We developed this idea while searching for a more effective way to connect with peers who share similar goals and interests. Being alone on a journey towards an important goal is extremely painful.
The online social space is occupied by vast networks hosting millions of individuals, which have often lost their personal and human touch. We believe in the potential of creating human-sized online communities, akin to your local church group, sports club, or art class.
We felt that among the available social interactions - feed, stories, posting board, live chat - the latter is the most impactful.
We’re all part of various group chats with friends, maybe centered around events or interests, or simply with our family. They’re often the most valuable way to stay connected.
This is the cornerstone of Meerchat.
We've grown the platform to 7,000 users, with a 1D of 50%+ and a 30D of 20%.
Our primary hypothesis was that for micro-communities to thrive, we needed to remove inactive users and continually refill the groups to maintain the optimal number of users, thereby constantly maximizing the group's value.
Each group size functions and behaves slightly differently, depending on the interests and demographics of its members. Boomers and moms appreciated its simplicity, the intimacy of small groups, and the ease of meeting new people. In contrast, Gen Z members sent 10 to 50 times more messages and preferred larger groups and a more chaotic environment.
Activating the groups initially requires a significant flintstonning effort. Once a group reaches 15-20 users, it becomes self-sustaining and no longer needs this initial manual flintstonning. But this means we inevitably lose those first users.
Initially, it was challenging to grow many interests simultaneously. Mid-way we pivoted to prioritizing the most active one, tailoring the entire app around it and we managed to double our user base in less than three weeks, jumping from 3,000 to 6,000 users.
At one point, we concentrated on increasing Day 1 to Day 7 retention for a month, which we achieved. We nearly doubled our Day 1 and Day 7 retention rates and the number of notifications enabled. However, this wasn't enough to retain users beyond Day 21.
User interviews and retention data suggest that we've addressed the needs of those who stayed, but the feedback was generally just 'good' or 'okay.' However, 'Okay' isn't enough; we need to achieve higher retention and organic word-of-mouth.
Since the product was somewhat seen as a commodity (a group chat app), it failed to generate a significant 'waoh' effect, resulting in extremely limited viral impact and word-of-mouth. Although we successfully attracted a substantial number of new users through Reddit and then TikTok, sustaining growth without a stronger R coefficient was challenging.
Ensuring users enable notifications is crucial on a social messaging platform. Early retention is disastrous without notifications, particularly when the 'aha' moment might not occur on day one. It's essential to build viral loops from the design stages, or at least to engineer those 'aha' moments and have a pre-defined viral distribution strategy with highly sharable content.
No amount of additional polishing or features will be effective if the core feature doesn't engage users or fails to function properly. While users requested more communication methods (like GIFs and reactions), adding these had no significant impact, as the problem was already solved with just text and images.
On the technical side, we faced challenges with major updates, occasionally breaking the app for users who didn't update. Implementing a built-in method to compel users to update is crucial in the long run.
Choosing Flutter for immediate cross-platform distribution was an interesting decision, and we believe it was the correct one, as our user base was evenly split between both app stores.