How might we catalyze conversation amongst strangers within a social network competing with other networks for attention?


User research, Strategic innovation, UI/UX design, Prototyping and front-end programming


I chose to focus on Nextdoor for this redesign challenge. Through detailed research, I encountered instances of Nextdoor neighborhoods that were 'ghost towns'. Within this project, I raised some high level strategy suggestions, and chose to focus on improving the onboarding and post initiation flows at the product level. I was able to prototype and test my designs, which can be found in the Redesigns sections below.


2 weeks


Interviewed potential users, conducted heuristic evaluation and broader competitive landscape analysis, redesigned two specific features and prototyped these designs.


The Challenge

Redesigning Nextdoor

KPCB has many amazing companies that I’d have loved to tackle in this challenge. However, Nextdoor quickly caught my eye as an interesting project. I have been a member of my Nextdoor community for a year, and have occasionally considered how my experience with the product could be better. I am also passionate about cities, and have worked on urban problems in other interaction design projects. Neighborhoods within cities are unique entities. The turnover of residents in urban neighborhoods can be higher, and there are alot of more strangers around. The density of a single high-rise building could mean that building has more residents than whole neighborhoods in less-urban regions.

Jane Jacobs in Washington Square Park, NY, 1963. (Photo credit:   Fred W McDarrah/Getty Images)

Jane Jacobs in Washington Square Park, NY, 1963. (Photo credit: Fred W McDarrah/Getty Images)

In fact, it was a prominent advocate for neighborhoods, Jane Jacobs, who contributed to my early education in user-centric design through her advocacy for a people-centric approach to urban planning. As she wrote in Downtown is for the People:

"There is no logic that can be superimposed on the city; people make it, and it is to them, not buildings, that we must fit our plans."

Downtown is for the People, 1958

Just like city plans, it is to people that we must fit our designs for products.

I have documented my design process comprehensively below; you can cut to the meat of the project by skipping ahead to the Redesign sections through the links below. tl;dr I did research, prototyped and tested my designs, and I had some general strategic recommendations for Nextdoor.

The Process

What do people want from their neighborhoods?

I began as I always begin - by talking to people. I did quick interviews with five friends, asking them about what they were looking for from their neighborhood. The interviewees were all young professionals, fluent in English, well-educated, city-dwellers across different boroughs in NYC. I do recognize that this might not be representative of potential Nextdoor users, however I got enough valuable insights from these conversations. All five were non-Nextdoor users.

As a Nextdoor user, I have personal experience with the product to draw upon. My neighborhood’s network is a ghost town. I only see notifications of new neighbors joining, and the occasional babysitter request every 6 months. I posted a question about food recommendations, and a week later was still waiting for a response.

My experience isn’t unique. I was seeing the same feedback below from several sources:

"It seems like people either don't really know about it or if they do, they join and not much is going on… I see lots of empty pages in my neighborhood."

— a close friend

To supplement this primary research, I read widely around neighborhoods and technology, diving into research papers and books and blogs of urban planners. I also read articles around Nextdoor, and watched interviews by CEO Nirav Tolia to get a better sense of the challenges Nextdoor themselves had identified.

My research spanned articles, research papers, interviews, and books on cities and neighborhoods.

My research spanned articles, research papers, interviews, and books on cities and neighborhoods.

Finally, I had interviewees test out Nextdoor. The majority were seeing Nextdoor for the first time. We discussed initial impressions of the product at several steps, including the sign up process, the existing conversation topics, and how they could see themselves using Nextdoor. I have noted specific feedback on the current Nextdoor UI within the Redesign sections further below.

I identified two key problems areas. First, there was little activity within Nextdoor neighborhoods to attract users to engage. Second, neighborhoods on Nextdoor lacked unique identity.

Other key insights emerged from my research:

  1. Many see their home as a refuge to withdraw into after a busy day at work and socializing; the neighborhood takes a back seat to the comfort of the home. Those who expressed interest in Nextdoor were settling down in a neighborhood for a very long time e.g. buying a place.
  2. Many felt that they already had too many things competing for their attention - Facebook, messaging, LinkedIn, Twitter - and did not have bandwidth for another network.
  3. Some Nextdoor features already had dominant competitors that people turned to: e.g. Craigslist and eBay for buying and selling, Meetup and Facebook for interest groups.
  4. The most common use of Nextdoor was for searching for local services: either to find a babysitter, or a house cleaner.
  5. Interviewees found the Nextdoor sign-up process intimidating and would very likely have dropped off at some point

Prototyping and Testing

Sketching and testing low-fidelity mockups allowed me to iterate quickly through numerous ideas.

The majority of my design process is spent scribbling and sketching: I usually dive into actual UI design pretty late. Besides experimenting with UI elements, I fill my sketchbook with notes from my interviews, empathy maps and competitive analyses.

I first created sitemaps for the core actions I wanted users to take, and then thought about where I could combine actions to reduce the number of views needed. After settling on the broad aims of each view, I sketched thumbnails for each view, then used Sketch to create basic wireframes, before developing these to higher-fidelity layouts.

For me, UX design isn't just about layouts: it's about the entire experience of moving through the feature. The flow and transitions are just as important as each visual layout (I love this Frank Chimero piece on the interaction designer's role in designing flux.) I devoted time to thinking through the movement through my designs; I found FramerJS really handy for quickly prototyping certain interactions to get a feel for them on my phone.

Creating UI layouts in Sketch, and tweaking transitions in Framerjs.

