Neemz@Product
Published in

Neemz@Product

Software is eating the world” — Marc Andreesen

How We Launched a Social Network for Neighborly Interactions

Every product needs an initial insight to get started with, and when that insight is found, it needs to go through a series of hypothesis-driven customer discovery experiments. This is the story of how we built a product that brought neighbors and local stakeholders together and helped them build social capital using technology.

Initial insights and market validation

In September 2018, our payments startup had scaled from 15 full-time employees to 65, and with an office in one of the city’s most crowded and densely populated locations, very quickly, parking spaces close to the office and in a reasonably safe and legally viable location were short in supply.

The options employees had were not the best solution. They had to:

  • Arrive at the office in the early morning hours before the competition, or
  • Risk parking in an illegal location and get fined, or
  • Park within a 20–30 minutes walking distance from the office, or
  • Not bring a car to the office—which wasn’t a long term solution

What we knew, however, was that the residents in the neighborhood would leave their private parking spaces vacant during the daytime when they were out for work, and our staff were more than willing to rent the spaces on an ongoing basis. To start conversations, we spoke to local businesses in the neighborhood as to whether they could vouch for the team, build trust, and find private parking spaces for the staff to use.

Within a very short time, we found out that several neighbors were willing to rent out their parking spaces to ‘our employees ‘— and the key was ‘to our employees’, due to the established neighborly trust.

To manage the parking space rental interactions, including, finding out which neighbor was available in the coming day(s), communicating and arranging meetups, and paying for the transaction, we started a Telegram group which was the mainstream messaging platform at the time.

Sharing parking spaces was just the beginning

As the members of the Telegram chat group expanded to more than 100 users, mainly through peer-to-peer invites and local business referrals, the number, and frequency of interactions went beyond only parking sharing and expanded into a broader scope of local collaboration needs, including:

  • Sharing rooftops, barbeque spaces, swimming pools, and party rooms
  • Asking and making recommendations
  • Food and grocery ordering and delivery
  • Security-related conversations
  • Offering help during a stormy night when a tree had fallen onto a neighbor’s driveway
  • Mentioning found items
  • Charity events and non-profit advocations

At this moment, we realized that the Telegram group was no longer a convenient place for us and the neighbors to conveniently manage our local and neighborly interactions. The product that we were using, was not designed for this purpose. And that was when the requests from the neighbors started to come in:

“Since you guys are tech-people and software developers, can you suggest or build something that would help us manage our interactions?”

Lack of adequate substitutes

The problem was that there weren’t any good substitutes out there that provided the perfect “neighborhood” management tools.

  • Telegram was built for mass communication through its channels and large group sizes
  • Instagram was just a different social platform, built for photo-sharing among people close to each other
  • Twitter and LinkedIn didn’t have the functionality and design purpose
  • Facebook was filtered and didn’t have the neighborhood features it does today
  • Other messaging platforms were less flexible compared to Telegram including Whatsapp, Line, WeChat, and Kakao
A visual map of the neighborhood members to build acquaintance, trust, and transparency

Defining a ‘hedonic’ North Star

When we compared the interactions among the neighbors in the workplace to the social interactions we experienced in our neighborhoods, it left a lot to be desired. Therefore we set out to improve the lives of ourselves and our neighbors in our local communities and set a mission for ourselves:

We help build better neighborhoods for all to live in

This statement wasn’t going to be set in stone forever, but it became a guiding north star for our small team of 1 product manager, 2.5 developers, and 1 product designer to strive towards as a side venture before we could measure and validate demand for further investments.

Neighborhood watchdog

Keeping it human before becoming all ‘techy’

In the initial days of deciding to imagine a solution for the problem, it became very natural for a technically experienced team to dive into browsing the array of products in the marketplace and benchmarking possible solutions’ features and offerings. But this was just creeping bias into our efforts into finding product-market fit.

To better understand (i.e. utility and transactional needs) and feel for (i.e hedonic needs) what we were solving, we had to become human and go to the customers and their demands. We decided to think of the new product as:

Something that we were building for ourselves and for the small community of neighbors at the workplace

With an active sample of 300–400 engaged users on the Telegram group, we decided to identify pains, prioritize them, build a deserving value proposition, and pivot based on their feedback on paper until they were ‘fully’ satisfied. And I say ‘fully’ because we didn’t want to get into any major development work before we had the value proposition nailed down — it’s just more cost-effective and emotionally less painstaking.

