It’s lunch time, you’re ordering from a new takeout place, and you have one important question: What’s good here?
If you have the Foodler mobile application, the answer is now on your smartphone.
Boston-based Foodler has been rating restaurants and helping people order meals online since 2004. It’s been a helpful tool, but hard to distinguish from competitors, like GrubHub, Seamless and Yelp.
Now, Foodler (also available on your computer) aims to stand out with a recently launched rating system that evaluates individual dishes, rather than an entire restaurant.
A diner ordering from the Ginger Exchange in Cambridge, for instance, could learn that people who’ve eaten there give the restaurant an average rating of three out of five stars on Yelp, and the same score on GrubHub. But there’s no telling how the mandarin orange vegetables and tofu compares to the miso eggplant.
A Foodler user could make a menu choice, knowing that veggies and tofu is an average, three-star dish, but that the eggplant is a four-and-a-half star delight.
“Every rating is actually from someone who ordered that particular item,” said Foodler co-founder John Jannotti. “So if you don’t like a restaurant, you can’t just go and rate every item poorly. You can really learn a lot from the individual ratings that you can’t learn from an overall rating.”
Nine years in business, Foodler has expanded to 50 employees and 48 states. The service is free to use; the company makes its money from more than 12,000 participating restaurants, which pay commissions on orders placed through Foodler.
The dish-by-dish rating system is only a few months old, so small sample sizes diminish the value of ratings, in some cases. But dishes at many popular restaurants already have been rated hundreds of times, offering hungry menu browsers precise guidance when ordering.
As time goes on, Jannotti noted, the ratings will become more meaningful.
Similarly, the longer you use Foodler, the better its recommendations get. The service tracks ordering habits and—with increasing speed—can learn that you love dessert, or don’t eat meat.
“We used to treat every single word in a menu description as different,” Jannotti said. “You may never have ordered chicken, but we may not have had enough information to know whether you like sausage or venison, because that’s a weird word. Now, our software recognizes that all those things are meat, so we can start to realize you’re a vegetarian long before we used to be able to learn you don’t like venison.”