Executives hired from mobile ad service URX will help Pinterest better understand the kind of content consumers pin, recommend and click on.
At Tobi.com, an apparel retailer that emphasizes service from personal stylists, shoppers with computer webcams can view real-time images of themselves trying on recommended products. Gesturing with a thumb’s up saves the product in a virtual closet.
At Tobi.com, a fashion apparel retailer that emphasizes service from personal stylists, shoppers with computer webcams can view real-time images of themselves trying on recommended products. Gesturing with a thumb’s up saves the product in a virtual closet and increases the product’s value for all shoppers as a recommended item.
Tobi this week became the first online retailer to deploy Fashionista, an application that combines webcam motion capture, real-time personalized product recommendations and the blending of computer-generated product images with actual live video images of shoppers. Fashionista was recently launched by technology partners Zugara, which provides the interactive video and computer-generated content, and RichRelevance, which supplies the product recommendation engine.
Jeff Lee, vice president of products and technology at Tobi, says Fashionista complements the retailer’s policy of offering a personalized and fun online shopping experience not unlike what consumers expect from fashion boutique stores. “Fashionista not only helps answer the question, ‘How does this look on me?’ but it also provides for a more fun and engaging way to shop and discover products online,” he says. “This helps shoppers find new products on our site they might not find otherwise.”
Fashionista uses Flash technology commonly used on many personal computers. When a shopper clicks into the application on Tobi.com, a live webcam image of her standing in front of her computer appears within a window that features images of several recommended items like dresses and other apparel items. The recommendations that appear are based on the shopper’s known shopping behavior as recorded by RichRelevance web analytics.
To try on recommended products, the shopper simply moves her hand to make it appear near a chosen product on the computer screen; the “augmented reality” technology within the application then makes the chosen dress appear in front of the shopper’s image as if she were wearing it. If she likes the dress, the shopper moves her arm to make her hand appear on the computer screen near a thumb’s up illustration, which sends the image of the dress into a virtual closet and increases the relevancy score of the recommendation; if she doesn’t like the dress, she moves her arm to make her hand appear on the screen near a thumb’s down illustration that removes the dress from her recommended items list.
“The thumbs up or down feature is a huge help in making our product recommendations more relevant,” Lee says. Items placed in the virtual closet can be reviewed at any time and, with a mouse click placed into a shopping cart for purchasing.
Although Tobi.com emphasizes service from personal stylists who help shoppers pick out fashion items and who reveal their names and images in live chat sessions, the Fashionista application is not currently designed to allow shoppers at home to browse simultaneously with their personal stylists at Tobi. “We can envision that, but for now we’re just getting our customers comfortable with the augmented reality technology,” Lee says.
Fashionista is also designed to let the shopper use hand movements to save a still picture of herself appearing with a garment she likes, then use the mouse to send that image to her Facebook page to elicit comments from her friends.
Zugara and RichRelevance have not said publicly what they’re charging for Fashionista. But David Sellinger, CEO of RichRelevance and a former head personalization technology research and development at Amazon.com Inc., says Fashionista is designed with a flexable pricing model based on pay for performance.