This design challenge was a part of my application for the Kleiner Perkins design fellowship.
I was chosen as a finalist in the application process.
Prompt: Redesign one of the company's product features.
I redesigned Nextdoor's feed customization and ways in which users interacted with each other to reduce frequent experiences of negativity and frustration.
Duration: 1 week
I used both primary and secondary research methods to gain insight.
Synthesis & Problem Definition -
I analysed my data to find themes and define the problem space.
Ideation & Wireframing -
I sketched screens and digitized them, reworking them throughout the entire process.
I created my narrative for this case study.
I decided to start my research by getting a general sense of what people's experiences were like using Nextdoor: what they use it for, how they interact with others, what things they found frustrating.
I created a post and asking my neighbours about their general thoughts on Nextdoor. A couple of them replied publicly, and some private messaged me. I also sent out a survey to my friends who used Nextdoor and interviewed a few of them to learn more about their thoughts.
After gathering my notes, I sat down and began to look at larger themes. While many people had smaller complaints surrounding the technical issues of Nextdoor, I saw that most people experienced negativity/undesireable content and interactions. I also created two personas to uncover different ways that different people were experiencing these frustrations.
Given the time I had left to do this challenge, I then narrowed down my problem area to focus on the following pain points:
1. I want to see less content I find irrelevant in my feed.
A lot of people get tired of seeing posts about certain topics (e.g. lost pets, complaints). There was also general interest expressed about more overall categorization in the Nextdoor app.
2. I wish discussions were more civil.
Despite the full names being displayed on posts and replies, tones of hostility/aggression (even if very subtle) can often surface when people disagree. One of my research participants told me that when communicating with neighbours you've never met on a platform, "tone, nuance and good intent are hard to convey".
3. I want to encourage more kindness between my neighbours.
Going off the previous point, many people expressed their wish for more kindness and positivity in the way neighbours interact with each other via Nextdoor. One lady actually messaged me privately to share an intimate, positive moment she had when she reached out to help another neighbour. Even if kindness is present, it is often being dominated by more negative content.
Most of my research participants stated that they only use the Nextdoor mobile app, so I chose mobile as the platform for my solution. I came up with a couple ideas and sketched out a few wireframes as a way of exploring these ideas and revising as I went. (Feel free to ask more about why I chose certain ideas and discarded others!)
I eventually landed on 2 solutions:
Currently, from the homepage feed they can only choose the topics they want to follow:
I decided this space could be better suited for more options. By allowing choice for what content is displayed, the amount of frustration and negative feelings that arise from seeing the same type of content repeatedly could be alleviated.
The feed can be customized in 3 different aspects:
1. Neighbourhoods - which neighbourhood posts does the user want to read about? The choices depend on where they are located.
2. Posts - which types of posts does the user want to read about? The choices are the options under the pre-existing categories of "neighbourhood", "topics you follow", and "groups".
3. Sorting - how does the user want their feed content to be sorted? They can choose whether to see latest posts first or top posts first.
I created a flow for introducing current users to this feed customization feature. While creating the screens, I referenced similar pre-existing flows/screens in the Nextdoor app (onboarding, group creation, selecting which topics to follow) to ensure visual and structural coherency of my design to the rest of the app.
From the very beginning, they can choose to ignore or try the feature. If they did want to try, they will be brought through a simple flow with their pre-existing feed settings presented as the default, in which they can adjust.
Toxicity and negativity are problems that most social network platforms experience. I thought there was opportunity to explore the way in which users on Nextdoor interacted with each other in relation to the localized, tightknit nature of Nextdoor.
The first method I came up with is to make it easier to shift a tense conversation on a public feed into an environment with less pressure. I increased the visibility of the option to privately message a neighbour by adding a "message" button below each post/reply, beside the "thank" and "reply" buttons. This could also encourage users who have privacy concerns to engage with their neighbours without having to unwillingly share their information publicly if replying to the post.
The second method is to humanize conversations by introducing an additional sense. Nextdoor states on their website that it "enables truly local conversations that empower neighbors to build stronger and safer communities". There is already a higher level of trust from the amount of information on display (full names and addresses), and a user's network on Nextdoor is not ambiguous and widespread like other social media networks (e.g. Instagram or Reddit).
Given this level of trust, what if voice messages could be a way of replying, alongside regular text replies? This is more an exploration, rather than a solution, of a rather unconventional method of communication for a social media network. Being able to hear a person's voice can do two things:
1. The tone and intention of a neighbour's message would be much clearer by hearing them talk out loud rather than reading what they wrote, and
2. It further reminds us that there is a human behind the screen, which can often be forgotten in the midst of a heated debate.
I created a simple reply flow to show these two features:
Overall, aside from application purposes, I treated this design challenge as an exciting opportunity to learn more about my own strengths and weaknesses.
I found the most difficult part of the challenge to be the problem definition. I wanted to work on tackling an area with high impact at an appropriate level of fidelity in terms of the solution, given the amount of time I had to spend on this challenge. Online toxicity is not something that can be solved overnight, and I did not want to create bandaid solutions for this complex issue. I decided to incorporate my interest in information architecture as a part of my solution (feed customization), which I saw as a very real way of addressing the problem.
If I were to continue working on this challenge, I would user test the feed customization and see what areas may work better than others. The three methods in which a user can customize the feed currently exist at the same level; could certain ways be more likely used than others? If so, how could they be displayed in a way relevant to the different level of need for each method?
I would also do much more research around the idea of a voice reply. Even though I didn't have time to flesh it out extensively, I decided to include it in my write-up not so much as a definite solution, but more of a door to an unknown territory that could be worth trying.
Thank you for reading!