What makes a review relevant?

Redesigning a review page to influence user input for an eCommerce review platform

OVERVIEW

Context
My role: Individual Contributor
Timeline: 6 to 7 weeks
Platform: Responsive web
Company
Sitejabber is a web platform for consumers to find trustworthy online businesses.

Their mission is to build a trusted community among shoppers to share their experiences and instill online transparency.
Business goal
Redesign the "Write a Review" page to encourage the reviewers to share relevant information about their online shopping experience.
Solution
Delivered within 6 and a half weeks by testing layouts to better influence users to input more relevant info into their reviews. Unfortunately, it could not be launched because it would have involved restructuring their databases to store the new information we could obtain.
Case Study Summary

THE PROBLEM

How does a layout influence the way a writer makes their review relevant, or not?
Our user and product research revealed a review is more relevant if it...
  1. Is not too short, or too subjective
  2. Includes when the experience occurred
  3. Includes advice
  4. Is fair and specific

RESULTS - BEFORE/AFTER

Asking questions in batches preserved user engagement and focused their responses
Our usability testing revealed the easier we made it for users to think, the more likely they were to engage. They cited it would respect their time and cultivate quality reviews.
I made sure we focused our iterations on fostering focus and encouragement to fill out the review forms. This led to...
  1. Prioritizing the "What affected your rating" sections to surface more useful details
  2. Removed "distracting" sections so the reviewers could focus on their review
  3. Added a "Review Strength" feature to help prevent low quality reviews from being submitted
BEFORE (Full review form)
AFTER (Full review form)
RESULTS

RESEARCH INSIGHTS

Surfacing the relevant details behind the "5-star rating review system"
I discovered research publications that suggested review systems which relied on the "5-star system" are less helpful to users because as more reviews are added, most reviews converge at 3 to 5 stars. This forces many users into laborious tasks like spending more time reading each review.

THE OPPORTUNITY

Engage and retain their most loyal reviewers
After also discovering numerous fake reviews, our goal was to improve the "Write a Review" page to help pave the way for Sitejabber's most loyal reviewers to contribute more reviews. To accomplish that, our research suggested to respect their time and contributions.

ITERATIONS

An unhelpful layout influences more irrelevant reviews
The team first learned that users would not further engage a review form if the design layout did not encourage it.
From there, the most impactful iterations were...
  1. "Review Strength" feature which came from how other sites encourage users to fill out their profiles
  2. Presenting limited options that are easy to answer which came from TripAdvisor's review form
  3. Copy iterations within review sections to make it more direct, expressive, and helpful
ITERATION #1 DIRECTION: Making the reviewer concentrate on providing more useful details
First screen users saw
Once they filled out the review
ITERATION #2 DIRECTION: Progressive disclosure works best when users knew the number of sections
Additional lessons learned...
First screen users saw
Once they filled out the review

VALIDATION TESTING

Making the asks easy can gain the most from your users
After conducting 5 more user interviews, the results were...

PROTOTYPES

Users knowing the end encouraged more engagement
Presenting all the questions through progressive disclosure improved user engagement because it showed how quick it would be for them to finish their review.

The GIFs below shows the improved design on desktop and mobile. For the mobile version, we focused on one review format and used a common design pattern found in other review platforms.

REFLECTION

Online opinions swayed by design
In trying to improve a user-generated content experience, I learned my business background proved helpful in specifying our research goals to navigate ambiguity. For instance, I led our second round of user interviews to uncover why they liked, or disliked the design changes we went for. This provided a clearer path to bring greater detail to online opinions within our designs.