Re-designing Yelp Reviews



the problem
Yelp is a platform that is essentially an online directory for various local businesses. The platform creates a community as users can leave reviews, ratings, and images from their experiences at the businesses and other users can then look back on these reviews to make their decision. I believe that these reviews were designed for "evil" and in this project I worked to come up with a more ethical design flow for these reviews.
my role
context
-
critiqued current Yelp review design process
-
conducted and analyzed user surveys
-
designed new screens in Figma
6 week project for Designing for Evil class to understand how certain features could be designed unethically.
Team size: 4
tools
-
Figma
-
Google Forms
-
Google Draw
the design process
01. understanding the problem
02. research
03. framing the solution
04. design critique
05. re-design
06. reflection
01. problem
We believe that Yelp Reviews are evil through three moral lenses:
utilitarianism:
actions are moral or right if they benefit the majority
-
The current state of reviews can both harm the application users as well as the business owners.
-
For example, in the instance of fake reviews, the user may end up choosing a different restaurant based on incorrect information.
-
Similarly, this can deter users from visiting a restaurant for these incorrect reviews.
libertarianism:
maximizes political freedom and autonomy, minimal intervention.
-
Yelp reviews do not give the user the freedom to look at all the restaurants in the area. By default, the application shows the highest rated restaurants and restricts the user’s freedom to choose their dining plans.
deontology:
the morality of an action should be based on whether that action itself is right or wrong, rather than on the consequences.
-
Specifically, with encouraging users to post often by giving them the title of an ‘elite,’ the application has a greater intention of getting users to post, or interact with Yelp, rather than help with finding a new restaurant.
-
Although this evil interaction push may be seen as a stretch, a corroborating factor is that Yelp forces users to download the application, rather than just search up a single restaurant on the web. The motivating factor is to get users to interact with the application, rather than simply to help users find new restaurants.



Recommended reviews can be or sorted by popularity, rather than reliability.
Right now, reviews can be non-descriptive, and unhelpful to users
Negative reviews, they may be hateful or biased, providing unrelated information.
02. research
User research was used to empathize with and understand the needs of the Yelp users. I wanted to understand how they were interacting with the review features and how the interaction could be changed to be more ethical. There were three methods of user research performed: interviews, surveys, and a competitive analysis.
53 surveys || 3 interviews || 1 insights
key insights
-
ratings matter though reviews are not reliable
Through all three of our interviews we found that though many people do rely on high ratings (anything 3.5+ stars) all 3 of our interviewees said that they did not find the reviews written were reliable. Most of these reviews are written in extreme situations, whether they had a great time or a terrible time. Also, we found that it is important at how the review was written. With longer paragraphs there is more credibility as the user took time to share their information.
-
photos matter
User’s heavily rely on photos as the aesthetic mattered a lot. The way the food looks and the way it is presented matters as well. Word of mouth is something all of our interviewees expressed but photos mattered just as much as they could take a look at something and make judgement for themself. They also are able to create an expectation of what they will see when they do get to the restaurant.
Users really care about who the person writing the review is. More user’s will trust someone that is a top reviewer, or has an elite badge next to them as they are reliable sources and have some credibility to their name.
-
type of reviewer matters
interview quotes
business owner on discrimination through Yelp
“We have had instances where someone has left a nasty review with incorrect information”
average reviewer on what makes a review reliable
“I always try to include a photo of what I got. And then I always talk about the customer service, since that's a big deal to me”
survey results
75% of users use Yelp less than once a week, usually to get other people’s opinions on a business
Over 65% of respondents look up a restaurant to decide if they want to eat there
28 respondents would believe positive reviews over negative ones
competitive analysis
This analysis compares the benefits among other platforms providing a similar service. The goal was to better understand the current market for this service
Platform
Strengths
Weaknesses
TripAdvisor
language, type of review, and date filter; provide specific details for reviews
Focuses only on traveler’s needs (rental cars, hotels, etc.)
Places/
Reviews
Very versatile, easily accessible, allows user to filter by time
no photos or explanations for rating; can't filter reviews by rating.
Zagat
Rates different aspects of restaurant: food, service, aesthetic
Geared toward restaurants in larger cities; no user reviews
03. framing the solution
design principles
honesty
the re-design would need to provide users with honest, unbiased feedback to not only help them, but also help the businesses; prevent dishonest hate.
credibility
the new design would have to provide more explanations or photos for users to read to ensure credibility of the review.
design challenge
How might we re-design Yelp reviews to be more credible and unbiased so they benefit the companies and the users?
04. design critique
utilitarianism -
Dishonesty is painful, and Yelp currently allows this. We would redesign the review to include a more visible and useful verification feature for if you have been to the restaurant. By adding a feature that restricts the users to only write a review upon verification that they did in fact have an experience at this restaurant, it increases reliability of the review. This can be done through scanning a QR code provided on a receipt or through scanning the receipt itself. These reviews can be pushed to the top and advertised as “verified” reviews. We were motivated to emphasize this because of our interview insights about reliability and credibility of our user.
kantian ethics -
For reviews that are negative, we would want to redesign how users leave negative reviews. Instead of simply text responses, we would encourage users to both add a picture and use descriptions from keywords, and give depth to their reply. This would promote doing what’s right, and helping the business themselves improve through these responses. This connects to our insight that our interviewees really care about the photos.
teleological -
The purpose of Yelp is to help users spread the word about a specific service or restaurant. Having reviews that are hateful, discriminatory, or unrelated to the service, is detrimental to the purpose of Yelp. In order to combat this, our redesign will include a pop-up that encourages users to think twice before posting hateful comments. In addition, we hope to have a better reporting feature for business owners to report hate comments.
05. the re-design
design sketches
These sketches implement the design critique solutions mentioned above. From left to right, each of the design pages show the Utilitarianism, Kantian, and Teleological solutions to making Yelp less evil.



