2-month project from Google News Lab Fellowship 2016-17
dec 2016 - feb 2017
Seoul, South Korea
5 members including a journalist, designer, programmers, and video artist
Sketch / Zeplin
Adobe Creative Suite
Project O is from Google News Lab Fellowship 2016-17. It is composed of four different media contents all dealing with the idea of filter bubble, "a state of intellectual isolation that can result from personalized searches when a website algorithm selectively guesses what information a user would like to see based on information about the user, such as location, past click-behavior and search history" (Wikipedia)".
All team members participated in the ideation process.
As UX/UI designer, I was especially responsible for B.I design as well as constructing the web-platform that puts all the contents with logical information architecture and navigation system. The following is the list of the products our team published. All of them are collected as a group on a website and individually distributed in diverse ways through multiple channels using inbound marketing strategy. Details about each product are introduced below.
The project is composed of three big parts:
how is your newsfeed?
a Chrome extension which analyses user's recent 200 posts on his/her Facebook Newsfeed and tells the top3 media of which post is most frequently appeared with along with the list of friends who share similar Newsfeed with the user.
this is your life, living in a bubble
an interactive storytelling animation using parallax scrolling. The story is about Michael (democratic advocate) and John(republican advocate) during 2016 Presidential Election in the US whose Newsfeeds are completely different from each other.
filter bubble, the invisible trap
An article and a short documentary film. The article is a data analysis report based on the data collected for a month from the first content. The video is about an interview and extreme political bipolarity conflict during the Korean president's scandal last year.
As Facebook has become a critical platform of news consumption in South Korea, there are rising awareness of filter bubble, resulting many people to the danger of consuming the news from certain media with specific political biases. Therefore, our team decided to make a series of contents that can raise awareness about this phenomenon.
To define target users, their pain points, and ultimate project goat with design solution, we used affinity diagram method through card-sorting process. The data is derived mainly from current journals, simple user interviews via phone call and in-person meeting.
Below is the list of the insights gained from research and team discussion:
Overall, I was responsible for designing a web platform that contains all of the separate contents. As a UX/UI designer, the most important thing aspect was to make users / audience to easily navigate over different contents in our website no matter how - through which content - they land on our page so that they can experience the whole project rather than separate pieces. Another responsibility as a designer was to have a clear and attractive visual system that not only goes along with the Brand Identity but also lightens the weight of the topic - filter bubble - that seems to be a bit intricate and difficult to understand for the target user.
user flow overview
The diagram below shows the overall user journey layout.
How is Your Newsfeed?
The following images show how the first interactive content is displayed.
After the chrome extension is launched, the extension collects recent 200 posts from the user's Facebook Newsfeed, and gives the percentage of the media contents that compose user's Newsfeed along with the top 3 media companies which the contents are from. Also, it gives the list of the top 3 friends who share similar Newsfeed.
On top left is the percentage of the contents published from media company among the 200 posts that the extension program has crawled. Three blue bubbles on bottom-left side is the top 3 media company that appears most frequently, and top-right side shows the three friends who share similar Newsfeed with the user. Also at the bottom, there is a share icon by which a user can share the result on his/her social network. Finally, at the lowest bottom is the navigation button that allows user easily access other contents of this project.
limit and backup plan
One of the biggest drawbacks of our project was that the interactive contents are both working on pc web. Considering the overflowing proportion of audience consuming digital contents through mobile platform, it is critical to get a backup plan through web-app that can meet their expectation and also - hopefully - lead users to access through pc web as well. Therefore, instead of crawling the newsfeed posts, we devised a backup idea to use user's like lists to analyze his or her political bias. Whether the liked media contents is politically biased was calculated by the leaning ratings calculated by a recent study. Though the data and the result might not be correct, the purpose of this program is to get user's interest into the brand's contents.
According to the Google Analytics results, there was total 1,200 hits on our landing page during the first five days after launching. Most of the access points were via Facebook page where we shared our digital contents consistently. Also, the UX goal which was to make users to go through all other contents no matter how they have accessed our contents was quite successful considering there was low bounce rate - more than 70% of users actually navigated through whole other links within our website. The most effective element that led this behavior was the top navigation bar and the link to other contents on every page - which tells importance of valuing user journey and significance of storytelling that drives them to have further curiosity in relevance to the previous contents.
However, one of the biggest obstacles in this two-month project is that there was constant change in plan. Therefore, there was no enough time for our team to test other strategies other than first few without proper user tests.