Friendly App
Make new
Friends
Founder
Acquired - Sept 2021
My first solo venture into the world of entrepreneurship took place during my time at Atlassian, while I was travelling back and forth from the Vietnam office and building the JIRA app. During my time in Vietnam, I realised the only way of connecting with fellow expats was through dating apps, which wasn’t right for me.
I created a Kickstarter campaign to seek funding for the project and to see if it was an idea that people were excited about. Having secured funding, I sourced two developers to join the team. From there we focussed on launching our Version 1, which was released 6 months later.
Initially we trialled a focus on “events in your area” but found this approach ineffective. The app attracted many introverted types and as a result not many users came forth to host events. Next, we decided to pivot to a basic social platform, connecting people based on mutual interests. We applied an advanced matchmaking system to source high compatibility pairings.
The app was successful in building a loyal following in Australia, however I soon realised how much funding it takes to make a social app go viral. Ultimately, I decided to let the app grow organically. Meanwhile, it also brought in a very healthy stream of revenue. In September 2021, the app was acquired by a Canadian dating app company looking to merge their app with our ecosystem to utilise our very effective matchmaking algorithms for their user base.
Scope
- Branding
- Native iOS app
- Website
- Marketing
Tools
- Adobe AI (Brand)
- Adobe AE (Animation)
- Figma (UI)
- Miro (UX)
- Maze (Testing)
- Hangouts (Interviews)
Brand
A friendly & gender inclusive brand.
Making friends online is a very new and somewhat peculiar concept to some people, particularly males. Our brand developed from a much more feminine look to something more gender neutral to better appeal to our demographic.
Illustration style & animation
I wanted motion to play a big part in the app, so to make this possible I opted for a flat style with our brand colours in lighter and darker shades. I used purple instead of naturalistic skin tones to make the app feel inclusive to all.
Planning
I wore my PM hat very much in this role, engaging my developers in sprint planning, as well as identifying the biggest impact-versus-development-cost for our Version 1 release. Although the prototype was tested in detail, we also set up event tracking and split tracking of the main features to test our first release from the get-go.
Product Design
Registration
An optimised onboarding experience
Based on user feedback, the greatest challenge we faced in the market was the pre-existing stigma of high-commitment onboarding processes culminating in paywalls. We needed to simplify our 6-step process, although we also didn’t want empty profiles landing on our home screen. We optimised the experience so that only 3 steps were required, which significantly decreased drop-off.
+26%
Finished registration
9
Versions tested
Home & Matches
Smart user discovery with less clicks
The home page and matches page went through many iterations in order to simplify and reduce the cognition and clicks required to find the right users. Initially we sorted matches based on location, but quickly found that this resulted in many inactive users being suggested first, and therefore to a higher bounce rate for the new users. Instead, it was imperative for new users to start conversations on their first day. I created a points system that combined the variables of time since last online, match percentage, and distance away, to create a default sorting option that outperformed the previous options almost fourfold.
+32%
Day one connections
+14%
Requests actioned
Matchmaking
Quality over quantity. Smart matchmaking
Our matchmaking process is where the magic happens and our algorithms kick in to improve recommendations throughout the app. At the start, matchmaking was part of the registration flow, but we progressed to using a “matches” tab to reduce the amount of time it took for users to start accessing the service. We tweaked the questionnaire and looked strategically at specific answer combinations, to the point where 70% of our user respondents reported that they actually had something in common with their matches.
70%
Match success
23
Algorithm releases
Paid Membership
Making the “Freemium” model work
Friendly went live as a completely free service. On Version 3, we released a “freemium” model based on matches, which we discovered was the greatest selling point for our app. We wanted our users to have a free taste of their potential matches first, hence the “freemium” approach, which offered a premium service according to the Tinder model once users had spent a certain amount of time on the app in one sitting. This model also increased stickiness for those users who decided to remain free. To make our premium offering more appealing, we added features such as “Who’s viewed you” and “Profile boosts” that were also accompanied by a time-limited free option.
Use Profiles
Rich profiles for stronger friendships
There are two main influences on our choice of prospective friends: individual factors and environmental factors. Individual factors include approachability, social skills, self-disclosure, similarity, and emotional closeness. Environmental factors, on the other hand, include proximity, geography, activities, and life events. Unlike most apps, we included both these factors on users’ profiles as opposed to superficial imagery only.
Chat & friends
A simplified and easy-to-use chat
Sometimes the best experience is a familiar experience. Opting for an experience very similar to other popular apps, namely Messenger and Whatsapp, allowed us to release a feature that required little to no improvements. Our feedback indicated that our users navigated the experience smoothly. This feature did require particular attention to backend development, however, as these other above-mentioned apps have set the standard for performance expectations.
4
Versions split tested
10
Customer interviews
Iterations
Not everything went to plan. Here are some of the core pages that we developed based on in-depth user feedback and testing sessions.
Core navigation
The first version of our home screen had no tab bar, with core navigation taking place in the header, which made the app feel very easy to use. However, this approach also caused confusion with our users, since it meant that certain crucial information was hidden deep inside other sections of the app. Once we moved to a more familiar tab bar, it made the sections for requests and different friendship modes clearly visible. Following the implementation of the tabs system, there was a significant increase in friend requests and support emails.
+28% Requests actioned
User Profiles
Although the first iteration of the profile page worked well enough, we received feedback from our users that larger profile images were preferred, so that they would not have to enlarge the original images. Users were also confused about the treatment of match percentages, and some overlooked the primary actions, which were lost after they scrolled away from the page. Our biggest improvement came from a frequent request from our male users to see if the males they were seeking to form friendships with were straight. Adding this information increased male-to-male friend requests significantly.
+18% Male to male Requests
Paywall
The first version of our membership paywall was relatively basic, with a set of static graphics that each offered a different monthly payment option. Our goal was to boost the rate of subscription conversion, so I used the Maze platform to test around 8 different options. Eventually I landed on a contextual paywall, which gathers information based on where the user comes from in order to make the graphic and copy more appealing. We launched the 3 different payment options as a test, but found that users liked to see all the available options with the savings provided by each, so in the end we left all 3 in place.
+16% Conversion rate
Takeaways
A great team is a family.
Treat your team like co-founders and in return you’ll have a team who invest more time, detail and love into what they do.
01
You are going to fail.
A Version 1 feature release rarely goes viral. The key is to refine, refine, refine, and not be a “one and done” kind of entrepreneur.
02
It’s okay to walk away.
We marry our ideas, but forget that half of all marriages end in divorce. Don’t get carried away by passion if your statistics are telling you otherwise.