Company:
Wildcard
Role:
Product Designer
Project Type:
Mobile App, Internal Tooling
Lead Designer
Design Chair
2 Project Managers
5 Editors
1 iOS Developer
1 Android Developer
2 Front-end Developers
3 Backend Developers
2 Data Scientists
1 CEO
When Wildcard 2.0 launched as a news aggregator, it received around 90K installs. Retention was soon revealed to be a problem; the app would lose almost 90% of its users within 30 days after the initial download. It’s a common problem with apps.
I worked with the lead designer to understand where the app needed to catch up with the users and to devise solutions to minimize the gap between short- and long-term usage.
In-house Wildcard editors collected, curated and uploaded the news feed through the backend. This created a bottleneck for the app's frequency of news article updates. Another challenge for the editors was to keep the feed as neutral as possible. Fake news and siloed opinions began taking hold on social media, and the users questioned the integrity of a curated newsfeed. One of the significant challenges for the design solution was how to allow users to personalize what they see within our current infrastructure without creating an echo chamber.
We contacted old and new users to understand how they used the app. Multiple users said they only opened the app once or twice a day; during one of those interviews, one user said: “The top image slideshow makes it seem like nothing has changed.”
The existing structure doesn't make it feel like the app updates frequently. New articles are populated under the slideshow, and articles related to the top stories are accessed by revisiting the full story in the slideshow.
When asked about their news reading habits, one user said: “I normally use another RSS reader like FlipBook because I get to see what I like.”
When asked how the app could be improved, one user said: “I wished there was a way to hide what (news topics) I don’t want to see.”
Customization of the feed was also something users were interested in. The available version only allowed users to share news articles.
Surface more exciting content to users while working with existing content creation workflow.
Introduce and explore content personalization while considering technical debt.
While working with the existing infrastructure, we created a feature called the Topic Targets to help surface more content and optimize the workflow for the editorial team. I did some contextual inquiry by sitting with the editors to see how they worked together and separately. There were five editors, and they each could collect and push content to the app. Often, topics overlap, and there wasn’t clear visibility within the team to see who was covering what. That often leads to a lot of back and forth verbally and lost time by double posting within the group. By introducing the topic targets, the editor-in-chief can assign or recommend a topic that should be covered that day and see what has already been pushed to the app related to the target.
The first step we took for personalization was adding another interaction layer to the existing news cards. The previous version of the cards allowed users to save the card for later by swiping right or share the card by swiping left; by adding the ability to remove the cards, users can hide what they do not wish to see. We went with this approach because it worked with the existing flows and allowed us to start saving data on what the user's likes and dislikes as a first step to help further personalize their feed.
With the rise of machine learning and virtual assistants, Wildcard used its data scientists and engineers to explore creating an assistant for news. This attempt was not only to do something groundbreaking but could also solve personalization for users, surface more content, and increase engagement overall. We created a v.1 of Zoe using personalization through a series of surveys made by editors and allowing the answers selected by users to connect to various topic tags. These will then populate more or fewer news articles on a user’s feed. In Zoe v.2 we included a Tinder like card selection for an additional layer of personalization.
I was privileged to work with and learn from some of the industry's most brilliant engineers and designers. I added many skills to my design tool belt, and I was able to merge UX with my visual design background in a work environment. I learned much about startup culture and the work needed to make a good app. Unfortunately, their investors cut the runway for the product, and the company folded. The product lacked clear growth goals, and pivoting from a native web browser to a news aggregator left too many open issues for the app to retain traction with users and convince stakeholders of its value. In the end, it is all about timing. There were features and marketing strategies could’ve been pushed sooner to help the app move further along.