Creating UI layouts in Sketch, and tweaking transitions in Framerjs.

A key lesson that I've learnt over my past projects is that designing in a silo without user feedback can produce ineffectual work. I took time to put prototypes in the hands of users to observe the challenges they faced in understanding what the software was asking of them.

Redesigning Nextdoor: The High-Level Proposals

I believe that at a broader level, Nextdoor needs to do a better job getting people talking.

Even though the company has a footprint of over 90,000 neighborhoods, I have encountered many anecdotes of neighborhoods that are silent on Nextdoor. As a social network, Nextdoor is dependent on network effects to succeed: users need to start talking to sustain long-term, broader interest in the product.

As I moved along my design process, I noted down several recommendations below that the company could consider for its general strategic direction:

Target the newcomers to the neighborhood

Nextdoor seems focused on having existing users invite their neighbors to their networks. I believe the people who are newly moving into a neighborhood are ones who are the most excited about the area, and most in need for connections and recommendations. If they have a great experience on Nextdoor, they could become the most active participants over time.

Streamline existing features

Nextdoor feels like a one-stop shop for everything people might want to engage their neighborhood with: classifieds, alerts, recommendations, interest groups etc. I fear that this puts the company in competition on multiple fronts with companies specialized in individual slices of this pie e.g. Yelp for local recommendations, or Meetup for interest groups. It might also detract from areas where Nextdoor could truly shine: as a local forum, for local professional services, and documenting local knowledge.

Cultivate/seed local leaders

My neighborhood has leads on Nextdoor, but they've been silent themselves. Perhaps Nextdoor could play a more active role in urging local leads to start conversations, or even hiring community managers to create discussions. In my observation of neighborhood groups on Facebook, it is often a small group of people who dominate the conversation, but they keep the group looking active.

Build upon the idea of the true local

Why would people engage in activities within their neighborhood? I suspect that the answer might lie in a sense of belonging: they feel great being a true local, knowing the best places to eat or shortcuts around their area. I believe Nextdoor has an opportunities to create this sense of legitimacy through new features, such as perhaps crowdsourcing guides to the neighborhood, or creating some kind of leaderboard of how legitimate a local a user is.

For the actual feature redesign challenge, I chose to focus this submission on redesigning two aspects of the product that could have the largest impact on increasing engagement: the self signup process, and the compose flow for a new post.

Feature Redesign Part I: Onboarding

What if neighborhoods within Nextdoor could feel like they've got unique identities? And can the pressure to push potential users down the funnel be dialed down?

I’ve observed that many NYC neighborhoods have very few locals using Nextdoor. My neighborhood has 181; Dumbo has 132; Chinatown has 58. I’d like to suggest that neighborhoods do not come across as appealing enough during the signup process, and that interested users can come under real pressure through the signup process. As you see below, there are several issues that can be identified within the onboarding experiences across web and mobile.

I therefore set out to create an experience that makes a potential user feel like they are joining a network that has its own identity, while giving users the space to explore and figure out for themselves if they'd like to try Nextdoor.

You can watch the Framerjs prototypes of both the mobile onboarding experience and the web onboarding experience below.


Throughout the redesign of the onboarding experience, I sought to make the user feel at home, or feel welcomed to a neighborhood that had a strong identity. I chose to always keep in mind the following themes:

Unique neighborhood profile

The user feels like she’s getting to know her own neighborhood, full of interesting information such as photos and quotes from other residents, instead of a generic network. A community lead would write a short message to appeal to new users to join the community.

Only necessary information requested

I removed questions like email and gender. The objective is to have a curious person first preview Nextdoor by searching for her address, and then to sign up.

Zero barriers to the first message

The new user gets quickly to her first message, instead of needing to fret about inviting another neighbor. She immediately feels welcome.

Visual appeal

Responding to feedback that Nextdoor felt utilitarian instead of social, I increased the appeal of each layout and transition to bring delight into the onboarding process.

Feature Redesign Part II: Initiating a Post

Users should be able to initiate a post to multiple neighborhoods quickly.

The improved onboarding flow described above is intended to raise the number of users within each Nextdoor neighborhood. This ideally should raise the number of people who can ask questions, contribute to discussions, and give recommendations.

But what if the lack of chatter could be traced to barriers in starting a conversation? I decided to evaluate the post initiation process, specifically targetting Nextdoor's mobile app. I believe that the mobile app presents the ideal opportunity to grow the volume of chatter. With the abundance of social channels today, it is perhaps getting rarer for people to have multiple networks open in tabs on their computers, especially not a network that is wholly focused on the neighborhood. The mobile app, however, is good for circulating and receiving alerts, and asking and answering questions on the fly.

After conducting a quick evaluation of the post initiation flow on the iOS app, I noticed a particular problem area: when a user initiates a post, she has to pick the neighborhoods she wants to contact. This, however, is presented to the user in a list-view. For someone that is not familiar with the surrounding areas, she would have difficulty picking the right neighborhoods to contact.

Having evaluated the existing flow, I decided to redesign the post initiation flow from the homescreen to the neighborhood selection view. I began by shifting the New Message button down to the right corner, nearer to the user's thumb, and thus making messaging easier for one-handed use.

To address the problem of posting a message to multiple neighborhoods, I designed a map view that showed all the adjacent neighborhoods. Users can quickly see which neighborhoods they want to message, and tap the neighborhoods to select them. The info box to the bottom corner keeps track of the number of neighbors that the message will reach.