Initial survey: quantifying scale and urgency of needs

Our initial questionnaire entailed a demographic segmentation questionnaire to better understand what various segments desired, along with a motivational and psychographic survey that focused on 5 categories of neighborly interactions including:

  • Communicating with public organizations and services
  • Improving co-living cultures and interactions
  • Collaborating to enhance local security
  • Building and delivering upon collective interests
  • Socializing actively and objectively, locally

Insights

Through referrals and shared invites, more than 3000 people participated in the survey, and below are the main insights:

  • All the main categories mentioned above were important to local living standards across all customer segments
  • ~3 out of 4 (75%) participants knew less than 5 neighbors in the neighborhood by name
  • ~4 out of 5 (80%) participants indicated a low level of satisfaction with their neighborly relationships and desired improvements in the frequency and quality of interactions with one another
  • ~4.5 out of 5 (90%) participants indicated that they had no source for neighborly news and content
  • ~3 out of 5 (60%) showed a large desire to attend local events
  • Having used Facebook, Instagram, Telegram, and Whatsapp, ~4.5 out of 5 (90%) participants trusted that similar technology tools could be used to improve neighborhood relationships
  • Only ~15% of the participants believed that current products could sufficiently help them build local online communities
  • Building trust and delivering on privacy concerns would be key for members joining a local neighborhood group/network
  • Female participants had a stronger desire to improve neighborhood relations than male participants
  • New residents of a neighborhood showed a higher desire to socialize locally
  • Married couples showed higher interest in neighborly interactions compared to those who were single

Qualitative interviews: discovering the underlying psychological pains, problems, and desires

With the survey validating demand and providing a priority of features for us to focus on, we needed to uncover what psychological drivers were creating demand and how a potential product would need to make consumers feel before, during, and after using it. And that’s why we embarked upon face-to-face coffee chats with 10 of the participants from the early adopter, early majority, and late majority to understand the personality archetypes we were dealing with and their emotional needs.

The main insights were:

  • All believed that a society that fosters close neighbor relationships is a healthy society with large degrees of social capital and civic engagement. This was a common string of stories that the participants remembered and told from their childhood which they no longer possessed and at times, yearned for
  • Those with public sector backgrounds, such as those working in the municipality or the police force, indicated that closely knit and interactive neighborhoods were easier to collaborate with on public projects and safer to live in, resulting in higher real estate values
  • Life in neighborhoods with high social capital is perceived better due to high levels of social trust with existing norms of socially beneficial reciprocity, delivers amplified reputation and status, and possesses a feel of collaboration rather than competition with a high degree of volunteer work and help
  • Observations of participants were that it wasn’t just neighborly connections that had deteriorated, but also familial ties had declined over the last 3–4 decades, which could indicate a macro-social change beyond just neighborly interactions. Which could have been a massive barrier to finding product-market fit if it was the absolute truth
  • Participants highlighted several thoughts as the driver of the change and erosion of social capital over the last decades including the entrance of women into the workforce, whether willingly or circumstantially driven by economic needs and lower real wages, lack of time to socialize due to a large share of daily time spent commuting, a seismic shift in social bonding beliefs including a decline in marriage and increase in divorce rates, fewer children per household, and the technological transformation of the entertainment industry from broadcast Radio and TV to smartphones

While we weren’t discouraged, this is mainly the output of most qualitative interviews. A reality check and deep dive into the extent of the problem and how deep-rooted it could be. It was obvious that if we are to make any change and deliver value to customers, we would need to build an ecosystem of various players including the public sector service providers and local activists, local businesses, NGOs, and the residents coming together to build upon each others’ social capital to benefit and stick around.

Neighborly news and updates

Building the MVP

Napkin testing possible high-level design flows

Rough sketches we discussed in the early days during 1:1 interviews with neighbors to gauge their understanding of the product and its utility and hedonic offerings

From day-1 we had rough sketches and mindmaps of the flow and the potential feature of the product and would cross-examine their stickiness at any touchpoint that we found the chance to interview and interact with users. This was key as we could gauge their reactions, and feelings and have a rough feel for the NPS or potential for referrals by the target customers which would have been key to growth.

