style and body type. This can give it a more personal feel and help you select clothes that are right for you.
After all, the app “knows” millions of fashion rules/hacks that help it select clothing to flatter various parts of the body, based on your preferences or desires. It also learns as you use it, in an effort to understand your unique body perceptions and style preferences. It then prioritizes that information accordingly.
The app is also a win for retailers, for a couple of reasons.
First, it’s a good way to generate business. These days, retailers are looking to stay competitive with online giants such as Amazon. In fact, Amazon sales account for over 60.5 percent of online sales growth, which has led others to keep up by utilizing third-party fulfillment centers.
Second, Lily attempts to provide users with very personal suggestions, which can lead to a very positive overall shopping experience. This helps paint retailers in a good light in the minds of Lily users, increasing the chances they will becomes return customers.
Why Lily won at SXSW
In short, Lily was honored at SXSW because it takes technology to the next level.
We know our electronic devices can “learn” our preferences, histories, and more over time, but very few applications make it feel this personal. What other app asks you which color you want to avoid more than kale fries?
On the business side, Lily already has a strong foundation of retailers on board. This helps in two ways. One, it gives the app instant credibility because it includes retailers that just about everybody has heard of. Also, it includes retailers that vary in price point and fashion sense, meaning there’s plenty of variety to go around.
And, assuming the rollout of Lily is as successful as it seems it will be, you can bet on even more of your favorite retailers to jump on board in the coming months.
And what does Gupta like most about Lily?
“My favorite thing about Lily is how she explains every item of clothing and how it flatters my body,” Gupta said. “We have built that feature ingesting more than 50 million data points and I’m very excited about how this product feature has shaped up in the Lily experience.
“I hear often from our users that they are getting spoilt by Lily as they now look for Lily’s personalized recommendation logic if they see any item of clothing elsewhere online or in-store.”