When you’re looking for a new place to eat in Boulder County, you’re liable to check Boulder Weekly, of course. But after that, you’ll probably Google some restaurant names or some food types and see what comes up.
Often what returns are links to user-reviewed food rating sites like Yelp, TripAdvisor, OpenTable, Google Reviews and Urbanspoon to name a few. Attached to those links are ratings that give you an immediate numerical judgment, usually out of five points or stars, as to how good a specific restaurant is. Your brain likely amalgamates all the rankings into a thumbs up or thumbs down vote on that restaurant. That means a 15-second Google search can mean the difference for whether or not a restaurant gets your business. Data from a University of California Berkeley study actually suggests that an increase in a half-star ranking on Yelp makes a restaurant 19 percent more likely to sell out on a given night.
But how effective are these sites at objectively reviewing restaurants? Who are the people reviewing on user review sites, and why do they rate restaurants the way they do?
Most of these sites have a five point or star rating system. A restaurant can be given one to five stars overall, and some sites allow users to give individual stars for aspects of the dining experience such as ambiance, service, food, etc. However it’s sliced, the first thing a user will see is an overall star total that averages all of the reviews of all the reviewers. Some sites are more specific than others, but it’s an oddly standardized system amongst all the social media review platforms.
Take Yelp for instance. Its five-star ranking system comes with guidelines for each star quality. A one means “Eek! Methinks not,” a two means “Meh. I’ve experienced better,” a three means “A-OK”, a four means “Yay! I’m a fan,” and a five means “Woohoo! As good as it gets!” There are also guidelines for decency and truth, and in fact you may have seen any number of news stories where restaurant and business owners sue Yelp reviewers who go over the top with their review.
Or take Google’s ranking recommendations. They range from “hated it,” to “disliked it,” to “it’s okay,” to “liked it” and “loved it,” with each correspeonding to a number one through five. The general gist of Google’s and Yelp’s ranking services is similar to most other review platforms. So assuming the five-point standard, I went through about 220 restaurants in Boulder and took the average of their Yelp, Google, TripAdvisor, Urbanspoon and Foursquare rankings.
I found that about 71 percent of restaurants in Boulder analyzed received a 3.5 to 4 point average. That statistic tells us a couple things. One, it means these ranking sites don’t rate restaurants relative to each other, or on a curve. If they did, we would then see an average ranking of three points or stars overall. It also tells us, perhaps more pertinently here, that Boulder diners either think 71 percent of Boulder restaurants are “Yay! I’m a fan!” or “Liked it,” or it means that the manner in which social review sites aggregate rankings is flawed.
The truth is that Yelp reviews are positive. After all, you would think that the majority of reviewers are likely to review a place if it was either exceptional or terrible in their opinion. About 83 million people visit Yelp every month, and data from the company shows that four or five star rankings take up 67 percent of all reviews, while one star reviews account for 14 percent.
But statistics show that the rating systems of these social restaurant review sites follows the so-called 1/9/90 rule. According to social media expert Susan Kuchinskas, about 1 percent of users create content; 9 percent of users engage with that content by editing, sharing or rating; and 90 percent of social media users just read the reviews. Yelp says its data abides by this rule, and judging by this non-statistician’s individual research and a general survey of review sites, it seems most follow this rule.
And so what we’re left with as hungry potential diners just looking for a decent review (I realize that this sounds like a long-con plug for the review section, but it’s not) is a bunch of restaurants that rate similarly. Diners have discerning taste and these sites don’t accurately represent that. And it is unfair that we use them as a quality marker whether we do it subconsciously or not.
But the savvy diner would look at this information and say that the only real way to judge a restaurant is to get out and try it for yourself. After all, people can rate a restaurant poorly because they care about napkins and the napkins were blue instead of black. People come to reviewing with different backgrounds, tastes and prejudices, and what these rating aggregation sites do is basically cancel everyone’s peculiarities out and provide a standardized test score. But we don’t shop or eat based on what the consensus says, we shop for and eat what we like. And so you have to wonder what is the value of these sites, and why do they have to carry such importance for nearly every restaurant and business.