Feature prioritization for mockups

From our quantitative surveys and direct interviews, we had pinpointed our early adopters:

  • Female
  • Married
  • 25–40 years old
  • Tech-savvy and varied in life interests
  • Extroverted and socially conscious
  • Homeowners
  • Had been living in the neighborhood for more than 10 years
  • Known/popular in the neighborhood
  • Held a neighborhood responsibility/role

We built a list of ~30 features that we thought could add value to neighborly interactions, and with the early adopters in mind, set out to identify the priority of the features. We quickly figured out a couple of experience design principles:

  • The early adopters, since they were mainly women, had habitualized using Instagram, and a similarly intuitive design was desired
  • Personalization and relevance of content on the newsfeed would be key to enhanced communication, engagement, and retention
  • Compared to those on Instagram, conversations and posts on a neighborhood product were quite transactional and utility need-based, hence conversations needed to have structure and be around specific categories and topics
  • In the long run, anonymity and fake user profiles would be a no-go in a somewhat private and closed community social network
  • Interaction with local businesses and public entities would be essential to local communication and the platform’s value proposition

Delivery of MVP in 5 milestones

With the goal of quick iteration with market feedback, we launched our initial prototype after 6 weeks and 3 development sprints with 2.5 developers, 0.5 designers, and 1 product manager.

However as suspected, the quality of the product and the scale of features it had to offer was not on par with what other substitutes were delivering — in other words, the new product did not offer any differentiated advantages or unique selling points to wow users and detach them from their Telegram habits. Something we had anticipated.

So we broke the MVP into 5 milestones that delivered the following main features in 16 weeks of development:

  1. Sign up
  2. Newsfeed
  3. Local businesses
  4. Messaging
  5. Profile and administration rights

1. Sign up

The signup process was built of two main subsections including:

  • User registration and profile building
User registration and profile building designs for a neighborhood social network
User registration and profile building design
  • Creating a new or joining an established neighborhood
on a neighborhood social network
Creating a new or joining an established neighborhood designs

Since there was no objective definition of neighborhood boundaries, and different people define their neighborhoods differently, to scale, we gave members the ability to build their boundaries and communities associated with them — feeling fully in control, empowered, and motivated to bring their local communities together.

New neighborhood login experience
Logging into an established, private, neighborhood with an invite code

2. Newsfeed: structured for the occasion

The newsfeed was a complicated feature as it drove engagement and retention rates. To maximize engagement the newsfeed was designed to have:

  • A structured posting regime that clarified topics of conversations with a quick conversation filter for ease of access
Initial newsfeed designs for local neighborhood social network
Initial newsfeed designs

We decided to have 7 categories of posts that brought meaningful conversations and a sense of belonging to each neighborhood:

  • Urgent Alerts: for safety and local notification purposes, very similar to a local amber alert, for example, notifying a car blocking an entrance of a house, burst water pipes, or detour due to maintenance and reconstruction
  • Crime and safety: which was a high priority across all segments
  • Polls: this was a massive ask from public figures, local influencers, and residential and high rise managers
  • Events: to build awareness and notify neighbors for offline gatherings, mainly driven by public agencies and local leaders
  • Local classifieds: to create a conveniently quick exchange medium with someone down the street, rather than try to replace the ‘Craigslist’s of the world
  • Recommendation: local requests and Q&A related conversation
  • Lost and found items
  • General topics

Since content created downtown would have little meaning for someone living 30km to the North of the city, we defined a geographic reach limit of 5km in radius to improve engagement with the newsfeed. In action, we gave users the option to choose neighborhoods within the 5km radius where they wanted their content to be seen and be interacted with.

