As you may have seen on our Yelp Official Blog, we recently did a post on a new feature we are testing: Yelpy Insights. Here on the Engineering and Product Blog, we wanted to dive into the work our engineering team did to help discover and display Yelpy Insights.
The Idea: Yelpy Insights was actually borne from feedback we had received from vegetarian yelpers (and in fact, some members of the engineering team!) that finding a good place to eat vegetarian food was tricky. For example, some things that are liked by vegetarians are also liked by omnivores, but some things are only liked by vegetarians.
The search team set out to help create a better search experience for veggie users (and Yelp engineers), but that also got us thinking: could the same be said for other groups? Why yes, it can! We discovered many such use cases we’re excited to test, including connecting yelpers with businesses that best fit their different age groups -- because I know my Dad doesn’t want to be seated next to a bachelorette party at dinner. Or actually, maybe he would.
What we look for: By combing through our vast review data, we’re able to pick out folks who share a vegetarian perspective in their reviews and highlight the businesses that they like. It’s not without its flaws, but it’s a great way to crowd-source insights about vegetarian-friendly businesses.
In the same way that we dealt with vegetarians, we’ve identified places that are great for 20-somethings, 30-somethings, and 40-somethings. For example, if we find a restaurant with both an abnormally high concentration of good 30-something reviews, and a large overall number of these reviews, we can be pretty sure that a lot of 30-somethings like that place. That said, we never presume that you are only interested in one age group and you’re free to use this filter as you see fit.
How we show it: There are several ways for users to discover these results. Users can search the 'Liked by XX-somethings' or “Liked by Vegetarians” filter on many searches performed on the site. Additionally, these businesses will also have a special badge above their reviews and a link to nearby places like it.
And don’t worry, we wouldn’t dare share anyone’s private information. What we can do is learn from signals in the reviews and user profiles how different groups like various businesses in aggregate and use that information to improve your search experience on Yelp.
What’s next: Of course, vegetarians and different age groups aren't the only ones asking questions like "I want a business that does X, but I want one that people like Y like". Our large and diverse community of reviewers can help us answer these questions for a variety of other groups that have a unique perspective on the businesses they review. Got an idea for the next Yelpy Insight? Let us know via firstname.lastname@example.org!