Kira Newman via Tech Cocktail
Apps based on reviews, and especially today’s smorgasbord of food apps, have a problem: some guy goes to restaurant X, turns his nose up at dinner, and crafts a one-star review dripping with disdain—“It would get zero stars if that were an option.” Reviews like this drag down overall ratings to an ambiguous three stars, and hungry readers are left wondering: Is this just his eccentric taste, or is he right? Is the whole place bad, or just the dish?
Food Genius, a Netflix-inspired food recommendation app, tries to solve this problem in two ways. First, users rate dishes, not restaurants. And second, their recommendation algorithm factors in dishes that are rated highly by similar users to ensure quality, plus dishes that fit your personal tastes. Food Genius might learn over time that you adore dinners from Greece, hate fried food, skip desserts, or are enchanted by cilantro. Deconstructing food “DNA” like this allows Food Genius to make connections to seemingly unrelated dishes.
Cofounder and CEO Justin Mazza sees this algorithm as Food Genius’s main asset, in contrast to ratings-focused apps like Urbanspoon, Foodspotting, and Nosh. He also promises that they won’t launch in a new city without a comprehensive database of dishes; in Chicago, they now boast more than 4,000 restaurants and 175,000 individual plates.
“We see ourselves as a data company,” says Mazza, who saw a need for Food Genius when he found himself eating falafel lunches four times a week.
Algorithmic and personalized curation, mining and socializing the data in food reviews. Smart. A fuller reveiw to follow, once I actually try to use it, which for now is limited to Chicago.
Dishes are the natural unit of food, just like songs are the natural unit of music.