Imagine you live in Williamsburg, Brooklyn. You know your neighborhood like the back of your hand. Including all of your favorite, top-secret restaurants and shops. On your home turf, you have things handled. But when you travel, things are a different story. You’re a fish out of water in Los Angeles, or Seattle. Which means you’re constantly ending up at the wrong restaurants, in the wrong neighborhoods. Even in a familiar environment, trying new places can be completely hit or miss. A big part of the problem is that online reviews aren’t written through your lense, and thus another person’s idea of chic and modern, to you seems drab and passe.
Enter personalized Web search by Nara. The Cambridge, Massachusetts-based startup believes that it’s no longer enough to deliver the most universally relevant answer to a query. In this day and age, it’s necessary to deliver the most personally relevant answer to each user. Today, the company is announcing $3 million of new financing following its $4 million June Series A round. “Call it a Series A-1,” says co-founder and CEO Thomas Copeman. “We had so much demand with our round this summer, and we’re seeing exciting new opportunities to invest in growth.”
The CEO and his co-founding CTO Dr. Nathan Wilson have been working for more than two years on the next-gen search technology which relies on an artificial intelligence to connect relevant information with interest graph data on each user. “We are now deploying massively parallel neural networks into the cloud to [automatically organize and personalize information] at Web scale and bring order to a growing number of categories,” says Wilson. “Nara’s vision is to build a more personal, actionable and liberating Web,” says Copeman.
The company has applied this technology initially to the restaurant search sphere, but this is little more than a jumping off point in its grand plans. The same level of inference can be applied to hotel recommendations, books, movies, music, and countless other categories.
Along the restaurant lines, in addition to today’s funding announcement, Nara is also announcing a partnership with online and mobile food ordering service GrubHub, to complement its existing OpenTable relationship. The new integration will go live by year’s end. Through these channels, the intelligent search platform collects semantic review and description data on each venue which it then structures into a database to compare establishments in different cities, genres, and price ranges.
The Nara platform currently analyzes over 250,000 restaurants, coffee shops, bars, and cafes in 25 major US and Canadian cities. Through its GrubHub, OpenTable, and FourSquare integrations, users can now see restaurant menus on merchant pages, as well as read reviews of each.
To get to know its users, and solve the “cold start problem,” Nara asks a series of basic but informative multiple choice questions. (You can also connect your social media profiles for further personalization.) First, users are asked to select “Which tables appeal to you?” – you can choose one or several representative photos among, formal, classic, intimate, trendy, hipster, and casual. Next, it’s which cuisine(s) to you enjoy? French, Chinese/Thai/Vietnamese, Japanese, Italian, Indian, Mexican, American, or Mediterranean? Next, how would you spend a day off? In the city, with nature, on the beach, with family and friends, exercising, staying at home? Finally, users are asked to name their two favorite restaurants in a city they know well. From this broad picture of a user’s tastes, Nara is able to deliver very specific and personalized recommendations that will only approve over time. The company’s tagline is appropriately, “A life well found.”
Nara is available on the desktop Web at Nara.me, and on iOS and Android. Since launching its mobile apps in mid-November, the company reports a doubling of its traffic, although it declined to quantify that statement further. Copeman did confirm that the platform will expand beyond restaurant recommendations into two additional “consumer interest categories,” in Q1. Given the planned breadth of his offering, I asked the CEO which company he thinks Nara is most competitive with: Google, Foursquare, or Yelp. He said, “I would say we’re most distruptive to Yelp. Foursquare and Google offer very different signals and value propositions than we hope to. There’s room for all three.”
Both Nara’s Series A and Series “A-1” rounds were funted entirely by Peter de Roetth’s Boston asset management firm Account Management, LLC. With the $7 million in capital raised in 2012, Nara has trippled its team to 20 members, including multiple MIT neuroscientists, computer scientists, and astrophysicists, as well as engineers and designers. Copeman deliberately chose Cambridge, as opposed to his previous home in Los Angeles, or the more knee jerk choice of Silicon Valley. “There’s no other place in the world to build this,” he says, citing availability of world class talent, lack of distractions, and the ability to engage with experts across a variety of disciplines.
The idea of personalizing and curating the Web’s information is an area of intense interest. Many companies are trying to tackle it within single verticals and use cases, such as online video discovery, fashion recommendation, and yes, even restaurant recommendation. The Nara we see today, however, is just a glimpse at what Copeman and Wilson have planned. Personalized restaurant recommendations are a nice, clean way to introduce the concept, but there’s far more in the works. Going forward, expect the company to tackle and increasingly wide number of categories and soon add international support. Its an ambitious and exceedingly challenging problem to solve. Here’s hoping for all of our sakes that they pull it off.