3. Local businesses: possible point of differentiation

This idea had gotten kicked off through our interaction and trust-building intermediaries and partners, the local businesses. We also knew that neighbors wanted to be able to interact with their local organizations and service providers in a more humanly fashion compared to for example on Google Maps. We tested several hypothetical features and benchmarked them against other substitutes to discover users’ key local business needs, but the key point of differentiation was that:

  • Neighbors and their local service providers desired an interface whereby they could communicate daily, get notified of recent developments including discounts and public opinions, and have businesses reach out to them in a professional local-feel fashion such as seeing a post about discounts on their local grocery store or updates on project developments. This was a feel and interaction that neighbors had not experienced before and wanted to see more of

4. Messaging: a must-have, but not required to be Whatsapp

Like any other social network, the main goal is to remove information asymmetries and ease human communication on a 1:1 and Many-to-Many basis. Now while this was a must-have, due to the nature of the interactions which was more of a functional interaction among neighbors compared to an emotional one with loved ones, the interface and quality of the messaging feature did not have to meet the experience of mainstream messaging platforms such as Whatsapp or Telegram. There were three key messaging requirements:

  • Direct 1:1 messaging: which was commonly used to transact classifieds
  • Private groups: that were used by small apartments complexes and condominiums to manage interactions — there was huge potential for further feature development in this segment of the messaging services in areas such as collecting maintenance fees, issues tickets, digital signatures, etc.
  • Public groups: which were used to build local communities of interest such as weekend soccer and volleyball tournaments, recycling initiatives, charity events, etc.

5. Profile and administration rights: the key to a localized feel

Unlike a global social platform such as Twitter where users can anonymously post any content with few restrictions and get into brawls over politics, a local social network needed privacy, real members, and moderation according to the needs and policies of each neighborhood and community. What made this complex was that the cultural and communication norms of an uptown neighborhood could be quite different from that of a downtown one and the product design needed to consider these variations.

To build a flexible product that would scale, we build administration and monitoring tools for the leaders of the neighborhood to set their respective rules and policies, giving control of the culture of the neighborhood on the platform to the users, while only stepping into their governance policies upon request and if high on demand.

The login structure, newsfeed, messaging, local businesses, and admin rights of the platform
Neighborly transactions

Launch: iterative pivots towards finding market-fit

Initial neighborhood: discovering engagement

While satisfied with the product’s usability and customer acceptance ratings, we used the neighborhood at the office for beta-testing and looking into users' consumption and interaction patterns. Within 3 months of observation, we delivered more than 20 iterations of the MVP to the controlled group and gradually migrated them from Telegram to our product.

Across the 3 months we tracked the following metrics, on a daily basis:

  • Number of users
Total numbers of users — the jumps indicate a new and adjacent neighborhood joining organically  on a local social network platform
Total numbers of users — the jumps indicate a new and adjacent neighborhood joining organically
  • Number of organic daily posts
Daily number of new posts on a local social nework platform
Daily number of new posts

The note here is that our content team was also pushing interesting and differentiated local content onto the newsfeed to keep people engaged. During early weekdays, and considering the number of posts we’d push, users would spend sometimes hours on the product catching up with local news, content, and updates. While this is not something that we could do forever and at scale, it was something that needed to be done for us to gauge user behavior and habit formations, and kickstart the cold launch problem.

  • Daily engagement: during the testing phase we had seen a range of daily engagement patterns including people spending hours on the platforms. However, on average, 12–15 minutes daily time was the norm towards the end of the 3 month period which had grown from the initial 5–7 minutes at the end of the first month
  • Post response speed
Response to a post within the first hour — as the network scaled, the response speed increased; in the initial weeks, our notifications were not working properly either. Another metric that we tracked was the response rate within the first 24 hours on a local social network platform
Response to a post within the first hour — as the network scaled, the response speed increased; in the initial weeks, our notifications were not working properly either. Another metric that we tracked was the response rate within the first 24 hours
  • The number of 1:1 messages: the general usage of this feature was quite low, and rightfully so as neighbors were not here for 1:1 conversations, but the newsfeed and daily updates in regards to their neighborhoods
  • The number of Many-to-Many messages: when we launched the product, considering that we didn’t have enough members, the usage of groups was zero! However, in the last month of the beta-test phase, we built public groups of interest that served across multiple neighborhoods, with potential options for local offline events, and the indications revealed that these groups need to be curated by our team and have potential in bringing people in neighborhoods together
  • The number of businesses: due to the nature of the initial neighborhood as a test group, we neglected to focus on local businesses, which would become a weakness as we went after other neighborhoods
  • Viral coefficient: while this was just one metric amongst a pool of others, our coefficient was not that great compared to other social platforms due to the nature of neighbors not knowing each other. This meant that we will need to invest in onboarding and invitation programs that would help kickstart growth
  • Cohort retention rates after 28 days: the figure stood at ~63%, but this was just too high and was inherent to the nature of the pilot neighborhood. After launching, in other neighborhoods, the figure stood at ~40% after 28 days

