Two years ago, Spark was nothing more than a weekend project. My co-founder Priya and I were frustrated with the messaging apps we used every day -- they felt impersonal, cluttered, and frankly, boring. We wanted something that made talking to the people we love feel fun again. That frustration turned into a prototype, the prototype turned into a beta, and the beta turned into the app that 200,000 people use to stay connected today.
This is the story of how we got here -- the decisions that shaped Spark, the mistakes we made along the way, and what we learned about building a product people genuinely enjoy.
The Early Days
It started with a question: what if a messaging app felt more like hanging out in person? We sketched the first concepts on a Saturday afternoon. No pitch deck, no business plan -- just two people with a shared frustration and a whiteboard full of ideas.
The first prototype was embarrassingly simple. Text messages with colorful bubbles and a handful of custom emoji reactions. That was it. But when we shared it with twenty friends, something interesting happened: they kept using it. Not because it was better than iMessage or WhatsApp in any measurable way, but because it felt different. It felt personal.
Those early testers taught us our first lesson: people don't switch messaging apps because of features. They switch because of feelings. The emotional experience matters more than the feature checklist.
Finding Product-Market Fit
For the first six months, we built and shipped a new feature every two weeks. Most of them flopped. A few stuck. The ones that worked all shared something in common -- they made conversations feel more expressive:
- Emoji reactions with animation -- not just a thumbs-up icon, but a bouncing, spinning celebration that made people smile
- Voice memos with waveform visualization -- seeing the shape of someone's voice made messages feel more intimate
- Group polls and plans -- deciding where to eat dinner shouldn't require a 47-message thread
- Mood-based themes -- your chat background shifts with the tone of the conversation, from warm and cozy to excited and energetic
We knew we had something when our retention numbers stopped looking like a leaky bucket. By month eight, 60% of people who tried Spark in their first week were still using it thirty days later. For a messaging app competing with giants, that was extraordinary.
"We didn't set out to build the best messaging app. We set out to build the most human one. Every feature we ship has to pass a simple test: does this make the conversation feel more like being together in person?"
Building the Team
Priya and I ran Spark as a two-person team for the first year. She handled design and frontend; I handled backend and infrastructure. We hired our first engineer, Jordan, when the server started crashing during peak hours. Jordan's first task was fixing the WebSocket architecture, and they did it so well that we still use a version of that original design today.
Hiring at a startup is an exercise in trust. We looked for people who cared about craft over credentials. Our best designer came from a game studio. Our lead backend engineer was self-taught. What they shared was an obsession with getting the details right -- the 60fps animations, the sub-100ms message delivery, the pixel-perfect alignment.
Today we're a team of fourteen, and every person on the team still ships code. That's not a flex -- it's a survival strategy. When everyone builds, everyone understands the product deeply enough to make good decisions fast.
Scaling to 200K
Our growth was never the hockey-stick chart investors dream about. It was more like a staircase -- long flat stretches punctuated by sudden jumps. Each jump came from the same source: people telling their friends.
We never ran paid ads. Our entire growth engine was word of mouth, driven by three deliberate choices:
- Make sharing delightful -- when you invite someone to Spark, the invite itself is an experience, not a generic SMS link
- Reward groups, not individuals -- features unlock when your group reaches milestones, creating natural incentive to invite
- Respect attention -- we send fewer notifications than any competitor, and users trust us because of it
The biggest staircase jump came when a college student posted a TikTok about our group poll feature. The video got 2 million views, and we gained 40,000 users in a week. Our servers held. Barely.
What's Next
We're working on features we can't talk about yet, but I can share our north star: Spark should feel like the digital equivalent of your living room. Comfortable, expressive, and uniquely yours.
The messaging market is dominated by apps built for scale, not for soul. We think there's room for something different -- something that treats every conversation as an experience worth designing for. Two years in, 200,000 users later, we're just getting started.
Thanks for being part of the journey. If you have ideas, feedback, or just want to say hi, find me on Spark -- I'm @alex. I read every message.