Machine learning is just summarizing data.
Ten Myths About Machine Learning
Pedro Domingos
1.1K21

Solving the paradox of choice with iPhoneX

The iphoneX projects a precise mesh over the face of one of our software engineers.

At Warby Parker, we design and sell hundreds of eyeglasses and sunglasses to fit a dynamic and diverse customer base. The amount of choice we provide is tremendous, but how does a customer narrow what might seem like an overwhelming assortment? Where do you start?

Barry Schwartz, a behavioral economist from the Wharton school, describes this phenomenon in his landmark book, The Paradox of Choice. In his famous salad-dressing study he found that sales of some items decreased as the number of options exploded, because we are confounded by the abundance of choice. Apple learned from this by bundling computer options into good, better, and best. New computer — easy peasy.

Selecting the right frames is partially a style choice, but it also has to do with what physically fits your face. And fit is complex. Here are some factors that Warby Parker retail advisors consider when helping people find their perfect frame:

Pupillary Distance: Your pupils should sit in the center of the lenses.

Lens height: The bottom of the glasses shouldn’t contact your smiling cheeks.

Eyebrows: Frames shouldn’t be so tall that they mask (or go above) those expressive brows.

Frame size: If your prescription is strong, avoid large frames (lenses tend to be thicker).

How do we take some of the expertise and advice from the retail experience and embed it in the ecommerce app? It would require a very sensitive 3D camera which would be too expensive for a retail store and certainly not within the reach of most consumers.

But last week the world changed. The new Apple iPhoneX introduced a technology that the Warby Parker Vision Tech and Research teams have been waiting for. The phone projects a grid of 30,000 points of infrared (IR) light, while a dedicated IR camera and neural network learn a depth map of your face. Apple calls the system TrueDepth.

The features of your face are totally unique, they provide a key to unlock your phone — just like a thumb print, except better, because it works with mittens on.

The same technology can map key facial features and alleviate the paradox of choice shopping problem. Using the TrueDepth camera, Warby Parker can now recommend a subset of glasses that are the most likely to fit your face.

How to use the tool:

If you have an iPhoneX, first download Warby Parker’s app. As you scroll through our frame gallery, you’ll see a new button that offers an option to filter your face.

In less than a second, the find-your-fit tool scans the contours of your face then recommends a set of glasses most likley to fit

3D photography has the potential to address “fit” problems not just in eyewear, but across other industries. According to the National Retail Federation, Americans returned $260 billion in merchandise last year, or eight percent of all purchases. Return rates of 30 percent or more are common for merchandise that’s bought online. Clothing is even worse. With the help of the 3D imaging, we may be able to transform how we size all clothing, from glasses to gloves to jewelry, shoes, shirts, pants, coats, and more.

A hat measurement tool from the 19th C. shows just how long retailers have been struggling to precisely measure the human body.

The ramifications for 3D measurements are also important for other categories: fit your living room with a couch or flat screen TV that fits your floor plan or wall space. At last, we can also address the age-old small apartment question, “Will this thing even fit through the door?” before it arrives. By empowering customers with new tools for precise measurements, we can limit waste, save retailers and manufacturers a lot of money, and streamline the shopping experience.

Thought experiment: What are some other product categories where computer vision and size-measuring tech can help address the paradox of choice?