Pivots from the pilot launch: user scale through building trust

The pilot launch was a great starting point as it had a niche and receptive user base ready for experimentation. However, as neighbors and businesses of adjacent neighborhoods were inviting members onto the platform, the metrics showed a significant drop in performance. Furthermore, public services and local businesses needed their communication to reach beyond immediate neighborhoods and in a more targeted fashion. Hence we added more features and made some pivots based on the feedback.

  • We added a demo version for people to check out the platform at scale before fully committing to ‘another’ social network
  • Removed the restriction to become a validated user before signing in and allowed 30 days till losing posting and communication abilities
  • Provided admins of the neighborhoods with a wide variety of tools to validate new members including referrals and private entrance codes. The private codes gave the network a feel of exclusivity that boosted the perception of its value proposition
  • Delivered monitoring and reporting tools to all members of the platform to report and mute members
  • Integrated easy sign-in using mobile numbers and Google and Facebook accounts
  • Built systematic daily and weekly summary notifications for notification in the neighborhood
  • Added a visual social map of where neighbors were in their neighborhood to bring visual transparency to the neighborhood

Businesses were key to growth: hacking scale

When moving from the pilot neighborhood to new ones, we hit a massive problem — when only 3 out of 4 people know less than 5 neighbors by name, this made the invitation and word of mouth virality of the network very challenging.

Unlike on Facebook and Instagram where one can invite friends and family through easy access to their smartphone’s contact list, or on LinkedIn via professional email address, on a hyper-local social network where no one knows each other and users are onboarding to get acquainted, viral and low costs growth becomes a challenge.

To grow, we went to the basics of our starting story, and very similar to the parking problem, we used the local businesses and local mom-&-pop businesses to invite new users. In addition, we approached local community centers for introduction to local leaders and influencers and onboarded them to become our word of mouth.

However, to make this work, we had to incentivize our growth partners including the local businesses, public organizations, NGOs, and local influencers with rewards including specially designed ads and campaigns for long periods some lasting as long as 12 months and at times providing admin rights to some local leaders to buy their commitment towards our growth.

Fundamentally, these growth partners were willing to help us grow because they saw a differentiated value in the partnership — and this was something we hadn’t planned for in our initial product design work as we were optimizing experiences for residents, which was a mistake. Unlike Twitter or Instagram, a local social platform offers local network effects. While you can scale the reach of your business on Instagram, the interactions are less meaningful because your followers could be from anywhere around the world, while the local pub, grocery store, or public sector leader needs the direct attention of its local constituents and stakeholders.

Launch strategy was critical for sustained retention

Our data indicated that it would take a neighborhood between 14–28 days to scale till sustained network effects kick in and user retention reaches ~40%. And this period was a critical period that needed care and attention. Data showed that we needed a minimum of 30 members with 6–9 active participants that post content and respond to others.

To help catalyze growth, we had two dedicated launch teams:

  • Operations on the ground: who would be going into neighborhoods, onboarding the growth partners. They would look into the neighborhood dynamics including the scale of the early adopters, the willingness of public entities to work with us, the number of local influencers and their public relations needs, and the fragmentation of the local businesses. The team's work would kick off 2 to 3 weeks before the push to invite neighbors where they would negotiate collaboration with the partners and tie them in before the launch.
  • Content team: the content team would supply engagement materials including sharing local news, urgent alerts, events, signaling charity gatherings, etc. so that new joiners would grasp the benefits of the network upon landing. The ‘on the ground’ team would also provide them with key local stakeholder contact for the team to broadcast their content on the platform. This would continue for ~3 months until we felt the network had enough scale of engagement for organic user content to ensure engagement and retention going forward

--

--

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store
Nima Torabi

Nima Torabi

1.3K Followers

Never not learning, always growing — nothing is written, and nothing is not written, the journey is the goal