The App That Knows Your Food Likes Better Than You Do
Two weeks ago I arrived late in the day in Singapore after a long flight. I had not done much research about restaurants in town, so I dropped off my bags and wandered to Club Street, a hilly lane of restored traditional shop houses in Chinatown. Lively, attractive restaurants lined the street. I randomly picked one place. It was pretty good as well as expensive, as are many things in super modern Singapore. But I left feeling I might have done better with a little more planning.
I thought about that meal after I learned about , a company that promises to find restaurants that you love anywhere, based on your own preferences. The site calls itself a “personal discovery engine,” which means it painlessly leads you to restaurants based on the data you have entered into the site.
There plenty of apps to help you find a place to eat: Foursquare, OpenTable, UrbanSpoon, Yelp. Nara claims it uses a new type of neural-network algorithm to do the job better, using more sensitive analysis of your (and everyone’s) picks and pans. To date it has made its money through revenue-sharing deals with , Uber and GrubHub. The service is opt-in only and Nara says it does not share any data about its end users. The company started in 2010 and last year raised $7 million from angel investors.
On Tuesday, the Cambridge, Mass. company announced a deal to license its technology to Asia’s leading telecommunications group, . The deal, the first of its kind for Nara, marks the latest step in greater personalization of the Internet. It also raises questions about how companies will use the ever greater amount of personal data they are learning about us.
SingTel is not disclosing details about how it will use Nara’s recommendation engine, saying only that it will deploy first in Malaysia, then Australia and Singapore, for restaurant recommendations. But Nara is planning to move beyond dining into financial services and other commerce categories and sees SingTel as a way to do that. SingTel has made a greater effort to buy and license apps for its 468 million mobile customers than typical U.S. wireless carriers. It even bought a mobile advertising business last year.
As more people enter their likes and dislikes in Nara’s engine, the overall data network will improve in its sensitivity. “An important component of Nara’s engine is its ability to learn from all of its users across the platform,” Nara CEO and founder Thomas Copeman said. “Eventually, all users across all businesses that are powered by Nara can improve our knowledge about both user tastes and preferences and related venues and entities.”
As its sophistication grows, the service should be able to recognize national differences in taste. For example, it would steer Brits and residents of its former colonies toward restaurants that doled out the Marmite while sparing Americans the unpleasant sensation of eating it for the first time. “Over time, we believe that this could be an incredible hyper-local value-add to SingTelGroup’s half a billion mobile customers,” said Loo Cheng Chuan, head of SingTel’s Digital Life.
But blending services into one huge personal profile may not go down well with those who don’t like being data-profiled as they move around Asia and across multiple apps. Nara sees itself more as part of the solution, not the problem. “Nara provides a ‘privacy silo’ where the system can work and transact on the behalf of the user,” Copeman said. “We almost imagine it as a personal anonymizing avatar that acts on your behalf to gather information and act with the internet, which is especially crucial given the growing number of services who are dedicated to linking your transactions to your identity.”