In an age where we are so connected through social media, how is it that so many people who still feel isolated?
In today’s day in age this has become somewhat of a trite question, but it still holds merit. Over 90 Million adults are reported to experience loneliness. And this is not a stagnant issue. Since 1980, the percentage of Americans who say they are lonely has doubled, and movement away from this trend does not seem likely. This is the problem that myself along with two others set out to solve.
How do you make friends?
To begin handling this issue a simple question must be asked: how does one make friends? After a lot of conversations, and reading research article after research article, we boiled it down to 4 factors: proximity, small group settings, repeat interactions, and having similar interests.
Social media, while it has its place, has failed to meet these criterion. Event hosting platforms such as Meetup and Facebook Events are often used for groups that are far too large to allow for real friendships to be formed and further do not foster repeat interactions. Our goal at Frienli was to create a solution to this problem. To allow people to easily meet others, and to turn acquaintances into friends.
Building a product
Our initial approach to this was to design a web app that would help coordinate local small group experiences and repeat interactions.
The site functioned as a marketplace for hosts and attendees. A host could post an event (i.e. a bike ride, cooking lesson, hike, etc.) and attendees could join. The host could set a minimum number of attendees in order for the event to take place and could also charge for the event, as Frienli included payment processing.
Once the attendees signed up they would be automatically added to a group message where they could chat with other attendees and the host. Furthermore, at the end of the experience, attendees would be given the option to friend other people at the event. This feature would allow them to chat over Frienli, and also let them be notified when their friends were going to other events nearby.
Launching a product
With a usable product, our team began the difficult task of solving the marketplace problem: how can you get hosts without having attendees or vice versa? After listening to one too many podcasts on building marketplaces, we opted to first acquire hosts, as they could be more easily incentivized. We did this by offering to fill any event they created with attendees, guaranteed.
Once we had an event posted to Frienli we tabled, spoke at events, reached out to email lists, and talked to person after person until our events were filled. And so our social marketplace was born.
Iteration 1: Sign up funnel
Once events started rolling out, product and data analysis began, and it didn’t take long to find things that needed to be fixed. Our first problem was that 43% of the people who started the sign up process didn’t finish. Our team originally hypothesized that forcing a profile picture upload was probably what was leading to the massive drop off.
Sign up incompletion instead was primarily driven by the inclusion of phone number verification and the overall length of the signup process. To resolve this issue I went to our web development team and suggested use of Facebook login to shorten sign up time and for the addition of an email verification option in substitute for the phone number verification.
With these changes implemented, we saw a 46% increase in sign up completions in the first month that they were deployed.
Iteration 2: Hostless events?
As time went on our product started running into more serious issues. Our primary concern was that our retention rates were extremely low. While new users continued to join, our original users kept dropping off. As shown below, three weeks after they signed up, almost no users returned to the site.
I went back to work trying to discern what was going wrong. After talking to users and looking at the numbers, the problem became much more clear. There were not enough events being hosted to keep users coming back to our site. And this problem was not going away as fewer and fewer events were being created.
The root cause of this decline in events was fundamentally a result of not having enough hosts, especially relative to the number of users that we had. This can clearly be shown in the graph below.
As our number of hosts flat lined eventually our attendees followed suit, as their were not enough events to sustain our user base. We attempted to solve this problem in many ways whether recruiting more hosts, increasing host incentives, or even piloting a hostless event concept, but each attempt failed.
Lessons worth learning
As Frienli was loosing steam we continued to interview users. The resounding message communicated was that their were limited incentives for the hosting of small scale events, which happened to be the ones that most attendees were interested in. The hosts that we did maintain were specialized and continued hosting due to the high prices of their events, which brought in supplemental income.
Fundamentally, the motivations of our attendees and hosts were not aligned. Attendees wanted cheap events where they could meet people, but there were no hosts willing to do this, as it brought in no revenue. Without any good incentive structure to encourage small scale events, the only events we could attract were ones with high costs.
While this project fundamentally failed, there was much I learned in this process. If given the chance to share one takeaway it would be this: empathize. While we started with a strong understanding of user needs we failed to focus on the needs of the host, which, in the end, lead to the failure of Frienli.