Top Channel redesign
Launched in 2011 for a Japanese audience, the SmartNews app has remained largely unchanged. Naturally, there is a fair amount of tech and design debt when trying to create new interfaces or UI. When the US app was launched in 2014, not much was changed architecturally.
Many of today’s design and engineering constraints revolve around our legacy tech stack, and how it is catered to the Japan edition of the app. We have mostly been able to get by until recently, when engagement and retention numbers are less than desirable.
Company
SmartNews, Inc.
Role
Lead Product Designer
Goals
Refresh Top Channel UX/UI to boost retention and engagement
Define design principles for designing future feed experiences
Upgrade UI to meet WCAG 2.0 standards
Opportunity
A large percentage of users are arriving at the Top channel for the first time, leaving without engaging at all, and never coming back. By considering the design of the app from a holistic perspective, as opposed to the many piecemeal tests done in the past, we can create a more cohesive browsing and reading experience. The hope is that this will drive user satisfaction and therefore retention.
Problems
User
Poor Accessibility Design
Confusing Navigation
Hard to Discover Content
Inconsistent and Outdated UI/UX
Business
Declining engagement and retention metrics
Feed architecture based on JP framework, content, typography
Legacy tech stack = Inflexibility of UI in the feed
Lack of process for creating new UI or using existing patterns
Goals
Reconsider the top channel experience for users, and realize a revenue increase, along with at least neutral engagement + retention
Explore implementing large cells (rich content/revenue)
Improve quality of UI, including meeting WCAG 2.0 guidelines
Consistent design language, principles to guide future feed decisions
Challenges & Constraints
Any changes to the feed needed to see positive or neutral engagement/revenue metrics
Any redesign needs to align with JP version of the app, where the SN brand is much more established and
Legacy tech stack based on JP framework launched in 2011
Cultural/Geographical differences between US/JP audiences
Colorful channel tabs were not easy to remove/replace as they became a branding element in US/JP app editions
Due to no previous email capture, we had no way of contacting our most engaged users for interviews, usability tests, A/B tests, etc.
How to construct an information architecture
which encourages content and feature discovery, and allows for easy navigation between areas of interest
Discovery & Research
Part one: User segments and pain points
SmartNews had some pre-existing user segments defined by our product marketing team, so we started there with our research. We decided to recruit both SmartNews and non-SmartNews users, and conducted usability tests using shell versions of our app, as well as competitors. This allowed us to see what first time users as well as churned users thought about our app side by side with our competitors. After synthesizing our results, we came up with this list of user problems:
App Felt Overwhelming / Cluttered
Confused by Where to Find Our Key Features or Content Types
Inability to Customize
Messaging Confusion - Don’t Understand What’s Being Offered
Content & Content Presentation is Off-putting
Lack of Categorization, Cognitive Overhead in Feed
Next, we worked with our Data Science team to view some data and click funnels, to better understand how traffic was moving throughout the app. As you can see in the orange squares, there was a sharp drop-off by a large percentage of users without really engaging with the rest of the app.
Taking into consideration the problems listed above, there was a clear correlation of why users would not be engaging in feed content or retaining overall.
Part two: Research Objectives with answerlab
We partnered with AnswerLab, a research agency, to conduct series of user interviews, usability tests, and competitor/industry research.
1. Gain a good understanding of target users’ mental modes of news feeds UIs
2. Identify best performing usage patterns for large cells in the news feed which bring real value to target users
3. Determine how to reconsider the usage of the top channels tab system and the GNB, in the context of the feed, to make for a better experience for target users
In preparation for the round of diary tests being conducted by AnswerLab, we prepared the following prototypes:
Define
From our research sessions, we defined some user personas, which helped us decide who to think about when prioritizing who we were building for.
Ideating, Prototyping, Tests
Initial testing key findings
Participants felt that the ability to customize their feed was very important to them and is something they would want in any news platform they use
Those who preferred smaller cells liked that they could see more articles in one view and the smaller size/uniformity of the articles, which allows for quick skimming, however most also preferred cells with “additional features” over uniform cells
Participants indicated a heavy reliance on the top channel tabs as a means of navigation, stating that it prevented them from having to scroll excessively to find relevant content
internal dogfooding / stakeholder feedback
After reviewing the test results with the team and XFN stakeholders, we determined a few things:
Although the large cells were popular and visually appealing, they were still far off in terms of engineering capability
Smaller cells allow users to see more content, as well as satisfy revenue concerns
Keeping the colorful channel tabs was important because it provided a clear means of navigating to other content areas in the app. It also acts as a primary branding element, since the JP app is known for it
Design
We decided to scale back some of the explorations with large cells and minimal channel tab UI colors. This also gave us an opportunity to explore potential new UI for the colorful tabs, although we would stick with our existing UI for the upcoming A/B test.
Results
Impact
Concepts that were tested in usability test as well as A/B tests showed some promising metrics, although this was just a small sample size and more testing was planned before a full release. Some early numbers showed:
FQ7: +1% - Users were retaining at a higher weekly rate
TS: +3% - Users were spending more time in the app
PV: +4% - Users were discovering more content
Reflection
Throughout my time on the project, I was happy that we were able to accomplish some impactful things that will setup the product team for future success:
Updated UI to meet WCAG 2.0 AA guidelines
Introduced product principles for building new feed experiences
Discovered key learnings that will help the product team
Unfortunately my time at SmartNews ended before I could see the final iteration of this, but I am hopeful that we can see some of these updates in the app soon. I had an amazing experience working with so many talented folks, looking under the hood to really understand what was happening with our users’ journey.
Thank you!